Hi all tubies!!!
I'm wonder what tube to use. The E88CC/6922 or ECC83/12AX7A?
It's going to be used in my to come balanced pre-amplifier that I'm planning to build.
Also, any body out there that have a good preamp curcit for any of the tube models? Not that I'm and beginner but it is always a very goof thing to get insperation on new ideas for other placed.
/Johan Ch
I'm wonder what tube to use. The E88CC/6922 or ECC83/12AX7A?
It's going to be used in my to come balanced pre-amplifier that I'm planning to build.
Also, any body out there that have a good preamp curcit for any of the tube models? Not that I'm and beginner but it is always a very goof thing to get insperation on new ideas for other placed.
/Johan Ch
Hi Johan,Johan_Ch said:
If you want to build someone else's design, the best balanced pre I've ever seen/heard are the RTP series at www.vacuumstate.com . Go to schematics and look at RTP3C or RTP5.
Personally, I would never use a 12__7 series tube. EVER. And the 6DJ8/6922 I find to be generally a peice of sonic pig misery (tm), except when applied with skill like Allen has, or in a cascode.
There are so many good tubes available out there for use in a pre, like 6C45, 6H30, ECC99, 6SN7, 12B4A, E55F (trioded) and about 30 or 40 others I can think of off the top of my head. Don't limit yourself to the two worst tubes in audiodom. There are only used in new gear because they're available new and cheap and in quantity, and because they are known to the tube ignorati. As a DIY, you have many fewer limitations than a manufaturer.
As frank mentions above, more detail is required about your intended circuit before more detail can be given.
Cheers
Brett
Re: TUBE CHOICE
I would have to answer your question with a resouinding yes, and no.
The 12--7 and 6922 families are not particularly linear, and the 12ax7 especially can't deliver enough current to many following stages (esp power) to slew them well enough for my liking. So as either, a phono or line pre tube, or a PA driver, they fail for me.
If you use them in more creative, and less usually applied ways, such as cascodes (with either a tube or fet on the bottom) or with an active load, you can get good-ish performance out of them, particularly the 6922. But except in a phono pre where you usually need the gain, why not use a 6H30 or 6SN7, say? if you look at the curves they are FAR more linear. For power tube drivers, these as well as trioded EL84/7189, trioded C3m (or either as a pentode!) as well as many others are a far better idea IMO.
I'm still of the opinion that most tubes used in commercial products are used because of availability, cost, familiarity and lack of imagination or daring by many designers. The only new tube I can recall getting any sort of 'mainstream' acceptability in recent times is the 6H30, used in some BAT products and hyped as the 'SuperTube'. Because of good marketing I have had some (non-DIY) 'philes tell me it's unavailable because BAT bought them all, so it's impact is limited. I have no trouble getting them, even in -DR. Same goes for the 6CW4 used in the MF Nuvistor products.
In my experience, the tubes and the topology (incl PSU) have about 80-90% responsibility for the sound of the circuit, with passives accounting for the rest. I try to use the most linear tube for the application, and have it in a good circuit to optimise performance.
Cheers
Brett
Hi Frank,fdegrove said:Brett,
Wouldn't you agree that it's not the tube that's at fault here but the way it was abused?
Thirty years of it at least....
On a more general note I tend to agree,being DIY sets no limits whatsoever to one possibilities.
Cheers,
I would have to answer your question with a resouinding yes, and no.
The 12--7 and 6922 families are not particularly linear, and the 12ax7 especially can't deliver enough current to many following stages (esp power) to slew them well enough for my liking. So as either, a phono or line pre tube, or a PA driver, they fail for me.
If you use them in more creative, and less usually applied ways, such as cascodes (with either a tube or fet on the bottom) or with an active load, you can get good-ish performance out of them, particularly the 6922. But except in a phono pre where you usually need the gain, why not use a 6H30 or 6SN7, say? if you look at the curves they are FAR more linear. For power tube drivers, these as well as trioded EL84/7189, trioded C3m (or either as a pentode!) as well as many others are a far better idea IMO.
I'm still of the opinion that most tubes used in commercial products are used because of availability, cost, familiarity and lack of imagination or daring by many designers. The only new tube I can recall getting any sort of 'mainstream' acceptability in recent times is the 6H30, used in some BAT products and hyped as the 'SuperTube'. Because of good marketing I have had some (non-DIY) 'philes tell me it's unavailable because BAT bought them all, so it's impact is limited. I have no trouble getting them, even in -DR. Same goes for the 6CW4 used in the MF Nuvistor products.
In my experience, the tubes and the topology (incl PSU) have about 80-90% responsibility for the sound of the circuit, with passives accounting for the rest. I try to use the most linear tube for the application, and have it in a good circuit to optimise performance.
Cheers
Brett
More info.
Some one aske for more specific information about what I'm going to do. It's a line pre-amplifier, to use between source and poweramplifiers.
Balanced beqause I have two balanced power amplifiers and a CD-player with balanced outputs.
To broaden the question, what kind of curcit to us? SRPP, catode follower etc.? I know the carastics of the different types of curcits regarding impedanses and function etc., but what I'm woundering is whitch curcit sounds the best in your oppinion?
/Johan Ch
Some one aske for more specific information about what I'm going to do. It's a line pre-amplifier, to use between source and poweramplifiers.
Balanced beqause I have two balanced power amplifiers and a CD-player with balanced outputs.
To broaden the question, what kind of curcit to us? SRPP, catode follower etc.? I know the carastics of the different types of curcits regarding impedanses and function etc., but what I'm woundering is whitch curcit sounds the best in your oppinion?
/Johan Ch
Re: Re: TUBE CHOICE
Brett, my man!
It's as if I hear myself talking throughout all of your post.
Tube lovers get a bad name amongst solid state fanatics due to the amount of badly designed ECC88 / 6DJ8 stuff out there. I'm not talking about Allen's designs, but the 10% H2 stuff that has the 'typical tube sound'. Good tubes are intrinsicly linear and in a good application, -90 dB THD+N without feedback is easily obtained. In cascode or active load (often with solid state sources!) applications, I've achieved up to -120 dB. Typical tube sound? Yuckie.
The main advantage of DIY is the freedom of tube selection, including rare ones, since no mass production will be taking place. You can select a tube that is linear and has low Rp as well, driving the next stage with a much lower Zout which helps obtain a very wide bandwidth with low distortion.
Regards,
Remco
Brett, my man!
It's as if I hear myself talking throughout all of your post.
Tube lovers get a bad name amongst solid state fanatics due to the amount of badly designed ECC88 / 6DJ8 stuff out there. I'm not talking about Allen's designs, but the 10% H2 stuff that has the 'typical tube sound'. Good tubes are intrinsicly linear and in a good application, -90 dB THD+N without feedback is easily obtained. In cascode or active load (often with solid state sources!) applications, I've achieved up to -120 dB. Typical tube sound? Yuckie.
The main advantage of DIY is the freedom of tube selection, including rare ones, since no mass production will be taking place. You can select a tube that is linear and has low Rp as well, driving the next stage with a much lower Zout which helps obtain a very wide bandwidth with low distortion.
Regards,
Remco
TUBE CHOICE
Guys,
You can count me in as adhering to the same principles.
But why cascodes?
They would give you more odd order distorsion than even order,no?
I would agree if you'ld say :hey,look this 6DJ8 was designed for cascode circuits,look what they do now they're using it for audio.
Yuk!
Well.IMHO that's not such a problem if you know your tube and know what you want to use it for.
My stomac turns when I see 300B's as voltage regulator (badly designed ones to boot) or even worse as a rectifier:now that MUST sound good right??
On the other hand,not that many tubes were actually designed for audio service so you're back to square one.
Me,I would rather design using simple but tested topologies,using medium to lowish mu triodes even if that means adding an extra stage.
Moreover it is important to know what a certain topology has to offer,all design parameters considered,and not cascode the same topology all over:that's how you create a definite sonic footprint.
I don't think it is good design practice to go for a very high gain stage that distorts so much you need to resort to NFB to linearise it.
Not that I totally reject NFB,just if you don't need it ,don't use it.
Moreover a good tube design does not sound like "tubesound" either.It sounds like live music!
And let's not forget the importance of the PSU,please.In my book that would account for 55% of the endresult.
In other words,it's no good to have a great "topology" if you feed it with junk form the PSU.
Also,what I dislike even more than a manufacturer buying up a whole supply of a certain still manufactured tube type is that some now even use types designed for computer use for instance(7044,amongst others),buy up whatever is left of it and letting Joe Public in the dark about that too.
Here they deserve to be sued in court IMO.
Johan,
In such a preamp you need a bit of amplification,good drive and low output Z.
Any of the named topologies can be used to achieve good results.
Which is not saying all that our out there are well designed.
I would do the balancing with transformers since I suppose that's what in the CD player already?
We would also need to know how the balancing is done on the amps side to be able to help you more.
Cheers,
Guys,
You can count me in as adhering to the same principles.
But why cascodes?
They would give you more odd order distorsion than even order,no?
I would agree if you'ld say :hey,look this 6DJ8 was designed for cascode circuits,look what they do now they're using it for audio.
Yuk!
Well.IMHO that's not such a problem if you know your tube and know what you want to use it for.
My stomac turns when I see 300B's as voltage regulator (badly designed ones to boot) or even worse as a rectifier:now that MUST sound good right??
On the other hand,not that many tubes were actually designed for audio service so you're back to square one.
Me,I would rather design using simple but tested topologies,using medium to lowish mu triodes even if that means adding an extra stage.
Moreover it is important to know what a certain topology has to offer,all design parameters considered,and not cascode the same topology all over:that's how you create a definite sonic footprint.
I don't think it is good design practice to go for a very high gain stage that distorts so much you need to resort to NFB to linearise it.
Not that I totally reject NFB,just if you don't need it ,don't use it.
Moreover a good tube design does not sound like "tubesound" either.It sounds like live music!
And let's not forget the importance of the PSU,please.In my book that would account for 55% of the endresult.
In other words,it's no good to have a great "topology" if you feed it with junk form the PSU.
Also,what I dislike even more than a manufacturer buying up a whole supply of a certain still manufactured tube type is that some now even use types designed for computer use for instance(7044,amongst others),buy up whatever is left of it and letting Joe Public in the dark about that too.
Here they deserve to be sued in court IMO.
Johan,
In such a preamp you need a bit of amplification,good drive and low output Z.
Any of the named topologies can be used to achieve good results.
Which is not saying all that our out there are well designed.
I would do the balancing with transformers since I suppose that's what in the CD player already?
We would also need to know how the balancing is done on the amps side to be able to help you more.
Cheers,
Re: TUBE CHOICE
The best tube circuit I know so far is choke- or OPT-loaded.
However, those are expensive and large. I have been trying to create a virtual choke in an electronic form.
After much listening, measuring and trials I think the value of merit for chokes is their different feedback for DC and AC. For DC their resistance is very low, so the tube sees strong feedback to keep it in its proper operating point. You may call it a constant current situation if you like. For AC, the choke has a high impedance, so the signal to the tube is not allowed to alter the operating point.
If you look at a regular RC loaded application, the behavior for DC and AC is exactly equal. DC drift is only poorly countered while AC signals, however small, constantly pull the tube out of its operating point. This is simply (local) feedback. Although this doesn't sound bad, I find eliminating the AC feedback and reinforcing the DC feedback sounds better.
The only way I found to eliminate this, apart from choke loading, is to put the tube in between current sources. This is what a cascode circuit does. I usually cascode the cascode as well (when it's made of transistors), since the output impedance of the cascode may be high, but it's not purely real. It generates harmonics. Cascoding the cascode relieves each casode. To top it off, I put a current source under the cathode. All current sources are biased by the same circuit and temperature compensated to make sure they all work together instead of apart.
Computer tubes: be aware that those were made for analog computers - i.e. they're very good. Take the 5687 as an example.
Regards,
Remco
Well-
The best tube circuit I know so far is choke- or OPT-loaded.
However, those are expensive and large. I have been trying to create a virtual choke in an electronic form.
After much listening, measuring and trials I think the value of merit for chokes is their different feedback for DC and AC. For DC their resistance is very low, so the tube sees strong feedback to keep it in its proper operating point. You may call it a constant current situation if you like. For AC, the choke has a high impedance, so the signal to the tube is not allowed to alter the operating point.
If you look at a regular RC loaded application, the behavior for DC and AC is exactly equal. DC drift is only poorly countered while AC signals, however small, constantly pull the tube out of its operating point. This is simply (local) feedback. Although this doesn't sound bad, I find eliminating the AC feedback and reinforcing the DC feedback sounds better.
The only way I found to eliminate this, apart from choke loading, is to put the tube in between current sources. This is what a cascode circuit does. I usually cascode the cascode as well (when it's made of transistors), since the output impedance of the cascode may be high, but it's not purely real. It generates harmonics. Cascoding the cascode relieves each casode. To top it off, I put a current source under the cathode. All current sources are biased by the same circuit and temperature compensated to make sure they all work together instead of apart.
My main problem with the 6DJ8 is the grid bias, which is at a lowish -1V. Furthermore, its curves are fairly nonlinear there. For small signals (phono etc) it's fairly linear, but you often see it applied where the grid variation is much larger, resulting in high H2. Allen Wright forces constant current so that the effect is much smaller. More or less equal to what I'm trying to do. That, i.e. constant current, is where the 6DJ8 is linear and it's what it was designed for. I myself am not commercial however, so I can choose a tube even more to my liking.
In tubes, you often can't get both a high mu and a high gm. For high bandwidth and low Zout, I often favor tubes with a high gm at the cost of a high mu. Besides, under constant current, you have maximum gain which is equal to mu, so I choose a tube with mu equal to the gain I want / need.
I think you're talking about cascoding tube X with tube X. That's not what I do!
Agreed! When people start about the typical tubesound, I often play some dancemusic at high volume to shake their believes. Nothing like hearing some Depeche Mode through 845's and looking at those jaws droppingMoreover a good tube design does not sound like "tubesound" either.It sounds like live music!
Agreed again. However, I don't like too many regulators. The regulator is in the signal path. A cascode works in exactly the same fashion so I'd rather have that and forgo the regulator.
Make it so! I'll come and testify.
Computer tubes: be aware that those were made for analog computers - i.e. they're very good. Take the 5687 as an example.
Regards,
Remco
Re: More info.
As Frank said ypu need a circuit with little or no gain and a bit of drive. Most CDPs will drive most amps into clipping at full output and no attenuation.
I don't like SRPP's. They sound so different depending upon what sort of load you want to drive with them that I think they're best reserved for 'internal' stages in amps. I don't like CF's either, the exception being Allen Wrights ones.
So, what would I do? use the AW CF in the RTP5 (maybe even with 6H30's in place of the first 6922's), or, a differential pair of 12B4a's loaded by a Lundahl LL1660PP transformer in step down mode, with a constant current sink in the tails. Bias around 25 - 30mA/tube and 90-100V on the plates. I would try it without cathode bypasses first. I don't have the datasheet in front of me so I'm guessing based on my memory of the curves, so the op point might need tweaking.
Another solution might be some Stevens and Billington TX1022 transformer volume controls, a switch, and no preamp at all. I doubt you'll need any gain.
Cheers
Brett
Johan,Johan_Ch said:
As Frank said ypu need a circuit with little or no gain and a bit of drive. Most CDPs will drive most amps into clipping at full output and no attenuation.
I don't like SRPP's. They sound so different depending upon what sort of load you want to drive with them that I think they're best reserved for 'internal' stages in amps. I don't like CF's either, the exception being Allen Wrights ones.
So, what would I do? use the AW CF in the RTP5 (maybe even with 6H30's in place of the first 6922's), or, a differential pair of 12B4a's loaded by a Lundahl LL1660PP transformer in step down mode, with a constant current sink in the tails. Bias around 25 - 30mA/tube and 90-100V on the plates. I would try it without cathode bypasses first. I don't have the datasheet in front of me so I'm guessing based on my memory of the curves, so the op point might need tweaking.
Another solution might be some Stevens and Billington TX1022 transformer volume controls, a switch, and no preamp at all. I doubt you'll need any gain.
Cheers
Brett
CASCODE
Hi Remco,et all,
I agree and do prefer a choke loaded anode over a resistive one,
but a cascode does not IMHO achieve the same thing unless you add semi-conductor devices.(or prehaps as you describe:cascode your cascode?)
It's a matter of taste perhaps but I'd rather not use a semiconductor where a tube or choke can do the same and often more reliably.
What you seem to ommit (unconsciously perhaps) is that a cascode actually acts as a penthode.
And the curves show that in abundance.
Now,in my book a penthode does not have the same linearity a good triode has and has a tendency to produce a predominantly odd order distortion characteristic.
Cascoding semi-conductors is a design choice I can understand but let's stick with tubes here ?
I agree,that's why I would use it where it belongs,namely at the early stage of a phono preamp.
Still I get the feeling you're looking at the datasheets from a tube manual and draw the wrong conclusions (as I said,it's just a feeling).
A tube used in a different circuit does not behave the same way as what is shown in the datasheet where most curves show the tube's behaviour under the given condition (i.e. most of the time a purely resistive anode load for a given B+ and grid bias)
(see,above as in the cascode,where two triodes become a penthode.)
But I suppose you know all that.
However,what repels me is the amount of compensation and addition of ss devices one needs to make it behave.Surely these devices will have an impact on the linearity of the circuit and will have a sound of their own too? My advice is to look at the curves under certain conditions and choose an operating point where said tube is most linear.
Or go for a tube that does fit the bill.
I agree,high mu and high perveance don't go together.
Hence I generally prefer lower mu and higher transconductance over high mu (less linear) and high Z out as you do yourself.
That's a rather unfortunate typo,it should have read: not cascade the same topology: not only topology,but tubetype too.
The task of the tubes was to perform switching (on/off tubes as they were called)
To my knowledge the 5687 wasn't designed for computer use but it was used in it.I consider that not the same.
For the rest of that bunch,nobody cared if they were noisier or microphonic as long as they were reliable bitches,euh switches,right?
I have as yet not found a use for them and I am not looking either.
Now if someone could please design some n-p-n and p-n-p complimentary tubes we would all be happy??
Best regards,
Hi Remco,et all,
I agree and do prefer a choke loaded anode over a resistive one,
but a cascode does not IMHO achieve the same thing unless you add semi-conductor devices.(or prehaps as you describe:cascode your cascode?)
It's a matter of taste perhaps but I'd rather not use a semiconductor where a tube or choke can do the same and often more reliably.
What you seem to ommit (unconsciously perhaps) is that a cascode actually acts as a penthode.
And the curves show that in abundance.
Now,in my book a penthode does not have the same linearity a good triode has and has a tendency to produce a predominantly odd order distortion characteristic.
Cascoding semi-conductors is a design choice I can understand but let's stick with tubes here ?
I agree,that's why I would use it where it belongs,namely at the early stage of a phono preamp.
Still I get the feeling you're looking at the datasheets from a tube manual and draw the wrong conclusions (as I said,it's just a feeling).
A tube used in a different circuit does not behave the same way as what is shown in the datasheet where most curves show the tube's behaviour under the given condition (i.e. most of the time a purely resistive anode load for a given B+ and grid bias)
(see,above as in the cascode,where two triodes become a penthode.)
But I suppose you know all that.
However,what repels me is the amount of compensation and addition of ss devices one needs to make it behave.Surely these devices will have an impact on the linearity of the circuit and will have a sound of their own too?
My advice is to look at the curves under certain conditions and choose an operating point where said tube is most linear.Or go for a tube that does fit the bill.
I agree,high mu and high perveance don't go together.
Hence I generally prefer lower mu and higher transconductance over high mu (less linear) and high Z out as you do yourself.
That's a rather unfortunate typo,it should have read: not cascade the same topology: not only topology,but tubetype too.
The task of the tubes was to perform switching (on/off tubes as they were called)
To my knowledge the 5687 wasn't designed for computer use but it was used in it.I consider that not the same.
For the rest of that bunch,nobody cared if they were noisier or microphonic as long as they were reliable bitches,euh switches,right?
I have as yet not found a use for them and I am not looking either.
Now if someone could please design some n-p-n and p-n-p complimentary tubes we would all be happy??
Best regards,
Re: CASCODE
Regards,
Remco
True, but I like the research. I could just order a choke and be done; hobby over. Trying new stuff is what keeps me here.
If it passes signal, yes. In cascoding however it's just trying to follow a DC reference while someone or something is pulling at its Vce / Vds / Vak, which boils down to sliding left and right over the flat output characteristics. Vg/Vg/Vb remains the same. Harmonics of DC are rare...
Nope - I don't like transistors in the direct signal path but I can stand them in doing things they're better at than tubes. And this is one of those points. I would not want to close my eyes to modern technology, rather use the best of both worlds at the right places.Cascoding semi-conductors is a design choice I can understand but let's stick with tubes here ?
I'm using a spectrum analyzer to test and verify all my circuits and all of the modification work I do. And believe me, some commercial stuff is pretty bad. (And the Understatement Of The Year award goes to: ultranalog!)
Sure it does. I'm not saying I'm there already. But as to linearity: it's a whole lot better than RC. And it sounds better too. I'd say it goes 3/4 of the distance between RC and chokeload. In some circuits that, combined with the lesser cost and space consumption, is more than enough for me.
Regards,
Remco
CASCODE
Hi Remco,
Glad to see you're still alive and kicking!
I get the feeling we had a slight semantics problem:
What I think you do is isolate the amplifying triode from it's PSU with a cascode instead of say,using a penthode as was once proposed to improve on SRPP (mu follower) circuit ?
So summarizing : you use a cascode circuit on top of a triode to work as a CCS ?
And NOT as I originally thought stabilizing a classic cascode circuit by using semi-conductors as a ccs?
**By all means expimeriment to your hearts content**
Cheers,
Hi Remco,
Glad to see you're still alive and kicking!
I get the feeling we had a slight semantics problem:
What I think you do is isolate the amplifying triode from it's PSU with a cascode instead of say,using a penthode as was once proposed to improve on SRPP (mu follower) circuit ?
So summarizing : you use a cascode circuit on top of a triode to work as a CCS ?
And NOT as I originally thought stabilizing a classic cascode circuit by using semi-conductors as a ccs?
**By all means expimeriment to your hearts content**
Cheers,
Re: CASCODE
As to the srpp or mu-follower:
- srpp doesn't work. The assumption of a constant current is only valid for no load. Current, and hence output impedance and bandwidth, vary with load and signal, rendering the circuit unusable unless hugely overdimensioned (design for 1A if you need 100 mA).
- mu follower tries to do both infinite plateload (constant current) and combining a cathode follower with a gainstage. Obviously, it can't do both, so every mu-stage has to be designed to meet a target impedance and/or bandwidth. It can work great, but is often misdesigned.
I'm simply trying to emulate choke behavior, since it's the reference I have to work with. The AC - DC feedback system is at the basis of a choke's merit IMHO. Most tube gear uses a choke or OPT and hence gets more linear as frequency increases. The energy content of audio is roughly 1/f correlated, so it would be fair to say that chokeloads have a constant resolution throughout the band.
Regards,
Remco
The rumours of my death have been greatly exaggeratedGlad to see you're still alive and kicking!
Yes. And one under the triode as well. And one cascoding the cascode. The triode has no say in determining its operating point
As to the srpp or mu-follower:
- srpp doesn't work. The assumption of a constant current is only valid for no load. Current, and hence output impedance and bandwidth, vary with load and signal, rendering the circuit unusable unless hugely overdimensioned (design for 1A if you need 100 mA).
- mu follower tries to do both infinite plateload (constant current) and combining a cathode follower with a gainstage. Obviously, it can't do both, so every mu-stage has to be designed to meet a target impedance and/or bandwidth. It can work great, but is often misdesigned.
I'm simply trying to emulate choke behavior, since it's the reference I have to work with. The AC - DC feedback system is at the basis of a choke's merit IMHO. Most tube gear uses a choke or OPT and hence gets more linear as frequency increases. The energy content of audio is roughly 1/f correlated, so it would be fair to say that chokeloads have a constant resolution throughout the band.
No, but that may be fun, just for the idiocracy of itAnd NOT as I originally thought stabilizing a classic cascode circuit by using semi-conductors as a ccs?
Regards,
Remco
As suggested earlier, I took a look at Allan Wright's RTP3C.
Any body that have tested it IRL? I mean that it ain't always that somthing that looks good on paper actuly sounds good.
Any way, I'm going to give it a try.
To awnser some earlier questions about what I'm actuly going to do.
I have build a DAC with a Crystal D/A-converter that leavs a balanced signal. The output stage, or actuly more of a buffer, is made up of a 12AX7 tube, one fore each channel.
So, what I'm going to build is a pre-amplifier with input selector and volume controll.
/Johan Christensson
Any body that have tested it IRL? I mean that it ain't always that somthing that looks good on paper actuly sounds good.
Any way, I'm going to give it a try.
To awnser some earlier questions about what I'm actuly going to do.
I have build a DAC with a Crystal D/A-converter that leavs a balanced signal. The output stage, or actuly more of a buffer, is made up of a 12AX7 tube, one fore each channel.
So, what I'm going to build is a pre-amplifier with input selector and volume controll.
/Johan Christensson
CASCODE
Remco and all academic tubeofiles still following,
I can't agree more here,it's a misnomer from the the day of its birth.
Shunt regulated push-pull??No way.
From a practical day to day implementation it can be used succesfully to achieve some design needs.I won't send it to the recycle bin just yet.
In some situations it can be used as an elegant solution to a common problem where e.g commercial constraints apply.
Then again most of what we see in the real world are just that and then some.
Let's just say it started out with misleading credentials.
(Life would be all too easy if one wouldn't have to account for the effect of the next stage loading down the entire current flow,inversion of absolute phase in the design and its effect on the entire circuit and so on)
Which brings us gently to your work..
To me it only holds water if you know exactly what you're aiming for.
In your case let's call it a study of the art of designing the wholy grail?Operating a tube in an ideal environment ?
After you've done that the hardest part is to compress (computer speak admitting) (synopsis being the better word) the entire situation into a usable simplified equivalent.
From what I derive from the data provided - and it isn't easy without seeing the whole picture- your design goal is to "lock-tite" the triode amplifying its input signal.
Ok,provided one doesn't starve the poor tube (Class A only here please) you will have a perfect lock down.
Admittedly purely academical and taking load of the next stage into account.
Practical experience will be the best teacher...
All in all,some friends at the university of Leuven more or less ran the same experiment a few years back.(on or off the official records,I don't dare to say).
Don't let that hold you back and report back to this wonderful community,please.
Yrs sincerely
Remco and all academic tubeofiles still following,
I can't agree more here,it's a misnomer from the the day of its birth.
Shunt regulated push-pull??No way.
From a practical day to day implementation it can be used succesfully to achieve some design needs.I won't send it to the recycle bin just yet.
In some situations it can be used as an elegant solution to a common problem where e.g commercial constraints apply.
Then again most of what we see in the real world are just that and then some.
Let's just say it started out with misleading credentials.
(Life would be all too easy if one wouldn't have to account for the effect of the next stage loading down the entire current flow,inversion of absolute phase in the design and its effect on the entire circuit and so on)
Which brings us gently to your work..
To me it only holds water if you know exactly what you're aiming for.
In your case let's call it a study of the art of designing the wholy grail?Operating a tube in an ideal environment ?
After you've done that the hardest part is to compress (computer speak admitting) (synopsis being the better word) the entire situation into a usable simplified equivalent.
From what I derive from the data provided - and it isn't easy without seeing the whole picture- your design goal is to "lock-tite" the triode amplifying its input signal.
Ok,provided one doesn't starve the poor tube (Class A only here please) you will have a perfect lock down.
Admittedly purely academical and taking load of the next stage into account.
Practical experience will be the best teacher...
All in all,some friends at the university of Leuven more or less ran the same experiment a few years back.(on or off the official records,I don't dare to say).
Don't let that hold you back and report back to this wonderful community,please.
Next lab perhaps?
Yrs sincerely
Re: CASCODE
Gee, I never knew I was the 'academic' type
Regards,
Remco
Gee, I never knew I was the 'academic' type
I had it in mind like this: designing the ideal line stage with (until now) the chokeload as reference. Come what may (I don't feel bound to tubes). The fun in the current design is that it uses only one tube per channel. I've built one stereo linestage with a nuvistor, and the whole circuit is smaller than a single miniature tube. I better never find that holy grill, experimenting is all the fun...
That's why the next preamp I'm building will have 'Tektronix-style' swappable linestages. Then I won't have to build a whole amp for every idea that makes it off the testbench. The ultimate proof is still putting on some music and watching if the foot will tap...
Regards,
Remco
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.