Wavebourn said:
Actually, modern standards and taste means stereo (actually, 5 channels) above 200 Hz only, a single channel for frequencies below 200 Hz with a resonant peak around 40 Hz, and a digital equalizer. It is what I heard today in a factory outlet of one modern flagman of an audio industry, they had a show for potential buyers... Quite impressive show, but an artificial water running on the screen sounded very artificially.
Good audio systems were good always, I mean systems that can fool perceptions creating imaginations of live sources of sounds, but today some modern parts may be used to make them more affordable for the general public, but anyway good systems are still expensive.
I respectfully beg to differ. I'm not comparing apples to oranges. I am comparing high quality solid state stereo amplifiers and speakers (not home theater) of modern times to the same levels of previous generations of equipment. The difference is HUGE. Just hear recording from the 30's, 50's and 60's (I am listening mostly to jazz) and compare to those of the 90's and 00's and you can understand what I refer to. If they did not have studio recording equipment that was good enough theses days, you can assume that home equipment was at least as bad as the studio one. The difference is very clear on a double blind test, and I conducted some. The fact that equipment is based on tubes does not guarantee good quality. I have tube radios that sound awful, regardless of their aesthetic value.
So I really do not think that vintage equipment sound as good as high end equipment of our times.
I think that one assumption is that modern SS stuff (typical electronics store stuff) is even striving for the best sound. I would argue that they are most often striving for the most watts from the lowest parts count possible. They get good imaging by using 4 to 7 speakers. I love "turning heads" of people that listen to my system when they are looking for where my rear channel speakers are (I have none). HT seems to be "the thing" and most of the guys I know that are into audio are mainly HT guys. They have some great setups (some mega-bux), but they are not very musical to my ears when the TV is off.
Now I do think that the speakers in this range have improved a lot over the years. Efficiency has been lost due to the copious wattage available now, but a medium-fi pair of speakers sound a lot better than they did 20 or 30 years ago. It's almost pointless to build your own set of budget speakers anymore, because you probably can't outperform what you could get at Best Buy for the same money.
I don't fear the sands of these times. I have a Tubelab SE now as well and I can't stop listening to the thing long enough to take it off the breadboard. My other amps have been sitting quietly for months.
Now I do think that the speakers in this range have improved a lot over the years. Efficiency has been lost due to the copious wattage available now, but a medium-fi pair of speakers sound a lot better than they did 20 or 30 years ago. It's almost pointless to build your own set of budget speakers anymore, because you probably can't outperform what you could get at Best Buy for the same money.
I don't fear the sands of these times. I have a Tubelab SE now as well and I can't stop listening to the thing long enough to take it off the breadboard. My other amps have been sitting quietly for months.
Re: Re: Re: Modern tube amplifier designs?
Sorry to hear of that.
Using a single loop of NFB around the multi-stage amp is *NOT* the way to go I think. The power supply needs multiple de-coupled nodes... Now the high quality OPT is useful anywhere, but to select a design that *NEEDS* one seems rather short-sighted. Cathode FB with tertiary windings is one way...Chicago has several designs of that type. E-Linear connection of the two-stage amp is another Way...
cheers,
Douglas
richwalters said:
Sorry to hear of that.
Using a single loop of NFB around the multi-stage amp is *NOT* the way to go I think. The power supply needs multiple de-coupled nodes... Now the high quality OPT is useful anywhere, but to select a design that *NEEDS* one seems rather short-sighted. Cathode FB with tertiary windings is one way...Chicago has several designs of that type. E-Linear connection of the two-stage amp is another Way...
cheers,
Douglas
From reading this thread... I think I'm missing something, but there have been a lot of innovations that have occurred since the 50s and 60s.
There are now tube amps that have 20Hz square wave response without any tilt.
There are fully differential balanced amps, innovative all-tube, effective CCS circuits, there's been a lot of research into how the human ear/brain system works, and designs that reflect the knowledge that came from that research. There's been a lot of materials and component improvements as well.
However its important to understand that tubes are a mature technology. So any improvement in the technology is likely to be an evolution rather than revolution.
There are now tube amps that have 20Hz square wave response without any tilt.
There are fully differential balanced amps, innovative all-tube, effective CCS circuits, there's been a lot of research into how the human ear/brain system works, and designs that reflect the knowledge that came from that research. There's been a lot of materials and component improvements as well.
However its important to understand that tubes are a mature technology. So any improvement in the technology is likely to be an evolution rather than revolution.
I used to think that way too, until tweaking with different concepts enough to realize that It really comes down to application. The Williamson concept can make a good sounding amp. The input and phase splitter are single-ended, so a nice spectrum can be attained. I generally prefer the sound of them over PP amps that are balanced front to back. Especially when the output tubes are an easy load and can be driven strait from the concertina (though this isn't a true Williamson).
Pete Millets E-Linear amplifier is a good one. But the "partial feedback" concept has it's own issues which can cause it to be mis-applied just as global feedback can be. It's by no means the cure all.
The driver needs a large amount of headroom which can make things difficult and the allowable amount of feedback is usually very limited.
When applied to SE, it can become a real juggle. In many cases the closed loop PSRR will be worse than open loop. One can get good PSRR, by using the ulta-path/WE connection on the output, and then adjusting the feedback ratio. However, the point at which the best PSRR is attained may not be the point where best sonic performance is attained.
The distortion spectra, gain and THD, is unpredictable because the driver is not in the feedback loop. Only the output tube is. Distortion is minimized by the principal of inverse distortion cancellation and not by the typical NFB rules. At a particular feedback ratio in an SE amp, you will find that even-order harmonics have completely nulled, and the odd have gone up to their highest value. This is the point where THD is lowest. Feedback ratio's other than that you will see THD raise as the even-orders start to come back with a slight reduction in odd-order.
Once you have tweaked enough to get what you want, you will find that tube rolling can cause the spectra and THD to significantly change. In an experimental amp I built, THD could vary by 3 fold (between .3 and .9 % @ 1W) simply by tube rolling. Even using different samples of the same brand. I'm not they type that really cares that much about what an amp measures (if it sounds good and channels are pretty closely matched), but I do like consistency.
Partial feedback is most consistent and fail proof when applied to push pull amplifiers (could easily be adapted to a Williamson). Selecting a driver tube that has tight tube-to-tube tolerance is a good idea as well.
Re: Re: Re: Re: Modern tube amplifier designs?
That is still the old way....note the Ekland Plitron amp (1st mail) with all the credit commentary about it.
Perhaps I should have mentioned the standard diff driver stage is often the source of unseen problems in many designs. It isn't stiff enough.The basic Williamson concept can be massively improved by regenerative cathode-grid feedback within the stage which reduces IM thd and further reduces loading the concertina. The use of video pentodes as triodes (as you mention the right tube for the job) with modest current offers good low Z drive to deal with o/p stage Miller effect. It is the simplest method for driving multiple p-p pairs.
One type of feedback loop which I've never really got a real performance benefit is the inverse feedback within the output stage anode to previous driver cathodes. Perhaps I should examine this in more detail with a spectrum analyser.
The real problem is Miller effect from 4x 6550 which I recon is one of the hardest tubes for combination Miller capacitance of 650pF and low driver Z. It makes sense for fidelity to keep B+ down. It appears any driver stage faces a tough time whatever trouble is taken. Switchmode power designers know too well the currrent drive formula for mosfet current drivers. The same can be applied to tube stages.
I'd be honest that I get exceptional results from the "modified" Williamson concept, 90% due to good quality o/p trannies.
I can see further improvemnts with the regenerative feedback Williamson diff stage, just a matter of getting around planting them.
richy
Bandersnatch said:
That is still the old way....note the Ekland Plitron amp (1st mail) with all the credit commentary about it.
Perhaps I should have mentioned the standard diff driver stage is often the source of unseen problems in many designs. It isn't stiff enough.The basic Williamson concept can be massively improved by regenerative cathode-grid feedback within the stage which reduces IM thd and further reduces loading the concertina. The use of video pentodes as triodes (as you mention the right tube for the job) with modest current offers good low Z drive to deal with o/p stage Miller effect. It is the simplest method for driving multiple p-p pairs.
One type of feedback loop which I've never really got a real performance benefit is the inverse feedback within the output stage anode to previous driver cathodes. Perhaps I should examine this in more detail with a spectrum analyser.
The real problem is Miller effect from 4x 6550 which I recon is one of the hardest tubes for combination Miller capacitance of 650pF and low driver Z. It makes sense for fidelity to keep B+ down. It appears any driver stage faces a tough time whatever trouble is taken. Switchmode power designers know too well the currrent drive formula for mosfet current drivers. The same can be applied to tube stages.
I'd be honest that I get exceptional results from the "modified" Williamson concept, 90% due to good quality o/p trannies.
I can see further improvemnts with the regenerative feedback Williamson diff stage, just a matter of getting around planting them.
richy
I don't want to make myself out to be some kind of expert on this, because I'm really not.
I built an amplifier(KT88 push-pull Stereo amp) which was first choke filtered (Hammond 193Q). I then replaced the choke with a Maida regulator with values tweaked to give good regulation at 200-500mA. The amp sounded more powerful, probably due to the lower output impedance and lack of sag in the power supply. Plus, the amplifier was about 10 pounds lighter and $50 cheaper.
I used FQAF6N80 as a pass device, which is now obsolete. I really liked the TO-3PF case which is huge and isolated, so there is no concern with electrically insulating it. They also make insulating thermal pads which should work to insulate(electrically) the heatsink or chassis. I used one for the LM317 (I found out later that they make plastic case LM317s as well). The thermal pad I used said that it was good for 600V per mil and was six mils thick, so it shouldn't be a problem for it. Still, it is scary because it looks so thin.
Traditional
Not that I am much clever or have a lot to say, but sometimes I feel the urge to say something.
1) The Article: I do not see what is modern or innovative in the schematics given, except for the Plitron transformers (innovative because of using a toroid core, I understand).
2) There is nothing innovative about tubes: they are OLD and where used in an era differing from ours in a lot of respects. Therefore, if you build an amp with tubes, it is old-fashioned by definition.
3) Modern concepts should represent some forms of hybridization (intended as combination of tubes and transistors). This has been done, and is done regularly by diyers. Still, those amps are not tube amps, but hybrids, and do not fit the bill for the discussion.
4) Basically, there is not much possiblity to INVENT something new, due to the fact that tubes were around for a very long time and almost all we do today with tubes has at some point or another been done. Whether you can or cannot patent some circuit today is a different matter. My RH amps should be original, but I am aware that amps with that type of feedback where made and used 60 or 70 years ago: what I see as a modern contribution is given by simulations (and FFT for those who like or need doing it) enabling you a more efficient design process where you can achieve far better results than they could at the time. It is than a different issue whether a circuit done correctly 60 years later is the real deal, or the "originality" remains with the original cirucit, good or bad that it was...
5) Modern active components (i.e. sand) can be used to improve the functioning of tube amplifiers, and were not available before the 2nd world war, but were obviously available in the 70s or 80s (like SS diodes, LEDs, CMOS, FET, jFET, etc.). Therefore it is very disputable what would a modern 21st century tube amp be in respect to a modern late 20th century tube amp? I dare say it's mere "poseur" behaviour.
6) Finally, what is the real contribution of SS devices (supposedly modern) to the sound of tubes? While in some cases it might bring an improvement, in most cases it is deletrious, and only introduces unncessary complications. Actually, in most cases. The discussion about "SS diodes vs. tube rectifiers" is one example where people are ready to go to great lengths to proove their point, but are not ready to try for themselves (in most cases). Some who try claim that they do not hear the supposed advantages, and others claim that there are no advantages, quite the opposite -- de gustibus... The same is basically applicable to the LCLCLC vs. regulation. I have tried it, and can say that LCLCLC is better sounding, although not as cheap. De gustibus, again? And, I would like to add, active devices can add a lot of their noise if quality is not impeccable, which can make such devices either more expensive, or difficult to source.
7) Last but not least, what is the real facilitation offered by the use of SS devices? One of the most quoted is DC heating. Well, while DC heating is very welcome in phono preamps, it is quite unnecessary in line preamps, if the tube is not direct heated... In tube amps, DC heating is utterly unnecessary if tubes are not direct heated (I am reluctant to say that some wire is cheaper and simpler and thus better than diodes, caps, regulators and the like... for what, if there is no difference...). If on the other hand, the output tubes of the amp are direct heated, there is an array of touted SS solutions ranging from DC, thru regulated voltage, to regulated current -- and all much more complicated and eventually unnecessary if compared to AC heating.
Since this is the main point I wanted to address, let me clarify a little bit. Why would you need to DC heat DHTs?
1) Because you own very efficient loudspeakers (partially true).
2) Because you can't stand the hum (well, you could try using 2A3 instead of 6B4G, 2.5V hums less than 6.3)...
And what is the advantage of DC heating, except for hum? Well, the only thing that comes to mind is the fact that the cathode is constantly heated (i.e. at the same temperature, able to emit electrons at the same rate all the time... is it?). BUT, there is a drawback: the cathode is not at the same potential, because there is a difference in potential between one end and the other (the higher the voltage and the need for DC heating related to hum, the higher the difference in potential).
On the other hand, a DHT heated with AC as it was intended to be heated, will hum very little and imperceptibly so if the necessary heating voltage is low, it's cathode will be at the same potential, and if you are not a rocket scientist, you will not take note that the cathode is not heated the same all the time (due to the wave characteristics of AC, the heating actually fluctuates unlike DC heating which is constant)
Not to mention that a piece of wire is something everyone can do and be happy about, while most will have to resort to buying circuit boards designed and made by others, if they want to implement some form of "sophysticated" SS DC solution..
Not that I am much clever or have a lot to say, but sometimes I feel the urge to say something.
1) The Article: I do not see what is modern or innovative in the schematics given, except for the Plitron transformers (innovative because of using a toroid core, I understand).
2) There is nothing innovative about tubes: they are OLD and where used in an era differing from ours in a lot of respects. Therefore, if you build an amp with tubes, it is old-fashioned by definition.
3) Modern concepts should represent some forms of hybridization (intended as combination of tubes and transistors). This has been done, and is done regularly by diyers. Still, those amps are not tube amps, but hybrids, and do not fit the bill for the discussion.
4) Basically, there is not much possiblity to INVENT something new, due to the fact that tubes were around for a very long time and almost all we do today with tubes has at some point or another been done. Whether you can or cannot patent some circuit today is a different matter. My RH amps should be original, but I am aware that amps with that type of feedback where made and used 60 or 70 years ago: what I see as a modern contribution is given by simulations (and FFT for those who like or need doing it) enabling you a more efficient design process where you can achieve far better results than they could at the time. It is than a different issue whether a circuit done correctly 60 years later is the real deal, or the "originality" remains with the original cirucit, good or bad that it was...
5) Modern active components (i.e. sand) can be used to improve the functioning of tube amplifiers, and were not available before the 2nd world war, but were obviously available in the 70s or 80s (like SS diodes, LEDs, CMOS, FET, jFET, etc.). Therefore it is very disputable what would a modern 21st century tube amp be in respect to a modern late 20th century tube amp? I dare say it's mere "poseur" behaviour.
6) Finally, what is the real contribution of SS devices (supposedly modern) to the sound of tubes? While in some cases it might bring an improvement, in most cases it is deletrious, and only introduces unncessary complications. Actually, in most cases. The discussion about "SS diodes vs. tube rectifiers" is one example where people are ready to go to great lengths to proove their point, but are not ready to try for themselves (in most cases). Some who try claim that they do not hear the supposed advantages, and others claim that there are no advantages, quite the opposite -- de gustibus... The same is basically applicable to the LCLCLC vs. regulation. I have tried it, and can say that LCLCLC is better sounding, although not as cheap. De gustibus, again? And, I would like to add, active devices can add a lot of their noise if quality is not impeccable, which can make such devices either more expensive, or difficult to source.
7) Last but not least, what is the real facilitation offered by the use of SS devices? One of the most quoted is DC heating. Well, while DC heating is very welcome in phono preamps, it is quite unnecessary in line preamps, if the tube is not direct heated... In tube amps, DC heating is utterly unnecessary if tubes are not direct heated (I am reluctant to say that some wire is cheaper and simpler and thus better than diodes, caps, regulators and the like... for what, if there is no difference...). If on the other hand, the output tubes of the amp are direct heated, there is an array of touted SS solutions ranging from DC, thru regulated voltage, to regulated current -- and all much more complicated and eventually unnecessary if compared to AC heating.
Since this is the main point I wanted to address, let me clarify a little bit. Why would you need to DC heat DHTs?
1) Because you own very efficient loudspeakers (partially true).
2) Because you can't stand the hum (well, you could try using 2A3 instead of 6B4G, 2.5V hums less than 6.3)...
And what is the advantage of DC heating, except for hum? Well, the only thing that comes to mind is the fact that the cathode is constantly heated (i.e. at the same temperature, able to emit electrons at the same rate all the time... is it?). BUT, there is a drawback: the cathode is not at the same potential, because there is a difference in potential between one end and the other (the higher the voltage and the need for DC heating related to hum, the higher the difference in potential).
On the other hand, a DHT heated with AC as it was intended to be heated, will hum very little and imperceptibly so if the necessary heating voltage is low, it's cathode will be at the same potential, and if you are not a rocket scientist, you will not take note that the cathode is not heated the same all the time (due to the wave characteristics of AC, the heating actually fluctuates unlike DC heating which is constant)
Not to mention that a piece of wire is something everyone can do and be happy about, while most will have to resort to buying circuit boards designed and made by others, if they want to implement some form of "sophysticated" SS DC solution..
Re: Traditional
The sound is... otherwise the kudos would have gone decades ago. Good to see tube heritage at its best.
Xenu; That 813 matrix amp. 800V B+. Happy with the challenges ?The Lundahls with Swedish iron aren't cheap, reputed to be best sounding, but there are still alot of 813' around-
richy
Alex Kitic said:
The sound is... otherwise the kudos would have gone decades ago. Good to see tube heritage at its best.
Xenu; That 813 matrix amp. 800V B+. Happy with the challenges ?The Lundahls with Swedish iron aren't cheap, reputed to be best sounding, but there are still alot of 813' around-
richy
Take a look at this ongoing thread with 813 involved:
http://www.intactaudio.com/forum/viewtopic.php?t=787
http://www.intactaudio.com/forum/viewtopic.php?t=787
While some of my designs resemble to what John "Platinum Brain" Broskie has done in the past years, I do not necessarily acclaim his work, particularly not the hybrid/SS/sandish alternatives he has made himself quite known for in the last decade or so.
Than again, all these designs that I actually do not like, could or should be regarded as new an innovative.
It's an issue of evolution, not revolution. Let's just keep from evolving like the Audio CD has evolved into the mp3 CD (instead of evolving into a 24bit LPCM Audio DVD, like some were promising us...)
I have always liked tube amps for their sound but I too was wanting some of the inherent limitations to be removed. I have had EL34, KT88, 845 and 300B amps and ended up selling them and currently have a D class amp.
So, when JLTi made what I see to be a true innovation in tube amp design I put down the dollars: http://www.customanalogue.com/jlti_el34.htm
I will have my KT88 variant of this by the weekend and will let you know. JLTi/Joe Rasmussen has been behind a lot of DIY projects as you know so I am hoping my post is not out of context with this forum.
So, when JLTi made what I see to be a true innovation in tube amp design I put down the dollars: http://www.customanalogue.com/jlti_el34.htm
I will have my KT88 variant of this by the weekend and will let you know. JLTi/Joe Rasmussen has been behind a lot of DIY projects as you know so I am hoping my post is not out of context with this forum.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.