Well, designs do make a difference, that is a given. But I don't know what Wang and Jim have heard in the past to state that KT88s can't soundly beat DHTs. This assumes one wants accurate sound, not colorations. And horns simply weren't as good as other types I heard at the CES and T.H. E. Show. Only one or two sounded good but the air motion transformers and ribbons were simply better. Horns were simply distorted, which is the case when one increases the bandwidth the horns had to reproduce. Horns must be limited in bandwidth to decrease distortion (see Radiotron Designers Handbook)
I think part of the problem stems from what parts are used with DHTs vs IDTs. The use of the parts necessary for DHTs will make the KT88s sound poor. (These seem to be the old type of designs.)
However, when using sonically neutral parts, as one should since there are so many of them in a design, the inner detail of IDTs really shines and sounds great, blowing away the DHTs as I heard at Alexis park and St. Topaz.
I have never heard a "steely" sound from Svetlana KT88s when the correct neutral sounding parts and designs are used.
As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used, to a greater degree than if one is using the best in neutral parts. This cannot possibly result in a truly neutral sound as one cannot completely compensate for so many sonic problems these parts generate.
Another problem is the grid drive necessary to the output tubes. I, personally, don't especially care for any of the tubes available. If a company were again to manufacture a rugged good sounding 8417, but with 50 watt plate dissapation, it would only take half the drive of a 6550/KT88 or 2a3, and about 1/4th that to drive a 300b. That would reduce drive requirements and I think help in the quest towards a better product.
Minimum drive requirements of the output tubes would result in the previous stages producing little distortion. There would be very little mixing of distortion products from the previous stages to the output stages, thus minimal higher harmonics. This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube. This clearly negates the advantages of the low distortion output tubes. Interstage transformers may help, but limited bandwidth, poor frequency response, and variable distortion vs frequency occurs.
Finally, I prefer the best amp and then choose the right speaker, not try to fit the amp to a speaker. Some speakers are so far out in left field that no amp will properly compensate. And do we need an amp that is so colored just to match to a particular speaker that is so colored?
Steve
SAS Audio Labs
I think part of the problem stems from what parts are used with DHTs vs IDTs. The use of the parts necessary for DHTs will make the KT88s sound poor. (These seem to be the old type of designs.)
However, when using sonically neutral parts, as one should since there are so many of them in a design, the inner detail of IDTs really shines and sounds great, blowing away the DHTs as I heard at Alexis park and St. Topaz.
I have never heard a "steely" sound from Svetlana KT88s when the correct neutral sounding parts and designs are used.
As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used, to a greater degree than if one is using the best in neutral parts. This cannot possibly result in a truly neutral sound as one cannot completely compensate for so many sonic problems these parts generate.
Another problem is the grid drive necessary to the output tubes. I, personally, don't especially care for any of the tubes available. If a company were again to manufacture a rugged good sounding 8417, but with 50 watt plate dissapation, it would only take half the drive of a 6550/KT88 or 2a3, and about 1/4th that to drive a 300b. That would reduce drive requirements and I think help in the quest towards a better product.
Minimum drive requirements of the output tubes would result in the previous stages producing little distortion. There would be very little mixing of distortion products from the previous stages to the output stages, thus minimal higher harmonics. This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube. This clearly negates the advantages of the low distortion output tubes. Interstage transformers may help, but limited bandwidth, poor frequency response, and variable distortion vs frequency occurs.
Finally, I prefer the best amp and then choose the right speaker, not try to fit the amp to a speaker. Some speakers are so far out in left field that no amp will properly compensate. And do we need an amp that is so colored just to match to a particular speaker that is so colored?
Steve
SAS Audio Labs
Hi,
if colouration is a problem with 300b's how about this beast as used by cdeveza
(drd amp)
http://www.diyaudio.com/forums/showthread.php?s=&threadid=13525&perpage=15&highlight=&pagenumber=2
it's the inherent simplicity of the circuit that appeals to me, albeit maybe not to the liking of everyone else. If I do build another amp I think this may well be it. I searched the web quite extensivley and also toyed with the idea of hybrid designs, but still have not found anything as appealing as this.
Thanks
Raja
if colouration is a problem with 300b's how about this beast as used by cdeveza
(drd amp)
http://www.diyaudio.com/forums/showthread.php?s=&threadid=13525&perpage=15&highlight=&pagenumber=2
it's the inherent simplicity of the circuit that appeals to me, albeit maybe not to the liking of everyone else. If I do build another amp I think this may well be it. I searched the web quite extensivley and also toyed with the idea of hybrid designs, but still have not found anything as appealing as this.
Thanks
Raja
Steve,
Unfortunately, as a general rule of thumb, the higher mu output valves have higher distortion, so you're back to square 1.
The mixing of distortion products is a very interesting one. It's a good reason to have as few stages as possible. Just recently, I've been in communication with a Swiss enthusiast who's going the single stage route.
Of course there will still be other stages - even the loudspeaker where the curves will add and multiply.
I don't find a lot of common ground with your posts, but at least we both have a common view of the problems, even though we tackle them in different ways😉
Konnichiwa,
Well, to list:
I have heard the EL34, KT88 & 6550 triode wired in DIY SE Amplifiers generally similar in parts quality and execution to my DHT Designs, plus many commercial designs, few of these SE admittedly.
But I have the idea that the DHT Amplifiers you have heard are the usual commercial stuff (I will refrain from naming names here) which are in many cases severely coloured, by design.
I do not what you heard or not and to what degree your perception of neutrality is same as mine (my "neutrality" comes from "comparable to a wire bypass"). However, conventional Cone/Dome type speaker sare completly unsuited to high fidelity reproduction in domestic settings, full stop. The reasons are manifold, from excessive compression and distortion over unsuitable dispersion pattern and unfavourable room interactions. If you consider such types of speakers as "neutral" than obviously a truely neutral a really neutral speaker (such as a really good main studio monitor) would of course not strike you as being neutral, not that many domestic "Horn" speakers approach neutrality.
So, the parts I use (all of which have to well in a "wire bypass" test make the KT88 sound poor? Sorry, but I like neutrality.
Yup, unless one uses Valves that are from the start comparably neutral.
Is it? I find engineering a driverstage for the 300B with > 6db headroom and tons drive using a single valve quite trivial, but that's me and WTFDIK.
A well engineered driverstage does that anyway. And with 50V+ RMS output. It really is trivial.
Yes, this is the way I enginner my designs.
I have no such problems up to around 80V RMS output grid drive, above that things get hairy.
I prefer to start from a very good speaker, as this makes the choise of a matching amplifier much less problematic, as more options are open. Very few speakers can be considered to provide "HIGH FIDELITY" under domestic conditions. And what is the value of the "BEST AMP IN THE WORLD" if the "best music" is then mangeled by these poor excuses for speakers so often found in this day and age. That is of course for the customer, the benefit to the Amplifier manufacturer are obvious....
Sayonara
Positron said:
Well, to list:
I have heard the EL34, KT88 & 6550 triode wired in DIY SE Amplifiers generally similar in parts quality and execution to my DHT Designs, plus many commercial designs, few of these SE admittedly.
But I have the idea that the DHT Amplifiers you have heard are the usual commercial stuff (I will refrain from naming names here) which are in many cases severely coloured, by design.
Positron said:
I do not what you heard or not and to what degree your perception of neutrality is same as mine (my "neutrality" comes from "comparable to a wire bypass"). However, conventional Cone/Dome type speaker sare completly unsuited to high fidelity reproduction in domestic settings, full stop. The reasons are manifold, from excessive compression and distortion over unsuitable dispersion pattern and unfavourable room interactions. If you consider such types of speakers as "neutral" than obviously a truely neutral a really neutral speaker (such as a really good main studio monitor) would of course not strike you as being neutral, not that many domestic "Horn" speakers approach neutrality.
Positron said:
So, the parts I use (all of which have to well in a "wire bypass" test make the KT88 sound poor? Sorry, but I like neutrality.
Positron said:
Yup, unless one uses Valves that are from the start comparably neutral.
Positron said:
Is it? I find engineering a driverstage for the 300B with > 6db headroom and tons drive using a single valve quite trivial, but that's me and WTFDIK.
Positron said:
A well engineered driverstage does that anyway. And with 50V+ RMS output. It really is trivial.
Positron said:
Yes, this is the way I enginner my designs.
Positron said:
I have no such problems up to around 80V RMS output grid drive, above that things get hairy.
Positron said:
I prefer to start from a very good speaker, as this makes the choise of a matching amplifier much less problematic, as more options are open. Very few speakers can be considered to provide "HIGH FIDELITY" under domestic conditions. And what is the value of the "BEST AMP IN THE WORLD" if the "best music" is then mangeled by these poor excuses for speakers so often found in this day and age. That is of course for the customer, the benefit to the Amplifier manufacturer are obvious....
Sayonara
Konnichiwa,
Implemented well the DRD/Monkey principle makes for some of the most transparent and neutral amplifiers I have heard, SE or PP.
Sayonara
Raj1 said:
Implemented well the DRD/Monkey principle makes for some of the most transparent and neutral amplifiers I have heard, SE or PP.
Sayonara
Drive voltages
Of course in the commercial world, no one would dream of supplying the driver stage with a higher HT voltage than the output stage; it'd be too expensive. That's where DIY really wins: We can use the best output valves, regardless of mu, and make low distortion drivers.
Of course you could just put ndB of feedback to "fix" all the ills
Of course in the commercial world, no one would dream of supplying the driver stage with a higher HT voltage than the output stage; it'd be too expensive. That's where DIY really wins: We can use the best output valves, regardless of mu, and make low distortion drivers.
Of course you could just put ndB of feedback to "fix" all the ills
Just like lightbulbs, if they're made properly, they only age in use.
Similarly, it depends on how hard you use them.
Small signal valves last for years. Sometimes the equipment's lifetime.
Power valves are another story. They are often used close to their limits, and will last much less time. It does depend on a lot of different factors, but I'd budget to replace normal output valves every 2 ~ 3 years.
Some will last longer, some even shorter. It all depends..
Similarly, it depends on how hard you use them.
Small signal valves last for years. Sometimes the equipment's lifetime.
Power valves are another story. They are often used close to their limits, and will last much less time. It does depend on a lot of different factors, but I'd budget to replace normal output valves every 2 ~ 3 years.
Some will last longer, some even shorter. It all depends..
Cleared some things?
There are some problems I need to address. First Dhaen.
"Unfortunately, as a general rule of thumb, the higher mu output valves have higher distortion, so you're back to square 1.
The mixing of distortion products is a very interesting one. It's a good reason to have as few stages as possible. Just recently, I've been in communication with a Swiss enthusiast who's going the single stage route."
I don't find a lot of common ground with your posts, but at least we both have a common view of the problems, even though we tackle them in different ways."
Unfortunately, this general rule is not true. 6550/KT88s have much higher transconductance than EL34s, yet have much lower distortion. 2A3 is another example except it has lower transconductance and has very high distortion. And with lower output tubes, distortion at one watt is generally higher as there is less "distance" between max power output and 1 watt.
Agreed, I try to have the fewest good stages possible to lower the distortion mixing effect.
-----------------------------------------------------------
Now Sayonara,
"I do not what you heard or not and to what degree your perception of neutrality is same as mine (my "neutrality" comes from "comparable to a wire bypass"). However, conventional Cone/Dome type speaker sare completly unsuited to high fidelity reproduction in domestic settings, full stop. The reasons are manifold, from excessive compression and distortion over unsuitable dispersion pattern and unfavourable room interactions. If you consider such types of speakers as "neutral" than obviously a truely neutral a really neutral speaker (such as a really good main studio monitor) would of course not strike you as being neutral, not that many domestic "Horn" speakers approach neutrality."
Me: (First off, my reference is a "Wire", look at my site.)
My source is the Radiotron Designers Handbook. "For example if a horn with a 40c/s cut-off is to reproduce a 4khz note, there will theoretically be 8% second harmonic distortion for an acoustical power of 0.01 watt per square inch of horn throat. The actual distortion appears to be about half the theoretical value. Thus for reasons of both efficiency and distortion, a horn should only be designed to cover a limited frequency range."
If the bandwidth is half of this example, the distortion still seems quite high.
Thus one need several horns with restricted frequency range to reproduce the sound with little distortion, thus needing crossovers just like regular speakers. There are also problems with dispersion from horns, I think more than with conventional speakers. That is why I always notice a lack of depth with horns. I agree that there are many cone speakers that sound bad, but you failed to mention ribbons and air motion transformers. I heard a pair of ribbons/cone woofers and air motion transformers that simply demolished the horns. The effect of distortion can also give the appearance of dynamics.
You: "So, the parts I use (all of which have to well in a "wire bypass" test make the KT88 sound poor? Sorry, but I like neutrality."
Me: I have done years of research determining the neutrality of parts, and I have seen zillions of schematics and parts listings for DHT amps, and from what you say above; but I can guarantee you they ain't the same parts my friend. You obviously haven't done exhausted testing of parts for sonic neutrality by your own admission. (See next statements below.)
Me: "As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used,"
You: "Yup, unless one uses Valves that are from the start comparably neutral."
Me: So you figure if the tube is neutral sounding you can use any parts? That is ridiculous. If one uses neutral sounding parts, the tube must also be neutral sounding too. If you use non neutral parts and a neutral tube, the sound will also be bad. If one uses neutral parts and non neutral tube, then the sound will also be off.
No, all the parts, including the tube must be neutral sounding if truly excellent sonic performance is to be achieved. Otherwise, every part, including tubes, presents sonic problems and it is impossible to compensate for all of them.
You: "Is it? I find engineering a driverstage for the 300B with > 6db headroom and tons drive using a single valve quite trivial, but that's me and WTFDIK."
You: "A well engineered driverstage does that anyway. And with 50V+ RMS output. It really is trivial."
Me: And are you getting 0.05% total distortion from All the previous stages (no cancelling distortion stages allowed) at max drive for full output of a 300b amp? (I include my preamp and non output stages of my amps.) We both know you're not if you are using this tube near max. By the way, at one watt, my driver/preamp stages produce virtually No distortion, period.
Me: "There would be very little mixing of distortion products from the previous stages to the output stages, thus minimal higher harmonics.
You: Yes, this is the way I enginner my designs."
Yes, we both agree to keep harmonic distortion down. However, there are other forms of distortions that parts create that must be addressed. These affect dynamics, soundstaging, depth, width, imaging, tonal balance etc. Your above comment, and from all the DHT world seems oblivious to this as from all the schematics, chat, and parts lists I have seen. Therefore, these problems don't even seem to be addressed.
Me: "This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube."
You: "I have no such problems up to around 80V RMS output grid drive, above that things get hairy."
Me: And so tubes that require higher drives have big problems with higher harmonics because of the previous stages.
Me: "This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube."
You: "I have no such problems up to around 80V RMS output grid drive, above that things get hairy."
Me: It suddenly occurs at 80 vrms, or 220v P-P? You really want to stick to that statement?
You: "I prefer to start from a very good speaker, as this makes the choise of a matching amplifier much less problematic, as more options are open. Very few speakers can be considered to provide "HIGH FIDELITY" under domestic conditions. And what is the value of the "BEST AMP IN THE WORLD" if the "best music" is then mangeled by these poor excuses for speakers so often found in this day and age. That is of course for the customer, the benefit to the Amplifier manufacturer are obvious...."
Yes, so you would design a speaker and try to fit the amp to the speaker? More options? I thought there was accurate and then sonic problems off to either side of optimum.
I think you make my point. It is very, extremely tough to make a speaker without having an amp to test it with, and speakers are so variable/volatile, and so I would rather start with a easier to design amp and then the speaker around the amp.
By the way, there are many speakers, non horn or single drivers (which has resonance problems), that are not compressed and produce exceptional music. As I mentioned before, the ribbons and air motion transformers I heard clearly out did the horns.
One last thing. Everyone knows I own a company. However, are you affiliated with any companies, groups, or individuals?
There are some problems I need to address. First Dhaen.
"Unfortunately, as a general rule of thumb, the higher mu output valves have higher distortion, so you're back to square 1.
The mixing of distortion products is a very interesting one. It's a good reason to have as few stages as possible. Just recently, I've been in communication with a Swiss enthusiast who's going the single stage route."
I don't find a lot of common ground with your posts, but at least we both have a common view of the problems, even though we tackle them in different ways."
Unfortunately, this general rule is not true. 6550/KT88s have much higher transconductance than EL34s, yet have much lower distortion. 2A3 is another example except it has lower transconductance and has very high distortion. And with lower output tubes, distortion at one watt is generally higher as there is less "distance" between max power output and 1 watt.
Agreed, I try to have the fewest good stages possible to lower the distortion mixing effect.
-----------------------------------------------------------
Now Sayonara,
"I do not what you heard or not and to what degree your perception of neutrality is same as mine (my "neutrality" comes from "comparable to a wire bypass"). However, conventional Cone/Dome type speaker sare completly unsuited to high fidelity reproduction in domestic settings, full stop. The reasons are manifold, from excessive compression and distortion over unsuitable dispersion pattern and unfavourable room interactions. If you consider such types of speakers as "neutral" than obviously a truely neutral a really neutral speaker (such as a really good main studio monitor) would of course not strike you as being neutral, not that many domestic "Horn" speakers approach neutrality."
Me: (First off, my reference is a "Wire", look at my site.)
My source is the Radiotron Designers Handbook. "For example if a horn with a 40c/s cut-off is to reproduce a 4khz note, there will theoretically be 8% second harmonic distortion for an acoustical power of 0.01 watt per square inch of horn throat. The actual distortion appears to be about half the theoretical value. Thus for reasons of both efficiency and distortion, a horn should only be designed to cover a limited frequency range."
If the bandwidth is half of this example, the distortion still seems quite high.
Thus one need several horns with restricted frequency range to reproduce the sound with little distortion, thus needing crossovers just like regular speakers. There are also problems with dispersion from horns, I think more than with conventional speakers. That is why I always notice a lack of depth with horns. I agree that there are many cone speakers that sound bad, but you failed to mention ribbons and air motion transformers. I heard a pair of ribbons/cone woofers and air motion transformers that simply demolished the horns. The effect of distortion can also give the appearance of dynamics.
You: "So, the parts I use (all of which have to well in a "wire bypass" test make the KT88 sound poor? Sorry, but I like neutrality."
Me: I have done years of research determining the neutrality of parts, and I have seen zillions of schematics and parts listings for DHT amps, and from what you say above; but I can guarantee you they ain't the same parts my friend. You obviously haven't done exhausted testing of parts for sonic neutrality by your own admission. (See next statements below.)
Me: "As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used,"
You: "Yup, unless one uses Valves that are from the start comparably neutral."
Me: So you figure if the tube is neutral sounding you can use any parts? That is ridiculous. If one uses neutral sounding parts, the tube must also be neutral sounding too. If you use non neutral parts and a neutral tube, the sound will also be bad. If one uses neutral parts and non neutral tube, then the sound will also be off.
No, all the parts, including the tube must be neutral sounding if truly excellent sonic performance is to be achieved. Otherwise, every part, including tubes, presents sonic problems and it is impossible to compensate for all of them.
You: "Is it? I find engineering a driverstage for the 300B with > 6db headroom and tons drive using a single valve quite trivial, but that's me and WTFDIK."
You: "A well engineered driverstage does that anyway. And with 50V+ RMS output. It really is trivial."
Me: And are you getting 0.05% total distortion from All the previous stages (no cancelling distortion stages allowed) at max drive for full output of a 300b amp? (I include my preamp and non output stages of my amps.) We both know you're not if you are using this tube near max. By the way, at one watt, my driver/preamp stages produce virtually No distortion, period.
Me: "There would be very little mixing of distortion products from the previous stages to the output stages, thus minimal higher harmonics.
You: Yes, this is the way I enginner my designs."
Yes, we both agree to keep harmonic distortion down. However, there are other forms of distortions that parts create that must be addressed. These affect dynamics, soundstaging, depth, width, imaging, tonal balance etc. Your above comment, and from all the DHT world seems oblivious to this as from all the schematics, chat, and parts lists I have seen. Therefore, these problems don't even seem to be addressed.
Me: "This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube."
You: "I have no such problems up to around 80V RMS output grid drive, above that things get hairy."
Me: And so tubes that require higher drives have big problems with higher harmonics because of the previous stages.
Me: "This cannot be said for tubes that require high drive voltages in which higher order harmonics, even 18th, even 27th, are actually produced because of driver and previous stages distortions mixing with the output tube."
You: "I have no such problems up to around 80V RMS output grid drive, above that things get hairy."
Me: It suddenly occurs at 80 vrms, or 220v P-P? You really want to stick to that statement?
You: "I prefer to start from a very good speaker, as this makes the choise of a matching amplifier much less problematic, as more options are open. Very few speakers can be considered to provide "HIGH FIDELITY" under domestic conditions. And what is the value of the "BEST AMP IN THE WORLD" if the "best music" is then mangeled by these poor excuses for speakers so often found in this day and age. That is of course for the customer, the benefit to the Amplifier manufacturer are obvious...."
Yes, so you would design a speaker and try to fit the amp to the speaker? More options? I thought there was accurate and then sonic problems off to either side of optimum.
I think you make my point. It is very, extremely tough to make a speaker without having an amp to test it with, and speakers are so variable/volatile, and so I would rather start with a easier to design amp and then the speaker around the amp.
By the way, there are many speakers, non horn or single drivers (which has resonance problems), that are not compressed and produce exceptional music. As I mentioned before, the ribbons and air motion transformers I heard clearly out did the horns.
One last thing. Everyone knows I own a company. However, are you affiliated with any companies, groups, or individuals?
This thread has generated a lot more traffic than I anticipated, which is good. 😀 I'll have to lump some of the replies together, and be a bit short about it, as I don't have much time today. Hope you'll forgive me for being short. 
Getting the balance of a push-pull right can be hairy at times. The number of components can be discouraging. There is a lot of tuning to do, with signal generators, spectrum analyzers and stuff to get accurate phase measurements. A single ended amplifier has simpler behaviour, and the distortion components universally reflect this.
Note that there is one particular kind of push-pull that I have no quarrel with, and that is the balanced single-end topology, implemented by having each stage of the amp be a differential stage. There are a number of topologies that are considered push-pull, such as the SRPP, that I am less happy about. Particularly any topology operated at less than class A1 or A2 bias (and the latter needs to be done rather carefully).
Efficiency and power are less of an issue than many people would claim. Yes, if you intend to drive some kind of Wilson or a Magnepan, you would do better to get the DP5.52 (2x400W @ 8ohm, 2x1600W @ 2ohm) or somesuch, but many (most?) of the people looking for valves will go for speakers with e.g. 90dB efficiency, which is driven plenty loud for a classical concert with just a few watts. My recommendation of 15-50W should give more than adequate output with a wide range of speakers.
Output resistance, or rather, damping factor (Zload/Zout), has been a big stick that people like to shake. And, in some cases, it does have a significant impact. I'll ignore very low damping factors, since they are gross enough to affect the crossover. Suffice it to say that it is simple enough to get below this.
But the validity of damping factor as an issue depends on the extent to which the woofer is mass controlled vs. current controlled, and so forth. Many people have argued that damping factor is not as critical as some would have us believe. I have heard low-damping amps perform well in the bass discipline.
Distortion is such a simple word, but such a complex topic. First off, do we agree that the extent to which the distortion has an audible effect (a tonal imbalance is audible), is the only criterion of relevance in audio? If not, you might as well skip the remainder of my posts on the topic. If yes, read on.
The distortion characteristic of a valve has several interesting features. First off, it is predominantly low order. For triodes, it is essentially second order only, when operated by itself, in a single stage. (Of course, that's rarely the case, so we see other distortion.) The distortion characteristic of a triode is similar to what occurs in our ear (transition to liquid medium), which leads us to sometimes perceive the sound as closer and more 'intimate' than it really is. That might be considered bad, but if present in very small amounts, it can be useful to improve imaging, rather than distorting it, by making the ear perceive something more like what it would perceive, were it actually at the event, than what it would perceive normally (room reflections et al.). A brief explanation is in order: this distortion performs a process which occurs in the ear upon a signal that has not yet reached the ear, causing the ear to perceive something closer to the signal, rather than the room-distorted version.
Also, the harmonic (and intermodulation) spectra of single ended are an issue. Studies, and knowledge of the physiology of our ear, shows that we have a 6th order masking effect from a fundamental. Thus, at 1% second order, and no other orders, the distortion is inaudible. At 0.5% third order, also with no other orders, the distortion is audible. This from a physiological point of view- I'm not getting into the debate about what all of this actually sounds like. Just pointing out what our ears are capable or incapable of.
It follows that any simple distortion figure, is worthless. We would do better to use a weighted figure, but that too is less than "the whole truth".
Yes, the SV811-3 is very nice. But I thought it had more distortion than the SV572-3 at comparable output levels ?
If not, then by all means, use the SV811-3 instead. That, at least, should still be in production, isn't it?
And, yes, at least 15W in SET.
SY said:
Getting the balance of a push-pull right can be hairy at times. The number of components can be discouraging. There is a lot of tuning to do, with signal generators, spectrum analyzers and stuff to get accurate phase measurements. A single ended amplifier has simpler behaviour, and the distortion components universally reflect this.
Note that there is one particular kind of push-pull that I have no quarrel with, and that is the balanced single-end topology, implemented by having each stage of the amp be a differential stage. There are a number of topologies that are considered push-pull, such as the SRPP, that I am less happy about. Particularly any topology operated at less than class A1 or A2 bias (and the latter needs to be done rather carefully).
Efficiency and power are less of an issue than many people would claim. Yes, if you intend to drive some kind of Wilson or a Magnepan, you would do better to get the DP5.52 (2x400W @ 8ohm, 2x1600W @ 2ohm) or somesuch, but many (most?) of the people looking for valves will go for speakers with e.g. 90dB efficiency, which is driven plenty loud for a classical concert with just a few watts. My recommendation of 15-50W should give more than adequate output with a wide range of speakers.
Output resistance, or rather, damping factor (Zload/Zout), has been a big stick that people like to shake. And, in some cases, it does have a significant impact. I'll ignore very low damping factors, since they are gross enough to affect the crossover. Suffice it to say that it is simple enough to get below this.
But the validity of damping factor as an issue depends on the extent to which the woofer is mass controlled vs. current controlled, and so forth. Many people have argued that damping factor is not as critical as some would have us believe. I have heard low-damping amps perform well in the bass discipline.
Distortion is such a simple word, but such a complex topic. First off, do we agree that the extent to which the distortion has an audible effect (a tonal imbalance is audible), is the only criterion of relevance in audio? If not, you might as well skip the remainder of my posts on the topic. If yes, read on.
The distortion characteristic of a valve has several interesting features. First off, it is predominantly low order. For triodes, it is essentially second order only, when operated by itself, in a single stage. (Of course, that's rarely the case, so we see other distortion.) The distortion characteristic of a triode is similar to what occurs in our ear (transition to liquid medium), which leads us to sometimes perceive the sound as closer and more 'intimate' than it really is. That might be considered bad, but if present in very small amounts, it can be useful to improve imaging, rather than distorting it, by making the ear perceive something more like what it would perceive, were it actually at the event, than what it would perceive normally (room reflections et al.). A brief explanation is in order: this distortion performs a process which occurs in the ear upon a signal that has not yet reached the ear, causing the ear to perceive something closer to the signal, rather than the room-distorted version.
Also, the harmonic (and intermodulation) spectra of single ended are an issue. Studies, and knowledge of the physiology of our ear, shows that we have a 6th order masking effect from a fundamental. Thus, at 1% second order, and no other orders, the distortion is inaudible. At 0.5% third order, also with no other orders, the distortion is audible. This from a physiological point of view- I'm not getting into the debate about what all of this actually sounds like. Just pointing out what our ears are capable or incapable of.
It follows that any simple distortion figure, is worthless. We would do better to use a weighted figure, but that too is less than "the whole truth".
AKSA said:
Yes, the SV811-3 is very nice. But I thought it had more distortion than the SV572-3 at comparable output levels ?
If not, then by all means, use the SV811-3 instead. That, at least, should still be in production, isn't it?
And, yes, at least 15W in SET.
Re: The "sound" of SE
I hope I did not come across as having feelings about what a topology should sound like. I have opinions on what the practical issues with certain topologies are, and what characteristics they exhibit. Physical, measurable characteristics. Such as power, distortion, standing primary current in the trafo, etc..
DHT valves do appear to have the least distortion, yes. However, I am not certain this is not something that has to do with the way in which they are used. Anyway.. The second problem with them, is that they are more susceptible to noise and hum. A solveable problem, though.
Essentially, I've heard people do great stuff with both DHT and IHT (directly and indirectly heated triodes for those who might be reading this while just starting with valves).
I do believe valves have a first order contribution to the sound. However, I do not believe this contribution is the only thing that figures into the equation. There is a result, which is the product of all the factors, including the 'sound' of the valve. I believe the valve has a set of characteristics, which will affect the sound. The characteristics of other components, such as the output transformer, will also affect the sound. The trick is finding a synergy between your components, so the least amount of effort and money can be spent getting the best sound.
I take it the best sound should be the final goal to everyone in this discussion, regardless of camp. Except when economy figures, in which case it has to be 'best sound your money can buy'..
There is much confusion regarding valves. Much of it is dispelled by realizing that people are not always discussing the same things. Put the facts and opinions on a sheet of paper, if it helps, and sort them, and relate them to each other. That way, a semi-clear picture might emerge. (I wouldn't know. One has not emerged for me yet 😉 )..
As far as building the amp is concerned, yes, quality iron is more expensive than most valves, the original Western Electric 300B being one noteable exception.
However, as dhaen said, valves have a finite lifespan. Unless great care is taken to extend the life of your valves, they will typically be limited to something on the order of 5K-20K hours. You can about quadruple this number by running the filaments with a more complex circuit, and by having a delayed turn-on for the high-tension supply, as well as having a sensible rise-time after this delay.
If you go for a SE 300B amp, you will find yourself spending USD 100 or so every time a valve blows, not just once. Also, bear in mind the potential for damage to the valves if you've wired something wrong.
As a sidenote, the Svetlana 6D22S, a valve diode with great stats, will allow you to build a full graetz bridge with valves, and also features an intrinsic delay of 20-30 seconds before it reaches full conduction.
dhaen said:
I hope I did not come across as having feelings about what a topology should sound like. I have opinions on what the practical issues with certain topologies are, and what characteristics they exhibit. Physical, measurable characteristics. Such as power, distortion, standing primary current in the trafo, etc..
DHT valves do appear to have the least distortion, yes. However, I am not certain this is not something that has to do with the way in which they are used. Anyway.. The second problem with them, is that they are more susceptible to noise and hum. A solveable problem, though.
Essentially, I've heard people do great stuff with both DHT and IHT (directly and indirectly heated triodes for those who might be reading this while just starting with valves).
I do believe valves have a first order contribution to the sound. However, I do not believe this contribution is the only thing that figures into the equation. There is a result, which is the product of all the factors, including the 'sound' of the valve. I believe the valve has a set of characteristics, which will affect the sound. The characteristics of other components, such as the output transformer, will also affect the sound. The trick is finding a synergy between your components, so the least amount of effort and money can be spent getting the best sound.
I take it the best sound should be the final goal to everyone in this discussion, regardless of camp. Except when economy figures, in which case it has to be 'best sound your money can buy'..
DVDHack said:
There is much confusion regarding valves. Much of it is dispelled by realizing that people are not always discussing the same things. Put the facts and opinions on a sheet of paper, if it helps, and sort them, and relate them to each other. That way, a semi-clear picture might emerge. (I wouldn't know. One has not emerged for me yet 😉 )..
As far as building the amp is concerned, yes, quality iron is more expensive than most valves, the original Western Electric 300B being one noteable exception.
However, as dhaen said, valves have a finite lifespan. Unless great care is taken to extend the life of your valves, they will typically be limited to something on the order of 5K-20K hours. You can about quadruple this number by running the filaments with a more complex circuit, and by having a delayed turn-on for the high-tension supply, as well as having a sensible rise-time after this delay.
If you go for a SE 300B amp, you will find yourself spending USD 100 or so every time a valve blows, not just once. Also, bear in mind the potential for damage to the valves if you've wired something wrong.
As a sidenote, the Svetlana 6D22S, a valve diode with great stats, will allow you to build a full graetz bridge with valves, and also features an intrinsic delay of 20-30 seconds before it reaches full conduction.
Re: Re: Re: Re: Build with 300B or something else???
I'm not sure how you get to associate "Excellent" sound with "Transistorish" sound.
I did not state that a transistor sounded excellent. I meant to point out that transistors have certain merits over valves, while I think valves, all in all, are preferrable. They are even more preferrable when they are used in a manner which imparts both sets of merits. If you're looking for neutral, accurate sound, while approaching things from an additive angle, as opposed to a subtractive one, then the best money can buy, will be pretty much the same thing, regardless of whether valves or transistors are used.
Yes, in "tradtional" Amp's the EL34 sounds warm and dark (not a "bad" thing per se) and 6550/KT88 rather steely (using the same Amp).
I would disagree with the "not a 'bad' thing per se", but that's just my personal preference. As the guy at Glass Audio put it, I like valves despite their colorations, not because of it. And if you've made the 6550 sound steely, then something's wrong with the topology, or your speakers.
Bear in mind that many people go for a laid-back sound from their valves to correct imperfections elsewhere in their system, such as an overly aggressive midrange, especially in the presence region.
In SE I find all "trioded pentodes" to give less than stellar sound, lacking resolution, involvement and pace, plus usually a less than well defined bass (less easy to follow base lines in the music).
This can easily be ascribed to improper design. The lack of defined bass is almost always due to either (a) transformer problems (transient saturation, etc.), or (b) phase problems. Even the slightest phase error in the bass region will give a feeling that the bass is 'slow', 'out of tune', 'poorly defined' or 'a step behind'.
"Pedestrian" is used here in the same way it usally is. Like "Car vs. Pedestrian". The sound is "slow", "opaque" and in general as if a few veils have been dropped in front of your speakers.
As I commented above, and others commented in the same thread, this depends more on what you use the valves with, than the valves themselves. "Veils" are generally associated with noise. "Slowness" is a phase phenomenon. Again, we're back to the topology issue. SE, even with pentodes, does not need to sound this way.
Oh, I should mention, all speakers used in recent years with me where of 95db/W/m (and not 95db/2.83V/m & 2 Ohm loads) or higher (up to 105db/W/m) in the majority, with only one or two of the common power hungry designerrors commonly sold as "High Fidelity" speaker).
Some power hungry speakers are not design errors, though most are. And power hunger is more commonly related to current controlled magnet systems, or odd crossovers. In those cases, a valve system can have trouble driving them, unless designed to cope with these requirements.
As said, I have little experience with the 6S33 and like with the Svetlana 572/811 "Dash" Series you don't get a choice of brands and are limited to a potentially early running out supply, also suitable Output transformers are not exactly common.
With the SV572/811 series you can still rather easily get hold of them, and the reason valves dry up, is supply and demand. People go for the 300B cult, rather than looking for the right valve for the job. This affects many valves, not just the Svetlana ones. But, yes, there is a potential concern here, though I would think you could order spares from Cary, whom I believe still has them manufactured by Svetlana.
A suitable output transformer could be e.g. the LL1663 from Lundahl, or any other 5K:8 transformer which can take the voltages and currents involved, with low distortion at 25W/30Hz.
As for the 6C33C-B, the LL1627 can be configured for 650:8, and copes with 25W/30Hz. I doubt you would have any trouble finding other companies delivering these impedances. And as for stock running out, the 6C33C-B is still in production, and will remain so for a good long time.
I'm not sure how you get to associate "Excellent" sound with "Transistorish" sound.
I did not state that a transistor sounded excellent. I meant to point out that transistors have certain merits over valves, while I think valves, all in all, are preferrable. They are even more preferrable when they are used in a manner which imparts both sets of merits. If you're looking for neutral, accurate sound, while approaching things from an additive angle, as opposed to a subtractive one, then the best money can buy, will be pretty much the same thing, regardless of whether valves or transistors are used.
Yes, in "tradtional" Amp's the EL34 sounds warm and dark (not a "bad" thing per se) and 6550/KT88 rather steely (using the same Amp).
I would disagree with the "not a 'bad' thing per se", but that's just my personal preference. As the guy at Glass Audio put it, I like valves despite their colorations, not because of it. And if you've made the 6550 sound steely, then something's wrong with the topology, or your speakers.
Bear in mind that many people go for a laid-back sound from their valves to correct imperfections elsewhere in their system, such as an overly aggressive midrange, especially in the presence region.
In SE I find all "trioded pentodes" to give less than stellar sound, lacking resolution, involvement and pace, plus usually a less than well defined bass (less easy to follow base lines in the music).
This can easily be ascribed to improper design. The lack of defined bass is almost always due to either (a) transformer problems (transient saturation, etc.), or (b) phase problems. Even the slightest phase error in the bass region will give a feeling that the bass is 'slow', 'out of tune', 'poorly defined' or 'a step behind'.
"Pedestrian" is used here in the same way it usally is. Like "Car vs. Pedestrian". The sound is "slow", "opaque" and in general as if a few veils have been dropped in front of your speakers.
As I commented above, and others commented in the same thread, this depends more on what you use the valves with, than the valves themselves. "Veils" are generally associated with noise. "Slowness" is a phase phenomenon. Again, we're back to the topology issue. SE, even with pentodes, does not need to sound this way.
Oh, I should mention, all speakers used in recent years with me where of 95db/W/m (and not 95db/2.83V/m & 2 Ohm loads) or higher (up to 105db/W/m) in the majority, with only one or two of the common power hungry designerrors commonly sold as "High Fidelity" speaker).
Some power hungry speakers are not design errors, though most are. And power hunger is more commonly related to current controlled magnet systems, or odd crossovers. In those cases, a valve system can have trouble driving them, unless designed to cope with these requirements.
As said, I have little experience with the 6S33 and like with the Svetlana 572/811 "Dash" Series you don't get a choice of brands and are limited to a potentially early running out supply, also suitable Output transformers are not exactly common.
With the SV572/811 series you can still rather easily get hold of them, and the reason valves dry up, is supply and demand. People go for the 300B cult, rather than looking for the right valve for the job. This affects many valves, not just the Svetlana ones. But, yes, there is a potential concern here, though I would think you could order spares from Cary, whom I believe still has them manufactured by Svetlana.
A suitable output transformer could be e.g. the LL1663 from Lundahl, or any other 5K:8 transformer which can take the voltages and currents involved, with low distortion at 25W/30Hz.
As for the 6C33C-B, the LL1627 can be configured for 650:8, and copes with 25W/30Hz. I doubt you would have any trouble finding other companies delivering these impedances. And as for stock running out, the 6C33C-B is still in production, and will remain so for a good long time.
Re: Cleared some things?
Konnichiwa,
Well, the understanding of "Horns" and their drivers has SOMEWHAT improved in the the inervening 70 or so odd years since this was written, primarily due to demand from the Studio/Sound Reinforcemnt sector (the serious stuff) for high, undistorted SPL's. We may argue 9or not) the implications technically, but that is for loudspeakers.
Nor are horns per se required to produce high SPL's from limited power (actually ALL HiFi type Cone/Dome drivers and many Studio/Pro Drivers I encountered show material compression [>0.5db] at 1 Watt power applied). My point is simple. A conventional moving coil speaker is subject to a number of limitations that make a high efficiency mandatory if high distortion and compression levels are to be avoided, that's just physics.
And a speaker to be used in an acoustic "life" envoironment (meaning outside anechonic chambers and studio "non envoironment" rooms) requires a well controlled dispersion and even off axis response if the room is not going to introduce colorations that automatically forfit any chance at "High FidelitY". I don't make the physical laws underpinning real electromagnetic systems and acoustics....
So, you are talking about what you think are "Horn" speakers, I merely talk about competently designed speakers.
I have no problems with these per se, except they are tweeters, not speakers.
I am not getting into a "I tested more parts than you" ****ing contest. Unless we put numbers on the table there is little point and quantety sill does not ensure quality. Do you actually use Electrolytic Capacitors, for example, ANYWHERE? Then good by neutrality from the outset.
What is ridiculous is your deliberate misstating what I have written. If you use Valves with a minimal inherent coloration level and passive parts that match this behaviour you end up with a fairly neutral sounding amplifier.
I have made driver stages that did provide this kind performance. In the real world I found that I find no further subjective improvement if the driver distortion is lowered to much more than around 12 - 20db below the Output Valves distortion AT FULL POWER fail to make further improvements.
And with all due respect, 0.05% Distortion from a 6922 providing around 20V RMS (triode connected EL34)? Do you mind if I qestion this sort of number?
And we both know that you do not get 15 - 20V RMS from a 6922.
Yet you Driver valves that actually are quite non-linear and have quite high levels of upper harmonics.
Again, I agree, yet I would question your use of Parts that minimise such distortion, based on what you have published about your designs on your webpages.
Now the schematics and parts lists of quite a few of my designs are available freely. And you have ACTUALLY looked at these? And noticed my use of parts that "introduce a lot of distortion"? I find this quite puzzeling. Maybe you would like to elaborate exactly which parts those are and what you feel is a "neutral" sounding part by comparison?
Of course. But I do not like to work with these valves to start. Though with intelligent design one can even work around such problems, but that's hard work.
No, it is a Value that is based on the Valves I prefer to use and their limits and thus the hence dictated limits in voltage swing. One can use Valves with higher permissable Anode voltages and Anode dissipations but the supply of suitable devices gets increasingly thin as you up the required voltage swing from the anode.
How do you arrive at that conclusion? A speaker that offers resonably resolution and neutrality in normal domestic settings with near realistic SPL's and provides very much an "Open Window" perspective does not need an amplifier to fit it. Assuming the Amplifier is competently designed and performas well the result will be good, even using (dread of dreads) solid state amplification. Naturaly there are differences in sound and presentation between different amplifiers and those are observable, easily so.
Hmmm. Has that got something to do with the fact that you make amplifiers? How do you know you are not compensating in the speaker design for deffects in amplifier performance? How can you judge the sonic performance of an amplifier without having first a speaker that is sufficiently resolving and neutral?
I view things from a system perspective. The biggest variables in ANY system are actually room interactions and the various distortion mechanisms in speakers (assuming a reasonably flat response). There is a LOT more difference here than there is in amplifiers.
Now, how about we agree that if a speaker reproduces input signals comparably well, in room (recognisable, if distorted squarewaves, even and short RT60, low levels of compression and distortion and of course even on axis and diffuse field response) that it will offer a resonably enough "open window" for performance evaluation, especially if augumented with headphone listening, using suitably high performance models?
Yes. And many of them are active, large format studio monitors and a few exceptions from the High End/HiFi side.
Which Ribbons/AMT's and which Horns?
Only in the loosest possible way in that I have either allowed companies to use my designs on condition of providing payments to childrens charities as "licence fee", plus in one case being not charged development costs on Transformers etc. the company later sells to the general market, which given the many iterations I tend to ask for is arguably a major benefit.
Oh yes, I also write the occasional review and tinker with all sorts of things, the results are usually published for general consumption.
And yes, I have been "holding back" some of what I consider "my best designs" in certain critical details or entierly with a view to eventually manufacturing and markting them.
Sayonara
Konnichiwa,
Positron said:
Well, the understanding of "Horns" and their drivers has SOMEWHAT improved in the the inervening 70 or so odd years since this was written, primarily due to demand from the Studio/Sound Reinforcemnt sector (the serious stuff) for high, undistorted SPL's. We may argue 9or not) the implications technically, but that is for loudspeakers.
Nor are horns per se required to produce high SPL's from limited power (actually ALL HiFi type Cone/Dome drivers and many Studio/Pro Drivers I encountered show material compression [>0.5db] at 1 Watt power applied). My point is simple. A conventional moving coil speaker is subject to a number of limitations that make a high efficiency mandatory if high distortion and compression levels are to be avoided, that's just physics.
And a speaker to be used in an acoustic "life" envoironment (meaning outside anechonic chambers and studio "non envoironment" rooms) requires a well controlled dispersion and even off axis response if the room is not going to introduce colorations that automatically forfit any chance at "High FidelitY". I don't make the physical laws underpinning real electromagnetic systems and acoustics....
So, you are talking about what you think are "Horn" speakers, I merely talk about competently designed speakers.
Positron said:
I have no problems with these per se, except they are tweeters, not speakers.
Positron said:
I am not getting into a "I tested more parts than you" ****ing contest. Unless we put numbers on the table there is little point and quantety sill does not ensure quality. Do you actually use Electrolytic Capacitors, for example, ANYWHERE? Then good by neutrality from the outset.
Positron said:
What is ridiculous is your deliberate misstating what I have written. If you use Valves with a minimal inherent coloration level and passive parts that match this behaviour you end up with a fairly neutral sounding amplifier.
Positron said:
I have made driver stages that did provide this kind performance. In the real world I found that I find no further subjective improvement if the driver distortion is lowered to much more than around 12 - 20db below the Output Valves distortion AT FULL POWER fail to make further improvements.
And with all due respect, 0.05% Distortion from a 6922 providing around 20V RMS (triode connected EL34)? Do you mind if I qestion this sort of number?
Positron said:
And we both know that you do not get 15 - 20V RMS from a 6922.
Positron said:
Yet you Driver valves that actually are quite non-linear and have quite high levels of upper harmonics.
Positron said:
Again, I agree, yet I would question your use of Parts that minimise such distortion, based on what you have published about your designs on your webpages.
Now the schematics and parts lists of quite a few of my designs are available freely. And you have ACTUALLY looked at these? And noticed my use of parts that "introduce a lot of distortion"? I find this quite puzzeling. Maybe you would like to elaborate exactly which parts those are and what you feel is a "neutral" sounding part by comparison?
Positron said:
Of course. But I do not like to work with these valves to start. Though with intelligent design one can even work around such problems, but that's hard work.
Positron said:
No, it is a Value that is based on the Valves I prefer to use and their limits and thus the hence dictated limits in voltage swing. One can use Valves with higher permissable Anode voltages and Anode dissipations but the supply of suitable devices gets increasingly thin as you up the required voltage swing from the anode.
Positron said:
How do you arrive at that conclusion? A speaker that offers resonably resolution and neutrality in normal domestic settings with near realistic SPL's and provides very much an "Open Window" perspective does not need an amplifier to fit it. Assuming the Amplifier is competently designed and performas well the result will be good, even using (dread of dreads) solid state amplification. Naturaly there are differences in sound and presentation between different amplifiers and those are observable, easily so.
Positron said:
Hmmm. Has that got something to do with the fact that you make amplifiers? How do you know you are not compensating in the speaker design for deffects in amplifier performance? How can you judge the sonic performance of an amplifier without having first a speaker that is sufficiently resolving and neutral?
I view things from a system perspective. The biggest variables in ANY system are actually room interactions and the various distortion mechanisms in speakers (assuming a reasonably flat response). There is a LOT more difference here than there is in amplifiers.
Now, how about we agree that if a speaker reproduces input signals comparably well, in room (recognisable, if distorted squarewaves, even and short RT60, low levels of compression and distortion and of course even on axis and diffuse field response) that it will offer a resonably enough "open window" for performance evaluation, especially if augumented with headphone listening, using suitably high performance models?
Positron said:
Yes. And many of them are active, large format studio monitors and a few exceptions from the High End/HiFi side.
Positron said:
Which Ribbons/AMT's and which Horns?
Positron said:
Only in the loosest possible way in that I have either allowed companies to use my designs on condition of providing payments to childrens charities as "licence fee", plus in one case being not charged development costs on Transformers etc. the company later sells to the general market, which given the many iterations I tend to ask for is arguably a major benefit.
Oh yes, I also write the occasional review and tinker with all sorts of things, the results are usually published for general consumption.
And yes, I have been "holding back" some of what I consider "my best designs" in certain critical details or entierly with a view to eventually manufacturing and markting them.
Sayonara
Hi,
I totally agree with this but the AMTs are ribbons too, aren't they?
The early ones didn't image all that well, hopefully things have improved over the past 25 years....
Ribbons + OTLs is like living dangerously but boy, do they sound good or what?
Cheers,😉
I totally agree with this but the AMTs are ribbons too, aren't they?
The early ones didn't image all that well, hopefully things have improved over the past 25 years....
Ribbons + OTLs is like living dangerously but boy, do they sound good or what?
Cheers,😉
Re: Re: Re: 300B SETs
Originally posted by James D.
All feedback referred to above is global feedback derived from the amplifier output. Non global feedback or global feedback not taken from the output doesn't necessarily show this variability.
We all know global feedback tends to have predictably unpredictable results. I like the approach used in one of the Lundahl application notes, where the secondary of the transformer is coupled in series with the cathode. The low static resistance (0.4ohm) of the secondary means that only a small DC offset appears over the speakers (about 150mV, IIRC), which they can easily live with (many SS amps have more). At the same time, it gives a small amount of local NFB with very limited phase shift.
So I mainly design and make PP fully balanced designs without feedback.
Consider trying a fully balanced single-end amplifier, consisting only of differential stages, and employing local feedback at each stage. For example, with the 6N1P double-triode, the mu is 33. Use feedback to get it down to, say, 20. That's less than 6dB feedback, and does wonders for stereo imaging.
I recently tried 211s in a cathode follower push pull ..... Finding a driver that would swing 600Vpp with similar performance is the problem.
Right. That explains a few things. I had not realized (as I had not looked closely enough at that valve) that the 211 requires 600Vpp. That is indeed a tall order, and an interstage transformer does deliver. Try driving it with, say, the 6C45P, and an interstage transformer. As the current capacity is high, and plate resistance low, you could step up very far, possibly getting by with only three valves (the 6C45P and two 211s).
Originally posted by Mike C
Really funny you should say that, as (here I may lose any street cred I may have built up) I've listened to Dido's latest CD and I like it.
I don't mind if people have different tastes than myself. I would mind if they didn't.
But where she can't quite manage the high notes it comes out as something resembling white noise!
That may well be (in fact, probably must be considered) a mastering problem. Which is getting more common with pop nowadays. People who prefer pop have an even tighter rope to walk than the rest of us, as far as the balance between fidelity to the source and insensitivity to flaws goes.
Good point - I love the last comment!
That was in reference to spending the money on misc, 'or (gasp) records)', for those who just started reading.
I find it simply amazing that so many people, with wide taste in music, spend less than a tenth of their HiFi investments on the source material itself. If you are going to buy a half-a-million USD stereo, why on earth would you spend less than 100.000 USD on records? Sure, not all of them will be your favourites, but I would be astounded if I were to find I am the only one who likes a sufficient amount of music that I could easily spend that much on records if I had it available.
It's about music, people.. not technology. Any stereo that doesn't get that point across is worth nothing, if you're not a mastering engineer.
Originally posted by James D.
All feedback referred to above is global feedback derived from the amplifier output. Non global feedback or global feedback not taken from the output doesn't necessarily show this variability.
We all know global feedback tends to have predictably unpredictable results. I like the approach used in one of the Lundahl application notes, where the secondary of the transformer is coupled in series with the cathode. The low static resistance (0.4ohm) of the secondary means that only a small DC offset appears over the speakers (about 150mV, IIRC), which they can easily live with (many SS amps have more). At the same time, it gives a small amount of local NFB with very limited phase shift.
So I mainly design and make PP fully balanced designs without feedback.
Consider trying a fully balanced single-end amplifier, consisting only of differential stages, and employing local feedback at each stage. For example, with the 6N1P double-triode, the mu is 33. Use feedback to get it down to, say, 20. That's less than 6dB feedback, and does wonders for stereo imaging.
I recently tried 211s in a cathode follower push pull ..... Finding a driver that would swing 600Vpp with similar performance is the problem.
Right. That explains a few things. I had not realized (as I had not looked closely enough at that valve) that the 211 requires 600Vpp. That is indeed a tall order, and an interstage transformer does deliver. Try driving it with, say, the 6C45P, and an interstage transformer. As the current capacity is high, and plate resistance low, you could step up very far, possibly getting by with only three valves (the 6C45P and two 211s).
Originally posted by Mike C
Really funny you should say that, as (here I may lose any street cred I may have built up) I've listened to Dido's latest CD and I like it.
I don't mind if people have different tastes than myself. I would mind if they didn't.
But where she can't quite manage the high notes it comes out as something resembling white noise!
That may well be (in fact, probably must be considered) a mastering problem. Which is getting more common with pop nowadays. People who prefer pop have an even tighter rope to walk than the rest of us, as far as the balance between fidelity to the source and insensitivity to flaws goes.
Good point - I love the last comment!
That was in reference to spending the money on misc, 'or (gasp) records)', for those who just started reading.
I find it simply amazing that so many people, with wide taste in music, spend less than a tenth of their HiFi investments on the source material itself. If you are going to buy a half-a-million USD stereo, why on earth would you spend less than 100.000 USD on records? Sure, not all of them will be your favourites, but I would be astounded if I were to find I am the only one who likes a sufficient amount of music that I could easily spend that much on records if I had it available.
It's about music, people.. not technology. Any stereo that doesn't get that point across is worth nothing, if you're not a mastering engineer.
I take some issue with that, other than the qualifier "at times." People have been making well-balanced p-p amps for many decades. Using very conventional phase splitters and output stages, balancing can be easily achieved with some pretty simple tools. In fact, to balance a simple phase splitter like a long tail takes only a voltmeter. Balancing a split-load is even easier- just use precision resistors. Balancing an output stage can be done with a 60Hz source (easy to generate internally) and a voltmeter. With a sensitive ear, you don't even need the voltmeter. To get perfect balance is unnecessary- you can routinely do it well enough to get even order distortion far below what any of the SE amps I'm familiar with have.
"At times" some people have designed bad p-p circuits which are tricky and difficult to balance, that is true. But just because there are a few bad designs out there using the concept doesn't mean the concept is bad. It is not that difficult to design a p-p amp whose output is indistinguishable by ear from the input (at least above the lowest octave), no matter what games one cares to play with audibility thresholds of various distortion components. The issues, IMO, lie more in questions of overload behavior at that point- and I submit that an 8 watt half-an-amp is going to have to overload much more gracefully than a 100 wat amp if it's to be used for music reproduction.
Unless you're using a single driver speaker (and to a lesser extent, even then) the source impedance issue will bite you in the midrage and treble, not just in the bass. Impedance curves aren't flat, and conjugate networks (limited in application, alas) can burn up a lot of that power of which the SE amp isn't blessed with an abundance. You may be able to cite SE designs with source Zs low enough to make this a non-issue, and I'd be interested in hearing about them.
Re: Drive voltages
Indeed a DIYer does not need to concern him/her self with commercial viability. However, some of us still need to concern ourselves with money, and a commercial designer *does* have options that realize pretty much the same thing. They just have to put more work into it first. For example, supplying different voltages can be realized by clever use of a switcher operating off one of the supplies to step voltages up or down. You shouldn't even need to do a lot of work for it to give decent results. Fine tune it properly, and voilla.
The AMTs are ribbons with an intrinsic linear horn loading from their folding, which gives rise to bellows-like motion.
I would have a look at the Piega's, but it doesn't seem as though that company sells their tweeters (and soon midranges) standalone. 120dB sensitivity, decent impedance, can take a beating, and does 50kHz with good off-axis response. Funny what a few bubbles can do...
As far as ribbons + OTLs go, have you considered using an OTL to drive the Bohlender-Graebner 28" ribbons? They can be crossed over at 150Hz with a 4th order L-R.. Add an 8" seas excel magnesium woofer, powered by some hot sand, and you should be all set... =)
dhaen said:Of course in the commercial world, no one would dream of supplying the driver stage with a higher HT voltage than the output stage; it'd be too expensive. That's where DIY really wins: We can use the best output valves, regardless of mu, and make low distortion drivers.
Of course you could just put ndB of feedback to "fix" all the ills
Indeed a DIYer does not need to concern him/her self with commercial viability. However, some of us still need to concern ourselves with money, and a commercial designer *does* have options that realize pretty much the same thing. They just have to put more work into it first. For example, supplying different voltages can be realized by clever use of a switcher operating off one of the supplies to step voltages up or down. You shouldn't even need to do a lot of work for it to give decent results. Fine tune it properly, and voilla.
fdegrove said:
The AMTs are ribbons with an intrinsic linear horn loading from their folding, which gives rise to bellows-like motion.
I would have a look at the Piega's, but it doesn't seem as though that company sells their tweeters (and soon midranges) standalone. 120dB sensitivity, decent impedance, can take a beating, and does 50kHz with good off-axis response. Funny what a few bubbles can do...
As far as ribbons + OTLs go, have you considered using an OTL to drive the Bohlender-Graebner 28" ribbons? They can be crossed over at 150Hz with a 4th order L-R.. Add an 8" seas excel magnesium woofer, powered by some hot sand, and you should be all set... =)
I take some issue with that, other than the qualifier "at times." People have been making well-balanced p-p amps for many decades. Using very conventional phase splitters and output stages, balancing can be easily achieved with some pretty simple tools.
I would think an input transformer would be a better choice than a phase splitter? In many cases, that can even eliminate a gain stage.
In fact, to balance a simple phase splitter like a long tail takes only a voltmeter. Balancing a split-load is even easier- just use precision resistors. Balancing an output stage can be done with a 60Hz source (easy to generate internally) and a voltmeter. With a sensitive ear, you don't even need the voltmeter. To get perfect balance is unnecessary- you can routinely do it well enough to get even order distortion far below what any of the SE amps I'm familiar with have.
My main point was that you have to redo this as the valve ages.
But just because there are a few bad designs out there using the concept doesn't mean the concept is bad.
True. I might have come across a bit strongly in some of my posts. Especially when operating in class A1, one can certainly realize designs that exhibit a characteristic that is very similar to SE. Apologies.
It is not that difficult to design a p-p amp whose output is indistinguishable by ear from the input (at least above the lowest octave), no matter what games one cares to play with audibility thresholds of various distortion components.
I've yet to hear any amplifier whose output was indistinguishable. If you have one, please point out its name, and I'll see if I can find a dealer in Norway which will let me put a divider network on its output to confirm this.
Of course, all bets are off as soon as a load enters the picture.
As far as "games" with audibility thresholds is concerned, I was pointing out the physical limits of the ear. The actual limits of perception might be different. My main point was that a SE amplifier distorts in a manner which is (if coupled properly) complementary to loudspeaker distortion, and identical to in-ear distortion, thus sometimes reducing part of the distortion spectrum, and in almost all cases leaving only distortions that the mind has had a lifetime of experience in disregarding.
The issues, IMO, lie more in questions of overload behavior at that point- and I submit that an 8 watt half-an-amp is going to have to overload much more gracefully than a 100 wat amp if it's to be used for music reproduction.
Yes. A point has been made about this in some online article that I cannot recall the title of. Distortion that is nonmonotonic in character, or distortion spectra that, when viewed against output power, have abrupt nonlinearities, tends to get offensive. Naturally a 100W amplifier has a lot more headroom before reaching such abrupt nonlinearities. OTOH, PP amps tend to be more abrupt about it, so it depends on what your amp will be looking at. This problem about discontinous distortion curves is one of the main problems with large amounts of feedback. That just sharpens the bend at the end.
Unless you're using a single driver speaker (and to a lesser extent, even then) the source impedance issue will bite you in the midrage and treble, not just in the bass. Impedance curves aren't flat, and conjugate networks (limited in application, alas) can burn up a lot of that power of which the SE amp isn't blessed with an abundance.
Mostly, impedance curves have peaks, not throughs, outside the bass. Again, it all depends on what the control mechanism for the driver is. High Qms + low Qes means you need low output Z. Low Qms + high Qes lets you get off with higher output Zs.
You may be able to cite SE designs with source Zs low enough to make this a non-issue, and I'd be interested in hearing about them.
Unless I misread my own post, the comment I made that partained specifically to SE designs, was that it is easy to get below the point where it does not interact significantly with a reasonably behaved crossover. Staying below 1 ohm will make this mostly a nonissue, and staying below a quart ohm is more than sufficient for any speaker that will thrive off the limited power of an SE design. To use a readily available reference, the Lundahl application note shows an amp which has a 20Hz-20KHz output resistance of 0.9ohm from an EL34. Using something like the 6C33C-B, with its plate resistance of 80 ohm or so, and a bit of local NFB, you should be able to get down even further.
I've got to cut this post short. My GF, who owns my speakers (don't ask), is threatening to take them away from me if I don't join her soon. (It's 4:27am here..)
I would think an input transformer would be a better choice than a phase splitter? In many cases, that can even eliminate a gain stage.
In fact, to balance a simple phase splitter like a long tail takes only a voltmeter. Balancing a split-load is even easier- just use precision resistors. Balancing an output stage can be done with a 60Hz source (easy to generate internally) and a voltmeter. With a sensitive ear, you don't even need the voltmeter. To get perfect balance is unnecessary- you can routinely do it well enough to get even order distortion far below what any of the SE amps I'm familiar with have.
My main point was that you have to redo this as the valve ages.
But just because there are a few bad designs out there using the concept doesn't mean the concept is bad.
True. I might have come across a bit strongly in some of my posts. Especially when operating in class A1, one can certainly realize designs that exhibit a characteristic that is very similar to SE. Apologies.
It is not that difficult to design a p-p amp whose output is indistinguishable by ear from the input (at least above the lowest octave), no matter what games one cares to play with audibility thresholds of various distortion components.
I've yet to hear any amplifier whose output was indistinguishable. If you have one, please point out its name, and I'll see if I can find a dealer in Norway which will let me put a divider network on its output to confirm this.
Of course, all bets are off as soon as a load enters the picture.
As far as "games" with audibility thresholds is concerned, I was pointing out the physical limits of the ear. The actual limits of perception might be different. My main point was that a SE amplifier distorts in a manner which is (if coupled properly) complementary to loudspeaker distortion, and identical to in-ear distortion, thus sometimes reducing part of the distortion spectrum, and in almost all cases leaving only distortions that the mind has had a lifetime of experience in disregarding.
The issues, IMO, lie more in questions of overload behavior at that point- and I submit that an 8 watt half-an-amp is going to have to overload much more gracefully than a 100 wat amp if it's to be used for music reproduction.
Yes. A point has been made about this in some online article that I cannot recall the title of. Distortion that is nonmonotonic in character, or distortion spectra that, when viewed against output power, have abrupt nonlinearities, tends to get offensive. Naturally a 100W amplifier has a lot more headroom before reaching such abrupt nonlinearities. OTOH, PP amps tend to be more abrupt about it, so it depends on what your amp will be looking at. This problem about discontinous distortion curves is one of the main problems with large amounts of feedback. That just sharpens the bend at the end.
Unless you're using a single driver speaker (and to a lesser extent, even then) the source impedance issue will bite you in the midrage and treble, not just in the bass. Impedance curves aren't flat, and conjugate networks (limited in application, alas) can burn up a lot of that power of which the SE amp isn't blessed with an abundance.
Mostly, impedance curves have peaks, not throughs, outside the bass. Again, it all depends on what the control mechanism for the driver is. High Qms + low Qes means you need low output Z. Low Qms + high Qes lets you get off with higher output Zs.
You may be able to cite SE designs with source Zs low enough to make this a non-issue, and I'd be interested in hearing about them.
Unless I misread my own post, the comment I made that partained specifically to SE designs, was that it is easy to get below the point where it does not interact significantly with a reasonably behaved crossover. Staying below 1 ohm will make this mostly a nonissue, and staying below a quart ohm is more than sufficient for any speaker that will thrive off the limited power of an SE design. To use a readily available reference, the Lundahl application note shows an amp which has a 20Hz-20KHz output resistance of 0.9ohm from an EL34. Using something like the 6C33C-B, with its plate resistance of 80 ohm or so, and a bit of local NFB, you should be able to get down even further.
I've got to cut this post short. My GF, who owns my speakers (don't ask), is threatening to take them away from me if I don't join her soon. (It's 4:27am here..)
Some problems in your answers.
Ok, let's take a look at your replies and see what we come up with. About half way down, we are going to make some important discoveries about your real knowledge vs what you try to lead us to believe.
"Well, the understanding of "Horns" and their drivers has SOMEWHAT improved in the the inervening 70 or so odd years since this was written, primarily due to demand from the Studio/Sound Reinforcemnt sector (the serious stuff) for high, undistorted SPL's. We may argue 9or not) the implications technically, but that is for loudspeakers....
My point is simple. A conventional moving coil speaker is subject to a number of limitations that make a high efficiency mandatory if high distortion and compression levels are to be avoided, that's just physics....
So, you are talking about what you think are "Horn" speakers, I merely talk about competently designed speakers."
Me: Unfortunately you don't seem to know that many horns use rear loading, which offers the "box" effect just like conventional speakers, with colorations too.
Secondly, according to the Radiotron Designers Handbook, "It is impossible to design a horn loudspeaker that covers a wide frequency band and is simultaneously free from non-linear distortion. Thus two separate units for low and high frequencies are essential for fidelity."
Next, distortion figures they gave have not been improved that much, so distortions are still coming from the throat, fm, loading enclosure etc.
Me: "but you failed to mention ribbons and air motion transformers.
You: "I have no problems with these per se, except they are tweeters, not speakers."
Me: Really, they go down to 100hz and they are tweeters? That is the best answer you can give?
You: "I am not getting into a "I tested more parts than you" ****ing contest. Unless we put numbers on the table there is little point and quantety sill does not ensure quality. Do you actually use Electrolytic Capacitors, for example, ANYWHERE? Then good by neutrality from the outset."
Me: Is that suppose to be a blockbuster comment? Did you have to copy that from my site? If you had really read my pages, you would have known the answer.
"Me: "As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used,"
You: "Yup, unless one uses Valves that are from the start comparably neutral."
Me: So you figure if the tube is neutral sounding you can use any parts? That is ridiculous.
You: "What is ridiculous is your deliberate misstating what I have written. If you use Valves with a minimal inherent coloration level and passive parts that match this behaviour you end up with a fairly neutral sounding amplifier."
Me: Look at your reply again. I said neutral parts are necessary and you said "unless one usesneutral valves".
You stated it. The word "Unless" says alot. If you didn't mean it, great, cause it isn't true.
Me: "And are you getting 0.05% total distortion from All the previous stages (no cancelling distortion stages allowed) at max drive for full output of a 300b amp? (I include my preamp and non output stages of my amps.)"
You: "I have made driver stages that did provide this kind performance. In the real world I found that I find no further subjective improvement if the driver distortion is lowered to much more than around 12 - 20db below the Output Valves distortion AT FULL POWER fail to make further improvements.
And with all due respect, 0.05% Distortion from a 6922 providing around 20V RMS (triode connected EL34)? Do you mind if I qestion this sort of number?"
Me: If you found that to be the case, great, but the higher the harmonic, the lower the amplitude needs to be to be heard, as just mentioned above in another post. Remember, there is also 6th, 9th, 12th etc. and many others to consider that are lower order.
.05%, you seem to question while mixing two different circuits. The preamp cannot generate 20 volts rms with .05% distortion, but 2 volts at .02% or less. The SET amp doesn't either. But the PP amp using the 6922 does. OF course I didn't mention which amp, my aplogies. But the 6922 in the SET amp does provide the drive with only around .3% distortion. Actual measurement of course.
Me: "We both know you're not (getting .05% distortion) if you are using this tube near max. (refering to low distortion in the driver gain stage running near maximum output signal ratings.)
You: "And we both know that you do not get 15 - 20V RMS from a 6922."
Me: Calling me a liar won't help you.
A second indication of how little you know about designing. I can easily get 20v rms out of a 6922 with fairly small amounts of distortion, about the same percentage as you compared to the output tubes.
Me: "Yes, we both agree to keep harmonic distortion down.
You: "Yet you Driver valves that actually are quite non-linear and have quite high levels of upper harmonics."
Again not knowing what you are talking about. Some newer 6922s have virtually no higher order distortions. Don't make a fool out of yourself with halfwit comments.
And the preamp and driver in the PP amp, 6922s, only measures .05% at full drive for full output power, measured. Not measureable at signals for one watt out. Yes, gobs of higher harmonics, LOL.
You: "Again, I agree, yet I would question your use of Parts that minimise such distortion, based on what you have published about your designs on your webpages."
Me: Interesting trick trying to defame me. What parts do I even list that you question? List them for all of us. And do you think I am going to give you my designs? LOL. You are not that clever, nor superior.
You: "How do you arrive at that conclusion? A speaker that offers resonably resolution and neutrality in normal domestic settings with near realistic SPL's and provides very much an "Open Window" perspective does not need an amplifier to fit it. Assuming the Amplifier is competently designed and performas well the result will be good, even using (dread of dreads) solid state amplification. Naturaly there are differences in sound and presentation between different amplifiers and those are observable, easily so."
Me: Question. How do you know you are even in the ball park without listening first, which requires an amp? A speaker has much wider variation in sound quality than an amp. Go to any store and listen to them. Horrible ones. Anyone can throw drivers into a cabinet and get absolutely horrible sound, much worse than an amp. And single drivers not in a cabinet are not flat at all.
Can really mess up an amp design there.
You: "Hmmm. Has that got something to do with the fact that you make amplifiers? How do you know you are not compensating in the speaker design for deffects in amplifier performance? How can you judge the sonic performance of an amplifier without having first a speaker that is sufficiently resolving and neutral?
I view things from a system perspective. The biggest variables in ANY system are actually room interactions and the various distortion mechanisms in speakers (assuming a reasonably flat response). There is a LOT more difference here than there is in amplifiers...."
Me: So, simply take the speakers outdoors. That was an easy answer. Now we simplify things don't we. Work on the room later.
Second, I work back and forth between tweeking amps and tweeking speakers. However, tweeking an amp is such a minor things compared to radical surgery for speakers that are way off in accuracy. If a speaker is designed wrong it is hard to get it right.
Me: Didn't write down any systems, didn't think necessary. Seems after dozens of setups, I never seem to hear the right ones.
You: "Only in the loosest possible way in that I have either allowed companies to use my designs on condition of providing payments to childrens charities as "licence fee", plus in one case being not charged development costs on Transformers etc. the company later sells to the general market, which given the many iterations I tend to ask for is arguably a major benefit.
Oh yes, I also write the occasional review and tinker with all sorts of things, the results are usually published for general consumption.
And yes, I have been "holding back" some of what I consider "my best designs" in certain critical details or entierly with a view to eventually manufacturing and markting them."
Me: OK, now are you receiving any Compensation from any company, group, people, or any other entity? Whether it is money, perks, cheaper price on parts or other perk?
It is obvious you know some general knowledge, but there is a whole other level that you seem oblivous too. Not knowing some answers to some critical questions, the cheap attacks exposed that. Even you knowledge about the ribbon being a tweeter revealed alot.
I could care less about your schmatics. If you want to give yours away, that is your business. Amateur schematics never appealed to me much. But I certainly won't give out my 44 years of information.
Ok, let's take a look at your replies and see what we come up with. About half way down, we are going to make some important discoveries about your real knowledge vs what you try to lead us to believe.
"Well, the understanding of "Horns" and their drivers has SOMEWHAT improved in the the inervening 70 or so odd years since this was written, primarily due to demand from the Studio/Sound Reinforcemnt sector (the serious stuff) for high, undistorted SPL's. We may argue 9or not) the implications technically, but that is for loudspeakers....
My point is simple. A conventional moving coil speaker is subject to a number of limitations that make a high efficiency mandatory if high distortion and compression levels are to be avoided, that's just physics....
So, you are talking about what you think are "Horn" speakers, I merely talk about competently designed speakers."
Me: Unfortunately you don't seem to know that many horns use rear loading, which offers the "box" effect just like conventional speakers, with colorations too.
Secondly, according to the Radiotron Designers Handbook, "It is impossible to design a horn loudspeaker that covers a wide frequency band and is simultaneously free from non-linear distortion. Thus two separate units for low and high frequencies are essential for fidelity."
Next, distortion figures they gave have not been improved that much, so distortions are still coming from the throat, fm, loading enclosure etc.
Me: "but you failed to mention ribbons and air motion transformers.
You: "I have no problems with these per se, except they are tweeters, not speakers."
Me: Really, they go down to 100hz and they are tweeters? That is the best answer you can give?
You: "I am not getting into a "I tested more parts than you" ****ing contest. Unless we put numbers on the table there is little point and quantety sill does not ensure quality. Do you actually use Electrolytic Capacitors, for example, ANYWHERE? Then good by neutrality from the outset."
Me: Is that suppose to be a blockbuster comment? Did you have to copy that from my site? If you had really read my pages, you would have known the answer.
"Me: "As any design uses many parts; if those parts are not neutral sounding, one has to continually compensate for each part, including the tubes used,"
You: "Yup, unless one uses Valves that are from the start comparably neutral."
Me: So you figure if the tube is neutral sounding you can use any parts? That is ridiculous.
You: "What is ridiculous is your deliberate misstating what I have written. If you use Valves with a minimal inherent coloration level and passive parts that match this behaviour you end up with a fairly neutral sounding amplifier."
Me: Look at your reply again. I said neutral parts are necessary and you said "unless one usesneutral valves".
You stated it. The word "Unless" says alot. If you didn't mean it, great, cause it isn't true.
Me: "And are you getting 0.05% total distortion from All the previous stages (no cancelling distortion stages allowed) at max drive for full output of a 300b amp? (I include my preamp and non output stages of my amps.)"
You: "I have made driver stages that did provide this kind performance. In the real world I found that I find no further subjective improvement if the driver distortion is lowered to much more than around 12 - 20db below the Output Valves distortion AT FULL POWER fail to make further improvements.
And with all due respect, 0.05% Distortion from a 6922 providing around 20V RMS (triode connected EL34)? Do you mind if I qestion this sort of number?"
Me: If you found that to be the case, great, but the higher the harmonic, the lower the amplitude needs to be to be heard, as just mentioned above in another post. Remember, there is also 6th, 9th, 12th etc. and many others to consider that are lower order.
.05%, you seem to question while mixing two different circuits. The preamp cannot generate 20 volts rms with .05% distortion, but 2 volts at .02% or less. The SET amp doesn't either. But the PP amp using the 6922 does. OF course I didn't mention which amp, my aplogies. But the 6922 in the SET amp does provide the drive with only around .3% distortion. Actual measurement of course.
Me: "We both know you're not (getting .05% distortion) if you are using this tube near max. (refering to low distortion in the driver gain stage running near maximum output signal ratings.)
You: "And we both know that you do not get 15 - 20V RMS from a 6922."
Me: Calling me a liar won't help you.
A second indication of how little you know about designing. I can easily get 20v rms out of a 6922 with fairly small amounts of distortion, about the same percentage as you compared to the output tubes.
Me: "Yes, we both agree to keep harmonic distortion down.
You: "Yet you Driver valves that actually are quite non-linear and have quite high levels of upper harmonics."
Again not knowing what you are talking about. Some newer 6922s have virtually no higher order distortions. Don't make a fool out of yourself with halfwit comments.
And the preamp and driver in the PP amp, 6922s, only measures .05% at full drive for full output power, measured. Not measureable at signals for one watt out. Yes, gobs of higher harmonics, LOL.
You: "Again, I agree, yet I would question your use of Parts that minimise such distortion, based on what you have published about your designs on your webpages."
Me: Interesting trick trying to defame me. What parts do I even list that you question? List them for all of us. And do you think I am going to give you my designs? LOL. You are not that clever, nor superior.
You: "How do you arrive at that conclusion? A speaker that offers resonably resolution and neutrality in normal domestic settings with near realistic SPL's and provides very much an "Open Window" perspective does not need an amplifier to fit it. Assuming the Amplifier is competently designed and performas well the result will be good, even using (dread of dreads) solid state amplification. Naturaly there are differences in sound and presentation between different amplifiers and those are observable, easily so."
Me: Question. How do you know you are even in the ball park without listening first, which requires an amp? A speaker has much wider variation in sound quality than an amp. Go to any store and listen to them. Horrible ones. Anyone can throw drivers into a cabinet and get absolutely horrible sound, much worse than an amp. And single drivers not in a cabinet are not flat at all.
Can really mess up an amp design there.
You: "Hmmm. Has that got something to do with the fact that you make amplifiers? How do you know you are not compensating in the speaker design for deffects in amplifier performance? How can you judge the sonic performance of an amplifier without having first a speaker that is sufficiently resolving and neutral?
I view things from a system perspective. The biggest variables in ANY system are actually room interactions and the various distortion mechanisms in speakers (assuming a reasonably flat response). There is a LOT more difference here than there is in amplifiers...."
Me: So, simply take the speakers outdoors. That was an easy answer. Now we simplify things don't we. Work on the room later.
Second, I work back and forth between tweeking amps and tweeking speakers. However, tweeking an amp is such a minor things compared to radical surgery for speakers that are way off in accuracy. If a speaker is designed wrong it is hard to get it right.
Me: Didn't write down any systems, didn't think necessary. Seems after dozens of setups, I never seem to hear the right ones.
You: "Only in the loosest possible way in that I have either allowed companies to use my designs on condition of providing payments to childrens charities as "licence fee", plus in one case being not charged development costs on Transformers etc. the company later sells to the general market, which given the many iterations I tend to ask for is arguably a major benefit.
Oh yes, I also write the occasional review and tinker with all sorts of things, the results are usually published for general consumption.
And yes, I have been "holding back" some of what I consider "my best designs" in certain critical details or entierly with a view to eventually manufacturing and markting them."
Me: OK, now are you receiving any Compensation from any company, group, people, or any other entity? Whether it is money, perks, cheaper price on parts or other perk?
It is obvious you know some general knowledge, but there is a whole other level that you seem oblivous too. Not knowing some answers to some critical questions, the cheap attacks exposed that. Even you knowledge about the ribbon being a tweeter revealed alot.
I could care less about your schmatics. If you want to give yours away, that is your business. Amateur schematics never appealed to me much. But I certainly won't give out my 44 years of information.
- Status
- Not open for further replies.