Hi
I still have your two tubes I haven't used yet. My immediate project is to make a 2 stage preamp. This has to have enough voltage gain to drive an output stage which I'm going to try in solid state. In the first instance I'll use a PP 2a3 output stage. Then I want to try a circlotron with lateral mosfets - Andrea Ciuffoli has a design called Amp End.
So my new plan is not a single stage 26 preamp but a two stage one with a PP low impedance output. Output transformer will be a Lundahl SE
P one, probably LL1660/18mA. The second stage could well be the 4P1L. Previous idea was a 46. I also am considering two 26 stages ending in LL1660/5ma in 4.5:1 stepdown, again SEP.
What do you think?
Andy
I still have your two tubes I haven't used yet. My immediate project is to make a 2 stage preamp. This has to have enough voltage gain to drive an output stage which I'm going to try in solid state. In the first instance I'll use a PP 2a3 output stage. Then I want to try a circlotron with lateral mosfets - Andrea Ciuffoli has a design called Amp End.
So my new plan is not a single stage 26 preamp but a two stage one with a PP low impedance output. Output transformer will be a Lundahl SE
P one, probably LL1660/18mA. The second stage could well be the 4P1L. Previous idea was a 46. I also am considering two 26 stages ending in LL1660/5ma in 4.5:1 stepdown, again SEP. What do you think?
Andy
You introduced me to the 46, and in my opinion it sounds as probably no other DHT.
I haven't tried yet the 4p1L with the LL1660, which is something that Iko suggested earlier.
I tried 10y, 112a, 01a, 31, 45,46, 26 and some other DHTs all with filament bias and a good PS and think I agree with Iko: with a good PS is difficult to notice a big difference between the best performing DHTs: 26, 46 and 4P1L in my view ...
I have to say I can't help further as its me who is learning from you
I haven't tried yet the 4p1L with the LL1660, which is something that Iko suggested earlier.
I tried 10y, 112a, 01a, 31, 45,46, 26 and some other DHTs all with filament bias and a good PS and think I agree with Iko: with a good PS is difficult to notice a big difference between the best performing DHTs: 26, 46 and 4P1L in my view ...
I have to say I can't help further as its me who is learning from you
I also am considering two 26 stages ending in LL1660/5ma in 4.5:1 stepdown, again SE
What do you think?
Andy
Not a good idea IMO.
You will easily overload the grid of the second 26.
A 46 will be better with some -30 grid voltage; besides it has lower Rp, and still enough gain to drive the circlotron stage directly through the interstage transformer.
Ah right! The F4 stage. We are thinking along the same lines here. I considered that but it seemed a bit complex for what it did, which is effectively just provide a current gain output stage if I understand this right - I have little ss experience.
I was considering Susan Parker's Zeus output stage - She's a friend and I have heard her system which is extremely clear and natural. Definitely works, but uses an output transformer for the usual lateral mosfets. Her schemo is online. Andrea's circlotron has no OPT so therefore considerably cheaper and possibly more transparent.
I've just decided that the ridiculous prices for 300b tubes (or PP 2a3) are not something I want to pay, plus good OPTs are equally expensive. Doesn't seem cost effective and I'm not even sure it's a better solution than a really clean mosfet output stage.
So my 26 preamp is going to be a two stage one now, and will then connect to some kind of PP output stage whatever that is - I can experiment. That's my current thinking. Susan Parker is also experimenting with a tube input stage for her Zeus - seems it's in the air.
The 46 is a no-brainer, as Pieter says, but curious about the 4P1L.
andy
I was considering Susan Parker's Zeus output stage - She's a friend and I have heard her system which is extremely clear and natural. Definitely works, but uses an output transformer for the usual lateral mosfets. Her schemo is online. Andrea's circlotron has no OPT so therefore considerably cheaper and possibly more transparent.
I've just decided that the ridiculous prices for 300b tubes (or PP 2a3) are not something I want to pay, plus good OPTs are equally expensive. Doesn't seem cost effective and I'm not even sure it's a better solution than a really clean mosfet output stage.
So my 26 preamp is going to be a two stage one now, and will then connect to some kind of PP output stage whatever that is - I can experiment. That's my current thinking. Susan Parker is also experimenting with a tube input stage for her Zeus - seems it's in the air.
The 46 is a no-brainer, as Pieter says, but curious about the 4P1L.
andy
Last edited:
Let's just say that I trust Nelson.
How does Susan's system compare to a tube amp sound wise, in your opinion?
It'll be interesting to see what you end up with. I'm still looking for affordable power tube alternatives. I have a gm-70 amp in mind, and looking at some other transmitting tubes. I find that it's not so much the price of tubes that's the real problem, but the price of iron.
I'm happy enough with my 26 preamp now; enough not to be chasing uber-expensive 46 tubes. The 4P1L though is much easier to get; in fact I have a number of them already.
Susan Parker's system was very clean - I was very surprised at the amount of low level detail. Hearing new things in the mix - I thought I was getting low level detail with tubes, but I'm wondering if this wasn't a tad clearer. The other striking thing was that the piano tone was extremely realistic. I was seriously impressed - genuine high end. Also, it had plenty of headroom so the sound was effortless. She was driving Jordan units in MTM. I have Alpair 10 full range - similar kind of thing - and my 300b output stage was struggling, and no doubt distortion was higher. Another reason for aiming at about 20 watts minimum into my speakers.
What's your plan for the F4 - how many output mosfets and what power? Is solid state your thing? As far as I can see from the schemo there's a 220uF cap in the signal path - that would worry me.
andy
What's your plan for the F4 - how many output mosfets and what power? Is solid state your thing? As far as I can see from the schemo there's a 220uF cap in the signal path - that would worry me.
andy
Last edited:
Andy,
Nice to hear confirmation of the quality of Susan's amplifier.
You are right: Andrea's circlotron output stage has no OPT, but that does not mean it might be more transparent.
With step-down output transformer you will have better damping and less distortion; close comparison of Susan's and Andrea's FFT distortion curves confirm this.
Yes, prices for the popular power valves are over the top, and their life expectancy is restricted; output transformers however normally survive us all so I think it is a worthwile investment for a good design.
You might know I am building a hybrid amp for a customer, and finally I decided to take Andrea's input stage (D3a - phase splitting interstage transformer) and Susan's output stage with 2:1 output transformer.
This way I have more pure class A output power, less distortion and bullet proof loudspeaker protection.
Andrea's circlotron needs active loudspeaker protection, and the circlotron power supply is quite a bit more complex and expensive (multiple floating power supplies with multiple rectifier diodes, chokes, capacitors) whereas Susan's output stage needs just a single positive power supply.
Be aware that a circlotron output stage is functionally not different from Susan's output stage. "Circlotron" seems to have (undeserved IMO) magic for some people.
For your driving stage I understand you will stay truthfull to your 26, but you should make a good choice for the possible other valve, and I don't think the 4P1L is a good candidate.
Let's say you put 1 VRMS into the grid of your 26 input tube, not unrealistic IMO.
With the inductive load it will amplify 8 times, so 8 VRMS comes out for the second valve.
8 VRMS is over 22 V p-p, so your second valve should have at least a negative grid voltage of 11; the 4P1L has lower grid voltage at a good working point, so the conclusion should be: out she goes...; besides she has too much gain for us.
A better approach is: what output power do you aim for.
You say that a 300B SE is not enough for your set up (apart from over the top prices for this valve); so let us say you want something in the order of 30 watts.
Why 30 watts? A SE 300B will deliver some 7-8 watts; clipping behavior of this amplifier will be friendly compared to a SS amplifier. You'd need a 15 watt SS amp for the same behavior; when 15 watts will not do, at least double it to 30 watts.
30 watts in an 8 ohms load requires 15,5 VRMS, so, taking Andrea's circlotron, you need some x3 of voltage gain between 26 valve and output stage (the output stage will have a gain of less than 1). A 1:1,5+1,5 interstage transformer between 26 and output stage would already give the required gain, but the 26 valve has not enough balls (Rp, current) for a good result. Possible is to couple the 26 to a sourcefollower, and put a 1:1,5+1,5 interstage transformer in the source. This way you need only your 26 valve, the rest is SS, with excellent drive for your output stage (it is much like Susan's topology).
For the Zeus output stage with this output power you would need a 1:2+2 IT.
The Zeus output stage will give more pure class A power than the naked circlotron (some 20 and 10 watts resp. when I am right).
Nice to hear confirmation of the quality of Susan's amplifier.
You are right: Andrea's circlotron output stage has no OPT, but that does not mean it might be more transparent.
With step-down output transformer you will have better damping and less distortion; close comparison of Susan's and Andrea's FFT distortion curves confirm this.
Yes, prices for the popular power valves are over the top, and their life expectancy is restricted; output transformers however normally survive us all so I think it is a worthwile investment for a good design.
You might know I am building a hybrid amp for a customer, and finally I decided to take Andrea's input stage (D3a - phase splitting interstage transformer) and Susan's output stage with 2:1 output transformer.
This way I have more pure class A output power, less distortion and bullet proof loudspeaker protection.
Andrea's circlotron needs active loudspeaker protection, and the circlotron power supply is quite a bit more complex and expensive (multiple floating power supplies with multiple rectifier diodes, chokes, capacitors) whereas Susan's output stage needs just a single positive power supply.
Be aware that a circlotron output stage is functionally not different from Susan's output stage. "Circlotron" seems to have (undeserved IMO) magic for some people.
For your driving stage I understand you will stay truthfull to your 26, but you should make a good choice for the possible other valve, and I don't think the 4P1L is a good candidate.
Let's say you put 1 VRMS into the grid of your 26 input tube, not unrealistic IMO.
With the inductive load it will amplify 8 times, so 8 VRMS comes out for the second valve.
8 VRMS is over 22 V p-p, so your second valve should have at least a negative grid voltage of 11; the 4P1L has lower grid voltage at a good working point, so the conclusion should be: out she goes...; besides she has too much gain for us.
A better approach is: what output power do you aim for.
You say that a 300B SE is not enough for your set up (apart from over the top prices for this valve); so let us say you want something in the order of 30 watts.
Why 30 watts? A SE 300B will deliver some 7-8 watts; clipping behavior of this amplifier will be friendly compared to a SS amplifier. You'd need a 15 watt SS amp for the same behavior; when 15 watts will not do, at least double it to 30 watts.
30 watts in an 8 ohms load requires 15,5 VRMS, so, taking Andrea's circlotron, you need some x3 of voltage gain between 26 valve and output stage (the output stage will have a gain of less than 1). A 1:1,5+1,5 interstage transformer between 26 and output stage would already give the required gain, but the 26 valve has not enough balls (Rp, current) for a good result. Possible is to couple the 26 to a sourcefollower, and put a 1:1,5+1,5 interstage transformer in the source. This way you need only your 26 valve, the rest is SS, with excellent drive for your output stage (it is much like Susan's topology).
For the Zeus output stage with this output power you would need a 1:2+2 IT.
The Zeus output stage will give more pure class A power than the naked circlotron (some 20 and 10 watts resp. when I am right).
Hello Pieter - that's really helpful. Are you making suitable output transformers for Susan's output stage? If you are, can you email me on performanceandmedia@gmail.com? I'm not specially keen on a circlotron - as you say, I'd prefer to protect my speakers!! Her amp sounds excellent so I don't doubt the quality.
I'd also like to know your option for a mosfet stage after the 26, if you could give me a schematic (remembering I have VERY basic knowledge of solid state devices). Certainly her preamp was contributing to the good sound. She is trying out an input stage with D3A right now - I'd like to stick to 26.
best
andy
I'd also like to know your option for a mosfet stage after the 26, if you could give me a schematic (remembering I have VERY basic knowledge of solid state devices). Certainly her preamp was contributing to the good sound. She is trying out an input stage with D3A right now - I'd like to stick to 26.
best
andy
Hi Andy,
Risk of getting too OT, but I can say that the 220uFs are required in all case for bias stability & noise suppression.
The multiple device pairs are used to reduce output impedance (gm increases!), which in turn reduces distortion. The nature of the amp would be very different for single pairs.
And before setting out to build, please buy the 2SK170 and J74 first - they are hurtling toward the land of unobtanium.
For my part, I pay out the money for an DHT-SE output transformer gladly - because they do not wear out! They give a lifetime of pleasure, and I need them every day - which makes them a bargain! I would buy again if someone stole my Sowter SA08s, and I would pay the highest price for best quality OT. Even speaking as a solid state design pro, I believe that it is unlikely that I would move away from DHT-SE power amps!
This is so, because even with the cheap ElectroHarmonix 300B I can compare listening to live orchestras - without feeling I am missing anything important, and the stereo presentation of the parts is better than most seating positions! Sure, this is only possible when you have everything right - power supply, filament supply (fully solid state), driver circuit (partly solid state) and speakers suitable and equal to the amps.
At the device level, the output stage is the one place where DHTs are unmatchable. At lower levels, the JFETs 2SK170, J74 and (for SE) the K369 can get very close, and even do things that valves cannot do (MC first stage, for sure). But Nelson's power amp with zillions of JFETs, or the Power JFET version, points to the shortcomings of ordinary power MOS in this position.
OK, I'm saying - Have fun with the F-4, but please compare the finished article to top-rank 300B-SE ! I would not like to see you miss out.
Risk of getting too OT, but I can say that the 220uFs are required in all case for bias stability & noise suppression.
The multiple device pairs are used to reduce output impedance (gm increases!), which in turn reduces distortion. The nature of the amp would be very different for single pairs.
And before setting out to build, please buy the 2SK170 and J74 first - they are hurtling toward the land of unobtanium.
For my part, I pay out the money for an DHT-SE output transformer gladly - because they do not wear out! They give a lifetime of pleasure, and I need them every day - which makes them a bargain! I would buy again if someone stole my Sowter SA08s, and I would pay the highest price for best quality OT. Even speaking as a solid state design pro, I believe that it is unlikely that I would move away from DHT-SE power amps!
This is so, because even with the cheap ElectroHarmonix 300B I can compare listening to live orchestras - without feeling I am missing anything important, and the stereo presentation of the parts is better than most seating positions! Sure, this is only possible when you have everything right - power supply, filament supply (fully solid state), driver circuit (partly solid state) and speakers suitable and equal to the amps.
At the device level, the output stage is the one place where DHTs are unmatchable. At lower levels, the JFETs 2SK170, J74 and (for SE) the K369 can get very close, and even do things that valves cannot do (MC first stage, for sure). But Nelson's power amp with zillions of JFETs, or the Power JFET version, points to the shortcomings of ordinary power MOS in this position.
OK, I'm saying - Have fun with the F-4, but please compare the finished article to top-rank 300B-SE ! I would not like to see you miss out.
Hi Rod - please remain off topic! I think for a few of us it's worth going through options of what we connect our 26 preamps onwards to.
I can see the LSK170, 2SK369 and the 2SK170 - are they all still obtainable? Do you see these as good SE choices directly after the 26 stage to use into a transformer to do the phase splitting? I'm also thinking PP2a3 here.
Presumably a 2SK170 stage would give low output impedence and allow a transformer volume control. This is getting into Thomas Mayer's territory.
andy
I can see the LSK170, 2SK369 and the 2SK170 - are they all still obtainable? Do you see these as good SE choices directly after the 26 stage to use into a transformer to do the phase splitting? I'm also thinking PP2a3 here.
Presumably a 2SK170 stage would give low output impedence and allow a transformer volume control. This is getting into Thomas Mayer's territory.
andy
I agree, many are adding headphone outs to the pre which the hybrid stuff applies too as well (historically a TOL pre has a headphone out, and trust me adding a 120ohm tap isn't the best solution.).
But speaking of unobtainable, I thought the zeux output was no longer an option as the mosfets are specific and obsolete, I did search but couldn't find that thread (not on SS much.)
You mean the Zeus output stage?
It is part of Susan Parker's design; you will find a long thread at this forum.
The output stage is a transformer coupled design, actually nothing special topology wise; it does not depend on one particular type of power fet.