Hi,
As Iam dumb in tubes, for me is easy to answer this question.
In my personal taste(of today) would be the Sakuma SET with self-adjusting Bias, if it is possible, or auto-Bias in second place.
harmonics - with the caveat that you don't pour in tons of feedback to linearise the amp.
No tube is truly linear over its entire operating range, so some distortion is pretty much assured. In any topology but especially SE, and assuming no bizarre feedback tricks, an amount of that is going to get through to the speaker.
It MAY be pleasant - but it will not be transparent or without colour.
No tube is truly linear over its entire operating range, so some distortion is pretty much assured. In any topology but especially SE, and assuming no bizarre feedback tricks, an amount of that is going to get through to the speaker.
It MAY be pleasant - but it will not be transparent or without colour.
...when it is linear enough without tons of feedback. But if not enough, I would pour rather many tons than few kilograms. Why? Because pouring a little bit of feedback I would reduce total HD number a little bit adding higher order of distortions. Pouring tons, I reduce distortions many tons below their previous level.
And, you know, a SE amp with tons of FB is still Single Ended, no **t!
There may be several factors at work in transparency. There is no lack of transparency in my 300b SET - it's the most detailed and clear amp I've ever built in 20 years. I attribute that to using all DHT tubes, but the topology helps.
Topology has taken some experimenting. For the first choke loaded 26 stage I ended up with a choice between battery bias on the grid and filament bias. I preferred filament bias, but only with a choke input filament supply feeding Rod Coleman boards, since this depends on the DC being super clean. I have a separate choke input HT supply for this stage with an AZ1 mesh rectifier and all polypropylene caps. I really have maxxed out on this stage, since everything after it depends on it. I even tried out 8 different kinds of 10 ohm cathode resistor before settling on the best sounding.
I then have two versions in different amps - direct coupled in one and cap coupled (Teflon FT-3) in another. Second stage is a 46 into a Hammond 126C interstage (or Lundahl LL1660/18mA). I found I could leave out the cathode bypass on the cathode resistor so I have.
Finals are 300b in cathode bias and here the cathode resistor is bypassed with a 100uF polypropylene cap.
I'm not too sure what could improve this, and certainly open to suggestions! I could battery bias the grid of the 300b for instance. But I'm pretty happy. It's quite a big amp on a few chassis, but I felt the topology was worth it.
Andy
Topology has taken some experimenting. For the first choke loaded 26 stage I ended up with a choice between battery bias on the grid and filament bias. I preferred filament bias, but only with a choke input filament supply feeding Rod Coleman boards, since this depends on the DC being super clean. I have a separate choke input HT supply for this stage with an AZ1 mesh rectifier and all polypropylene caps. I really have maxxed out on this stage, since everything after it depends on it. I even tried out 8 different kinds of 10 ohm cathode resistor before settling on the best sounding.
I then have two versions in different amps - direct coupled in one and cap coupled (Teflon FT-3) in another. Second stage is a 46 into a Hammond 126C interstage (or Lundahl LL1660/18mA). I found I could leave out the cathode bypass on the cathode resistor so I have.
Finals are 300b in cathode bias and here the cathode resistor is bypassed with a 100uF polypropylene cap.
I'm not too sure what could improve this, and certainly open to suggestions! I could battery bias the grid of the 300b for instance. But I'm pretty happy. It's quite a big amp on a few chassis, but I felt the topology was worth it.
Andy
Last edited:
It is not possible to say what is best. If our constraint is a working amplifier (nothing smokes, working points of active parts are correct, etc..) than every of the above possibility gives a sound reproducing device.
What actually changes is the harmonic spectrum and the bandwith. A grid has a capacity which has to be charged. If that's not fast enough (lack of current) you will get distortion.
A single high-gm pentode can drive a big DHT without effort. Two staged triodes can do that, too. The distortion spectrum will be different.
Putting the tubes aside isn't really helpful. 6C33 for example isn't really an audio tube. look at the curves, it has huge "punch through" distortion. (plate resistance varies greatly) when compared to a 300B/845 etc.
This distortion makes the sound and thats it.
So the question shouldn't be "which design is best ?" instead one should ask "which design gives less distortion ? / do you like most ?"
The truth is that the distortion figures for 6c33 SETs are way pass 5% from the first watts. So i need to reduce distortions VIA smart design.
The amp will have some form of feedback to bring down distortions and damping factor for NON-set-friendly music. But it will have a swich also
.
@Andy. can you post the schematic please?
The amp will have some form of feedback to bring down distortions and damping factor for NON-set-friendly music. But it will have a swich also
.@Andy. can you post the schematic please?
Here's the schematic for the first stage, which I have in another chassis as a premp. It will end up being direct coupled and in the same chassis as the amp - I'm still finishing this off and doing the filament supplies.
Haven't written out the amp schematic yet. Basically it's a 46 into a Hammond 126C, unbypassed cathode resistor, and the 300b is cathode bias bypassed with a 100uF polypropylene.
I'd stick to the 300b, and personally I would only use DHTs. Check out the thread "26 pre amp" for a lot of happy users and plenty of design details. Filament supplies are a big build but worth it for the sound you get. I don't know exactly why you are committed to the 6C33 when so much useful work has been done on using the better sounding 300b if you dig up all the best posts on operating points, distortion etc.
Andy
Haven't written out the amp schematic yet. Basically it's a 46 into a Hammond 126C, unbypassed cathode resistor, and the 300b is cathode bias bypassed with a 100uF polypropylene.
I'd stick to the 300b, and personally I would only use DHTs. Check out the thread "26 pre amp" for a lot of happy users and plenty of design details. Filament supplies are a big build but worth it for the sound you get. I don't know exactly why you are committed to the 6C33 when so much useful work has been done on using the better sounding 300b if you dig up all the best posts on operating points, distortion etc.
Andy
Attachments
Ra 600 ohms is a possibility for primary impedance; the tube itself has some 100 ohms of Rp.
The value of Ra is optional.
I don't see why the OPT can be cheaper than...what?
The 6C33 has more power (max. anode dissipation of 60 watts; in practice 40 watts is good) than for instance the 300B (max 42 - in practice some 30 watts).
Anode voltage for 6C33 is lower than 300B, but the 6C33 runs on much higher current, so the SE OPT needs a large airgap.
My custom wound 6C33 OPT's are larger than the 300B OPT's.
The value of Ra is optional.
I don't see why the OPT can be cheaper than...what?
The 6C33 has more power (max. anode dissipation of 60 watts; in practice 40 watts is good) than for instance the 300B (max 42 - in practice some 30 watts).
Anode voltage for 6C33 is lower than 300B, but the 6C33 runs on much higher current, so the SE OPT needs a large airgap.
My custom wound 6C33 OPT's are larger than the 300B OPT's.
Mine are huge. 3.5kg each. But they are an easy build considering the fact that you need fewer turn on the primary. You cand get a good bandwidth with not to much hassle wich is the reason they were so cheap.
About Ra. You are right. From what i know it can be anywhere from 300Ohms to 1.5K.
"Transparent" and "SET" don't belong in the same sentence. That isn't why people build them. Of course, you can always add gNFB to linearize any design, but then you lose the very reason why people want SETs in the first place.
Not too certain it makes as big a difference as you think. What's most important is designing well whatever topology you choose to go with. In cases like this, I would tend to go for keeping it simple. Design for linearity, and then add enough gNFB to improve what you have.
The A Number One consideration would be adequate grid drive. Power triodes tend to have some rather large Cmiller and input capacitance, and the grids have a nasty habit of drawing current even before Vgk actually goes positive. They also need a larger input swing. That means that you will need a decent driver for the final. A stiff cathode follower might be enough, and a source follower definitely something to consider.
Best I have heard that have a name
I can answer what are the "Best sounding SET amps I have heard", based on my limited experience and you can chase down the typologies and the builders. Otherwise, best could mean the best heavy object to slay your mother-in-law with.
I am only picking commercial builders for practical reasons. I don't think I have personally heard a DIY amp that I could give a name to.
In no particular order: Audio Note (UK) On-Gaku, or is it Gaku-On? (the power amp); Fi Audio 2A3, 45 and 300B mono-blocks, and various incarnations and re-incarnations of the Wavelength 300B Cardinal. There was also a commercial builder once upon a time who made some very, very nice sounding amps based on the Jack Elano DRD circuit, which you can find on the Electra-Print home page.
Most of these designs are somewhat, or a lot, different from each other, reflecting very different views about amplifier design, different experience and backgrounds -- at least between the US and UK, and different build approaches.
I can answer what are the "Best sounding SET amps I have heard", based on my limited experience and you can chase down the typologies and the builders. Otherwise, best could mean the best heavy object to slay your mother-in-law with.
I am only picking commercial builders for practical reasons. I don't think I have personally heard a DIY amp that I could give a name to.
In no particular order: Audio Note (UK) On-Gaku, or is it Gaku-On? (the power amp); Fi Audio 2A3, 45 and 300B mono-blocks, and various incarnations and re-incarnations of the Wavelength 300B Cardinal. There was also a commercial builder once upon a time who made some very, very nice sounding amps based on the Jack Elano DRD circuit, which you can find on the Electra-Print home page.
Most of these designs are somewhat, or a lot, different from each other, reflecting very different views about amplifier design, different experience and backgrounds -- at least between the US and UK, and different build approaches.
@ andy
Well.....6c33 is cheap, it is russian, has Ra=600Ohms(cheap OPT)...and i can get close to 20W from it
Domnul Rvrazvan indiferent ce topologie amplificatorului vreati involva cu tubul 6c33c-b eu va recomendez sokcetul asta Rusesk : PLK7 - collet ,din experienta mea este definitiv ce al mai bun !
: ''GSTube.com''. Tubes, sockets etc. Tube sockets of this type.
Salut si numai bine doresk !
Attachments
HAHAHA. NICE!
I have to buy sockets sometime in the near future.
How much to these ones cost? I think that the ones around here are chinese/ russians(brown with silver wires attached to them).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.