looking for very good SE output transformer

[Alan0354]

Well, there are a lot of popular things that I'd argue aren't much good. I could name cars, bands, equipment, whatever.

The other thing to consider here is that numbers don't always tell you everything. I'm a fan of measurements, its the only truly objective way we have to compare stuff, but the fact is you can have a very pleasant sounding amp that measures pretty badly. A lot of people love Magnavox console amps, but in stock form quite frankly they're not awesome on paper. Conversely for years I hated the sound of my Phase Linear amp. Measures great, meets all its specs, but it made me want to plug my ears. I changed some parts and its much better now, still measures about the same but the sound is not the same. I can't quantify why its better, but it is.

Also, just because its a simple design doesn't mean there can't be "gotcha" elements to it. Layout, component quality, lots of other little things go into it. I've had far more problems getting basic 5 tube AM radios to work right than I have with multi-tube complicated shortwave receivers. The simple stuff sometimes has to be precisely just so in order to work, where the complicated stuff can be somewhat immune to less than perfect component placement and such.

The ultimate test is the listening test, no doubt about it. But with 4 amp (this cheap amp, Acurus, PA-7 Stasis and two of my amps) tested in detail both FFT and listening test , also with 3 pairs of speakers ( JM LAB, KEF and Monitor Audio). I do start to see a pattern. To be honest, all amps sound very similar, particular the SS amps. I truly can say the sound signatures are only slightly different. It's not like one make the human voice a lot better, or this better or base better etc. The tube kit amp is definitely a little more mellow and the high is not as pronounced. But that's mainly because it roll off 3dB at 20KHz where all the other SS amp are flat to 20KHz.

The major difference is the sound stage, the 3D effect, the fullness and separation. There is definitely a correlation between THD and how good these are. I have two members here at the beginning of this thread think I gone to twilight zoo when I start talking about this, and that the speaker cable is the key to make an amp stand out ( or bury the amp). But I am sure you've seen my threads on these step by step experiment and demonstrate this.

This is what I found from testing all the amps I have, the lower the THD, the better the soundstage, separation, 3D, fullness. Even before I added the cathode follower, the sound of the kit amp just stay on a flat wall in front of me rather than coming out and come alive. It's the cathode follower that saved the amp from going to the trash. That cathode follower lower the THD by half and the sound just starting to come out of the wall.

I want to stress, all my amps and speakers are NOT low quality, even the cheapest Acurus is nothing to sneeze at compare to most of the Onkyo, Pioneer, Sansui, Kenwood, Marantz and the like, definitely a class higher. My bad pair of speaker is KEF, it is by no means "lousy" speakers. I still paid over $600 at the time. I put a lot of effort in doing this.

I am confident the layout of the tube amp is as good as it can, I spent a lot of time thinking about the layout wiring before I even start soldering. I turn the tubes and point to different directions to make the wiring smooth and avoid crossover. I trim to minimum length to optimize the layout. So the poor quality of the kit amp is not due to layout. This is something I am very good at.

Too bad we all live far away, all we can do is talk. One listen speaks a thousand words, but we cannot.

 

[gadget73]

Too bad we all live far away, all we can do is talk. One listen speaks a thousand words, but we cannot.

now thats something I can completely agree with.

 

[Alan0354]

That's the problem, each room is different, so many variables in the system, the speakers, the cables, the amp and all. It is so hard to have an objective comparison. I cannot have a pair of high efficient speaker just to proof a point. I only have what I have and I have play within that "rules".

 

[for_p1]

That's the problem, each room is different, so many variables in the system, the speakers, the cables, the amp and all. It is so hard to have an objective comparison. I cannot have a pair of high efficient speaker just to proof a point. I only have what I have and I have play within that "rules".

If you want to make SE amplifier able to drive common speakers, you have only two types of tubes to work with: 845/211 (high voltage, high impedance) or 6C33C (low voltage, low impedance). Both can give you 15-30W (depends on circuit) with relatively (<2%) low distortion. Both types are hard to cook though.

And room should be treated enough to keep sound from speakers clean. That allows proper evaluation of other components.

 

[Alan0354]

If you want to make SE amplifier able to drive common speakers, you have only two types of tubes to work with: 845/211 (high voltage, high impedance) or 6C33C (low voltage, low impedance). Both can give you 15-30W (depends on circuit) with relatively (<2%) low distortion. Both types are hard to cook though.

And room should be treated enough to keep sound from speakers clean. That allows proper evaluation of other components.

Is there anything wrong having two tubes in parallel?

 

[for_p1]

Is there anything wrong having two tubes in parallel?

Need close matching. Overall distortion will be higher, which is a big deal when little or no feedback used.

 

[Alan0354]

Need close matching. Overall distortion will be higher, which is a big deal when little or no feedback used.

Those other tubes are a whole lot more expensive, I rather buy more of the EL34 and do matching.

 

[for_p1]

Those other tubes are a whole lot more expensive, I rather buy more of the EL34 and do matching.

You may try GM70 if you do not mind wasting 60W just for filament heating. But need to find the way to suppress hum and provide soft start for 20 volts heater.

You cannot get more simple than this http://www.jacmusic.com/kits/amp/gm70/GM70-index.htm

 

[Alan0354]

You may try GM70 if you do not mind wasting 60W just for filament heating. But need to find the way to suppress hum and provide soft start for 20 volts heater.

You cannot get more simple than this http://www.jacmusic.com/kits/amp/gm70/GM70-index.htm

Thanks, I downloaded the article and schematics to study.

 

[Alan0354]

I was on a cruise and I read a few articles.

http://www.lundahl.se/wp-content/uploads/datasheets/cb-amplifier_8wse.pdf

https://www.scribd.com/document/244029905/Single-Ended-vs-Push-Pull-Parts-1-2-Eddie-Vaughn#
The first article talked about lower the output impedance by feeding back from secondary to the cathode of the power tube only, this avoids the global NFB. The second article talked about output transformer of SE vs PP, that SE definitely has an advantage as it avoid hysteresis of the core.

 

[RGA]

Well as an Audio Note preamp owner and eventually amplifier owner I would recommend having a look at Audio Note transformers as they sell to several other very high end tube amp makers.

https://www.hificollective.co.uk/components/audionote_transformers.html

And if you are into KITs these are probably the best kits for Single Ended amps available (maybe Push Pull too) but they are also NOT cheap! http://www.ankaudiokits.com/Interstage-Quad-C-Core-300B-Power-Tube-Amplifier.html

 

[Alan0354]

Well as an Audio Note preamp owner and eventually amplifier owner I would recommend having a look at Audio Note transformers as they sell to several other very high end tube amp makers.

https://www.hificollective.co.uk/components/audionote_transformers.html

And if you are into KITs these are probably the best kits for Single Ended amps available (maybe Push Pull too) but they are also NOT cheap! http://www.ankaudiokits.com/Interstage-Quad-C-Core-300B-Power-Tube-Amplifier.html

Thanks, I'll put it as one of the choices. It's a little too expensive for me as I want to stay within like $300 for each output transformer. OT like Lindahl and Sowter. My SS amp is not the best in the world, I have a Nakamichi Stasis PA-7 designed by Nelson Pass that is a little better than his Threshold S300 from the design and circuit. My amps are clearly better but by no means the top end audiophile quality. My cost of material for each of my amp is about $900, so I want to stay within like $1000 for this amp to make it comparable quality to give this tube amp a fair shake before I make my conclusion between tube and SS amp. The Nakamichi was $1800 in the late 80s, I would imagine the comparable quality amp today is about $3000 to $3500. My amp being better, so I think I am shooting for a $5000 to $6000 type of amps.

I usually don't do kit amp, this is just cheap and dirty first try. I am studying up different circuit and plan to design my own. Been building amps for a long time, many tube guitar amps, also successfully designed and build two SS audiophile amps already. I figure $250 for a first try is cheap and I already learn a lot from it already. I am designing a new frontend of a SS input stage and take a dap in a hybrid input stage with tube as first stage right now. I'll get back into designing the SE amp as soon as I send the pcb out for fab. I am planning to try the partial feedback like what Claus Byrith did on partial feedback from the secondary to the cathode of the power tube to lower the output impedance. Problem with the tube amps is the output impedance is very high and cannot control the speaker that well.

 

[Selmerdave]

I was on a cruise and I read a few articles.

http://www.lundahl.se/wp-content/uploads/datasheets/cb-amplifier_8wse.pdf

https://www.scribd.com/document/244029905/Single-Ended-vs-Push-Pull-Parts-1-2-Eddie-Vaughn#
The first article talked about lower the output impedance by feeding back from secondary to the cathode of the power tube only, this avoids the global NFB. The second article talked about output transformer of SE vs PP, that SE definitely has an advantage as it avoid hysteresis of the core.

I wonder what Eddie Vaughn recommended for speakers, while you're taking his advice...

 

[Alan0354]

One comment I want to make from the few days I studied tube circuit and from building the cheap amp. Really seems like tube amp mostly rely on the quality of the output transformer and the tubes, circuit design is secondary. The circuit is so damn simple there is not much variation. My experience with SS design is the design is very important, that you can get much better sound quality by design rather than paying boats load of money for the tubes and output transformer. Only thing one cannot go cheap in SS amp is the chassis. Chassis with huge heat sink is needed for high quality SS amp. The chassis alone is like $350 in my amp. That's only good for lower power high bias push pull amp only. If you want to run class A or higher bias, the chassis is like $500 to $600. Problem also the heat generated by the amp can be very bad. People claim tube amp runs hotting should really stop and try some higher end SS amp like Krell.

I almost bought the Krell KSA250 before the PA-7 Stasis, it was a good price of about $1800 on ebay. But I decided against it because it burns 1200W of heat just sitting idle. Unless you live in a very cold country, this is really bad. My bigger amp runs about 180W of heat and my smaller amp about 140W. I had to adjust it down to from my original design as the big amp heated up the whole living room when I had over 200W of heat from the big amp. It's funny the tube amp kit ran cooler than even my small amp. No even close.

 

[Alan0354]

I wonder what Eddie Vaughn recommended for speakers, while you're taking his advice...

No idea, these are all new to me. I just look at the design, there are some comments I don't even quite agree with him looking from the SS world.

Only thing I know is I absolutely do not want horn tweeter and maybe big woofer. I am old enough to live through those horns and big woofers. When I bought my first pair of German Heco in the early 70s, that was out of this world compare to all the horns. I love the sound of the dome tweeter. I do find the smaller woofers give better attack and transient.

On top of it all, there might be extreme high end high efficient speakers using horns, but if you look at those beat up altec on ebay, they look really bad and cheap. Those look like from the 50s or 60s? Some still want boat load of money. Far as I concern, electronics gone through quantum leaks since the 50s and 60s. I don't want to look back.

 

[primosounds]

One comment I want to make from the few days I studied tube circuit and from building the cheap amp. Really seems like tube amp mostly rely on the quality of the output transformer and the tubes, circuit design is secondary. The circuit is so damn simple there is not much variation.
.

You are looking at this the wrong way.
Because of the apparent simplicity of a single ended amp each component used becomes an important part of the signal chain. There is no noise cancellation from signal splitting in the phase splitter and recombination in the output transformer. And at the same time the signal has not been altered. In this way the signal at the input has far fewer connections, tubes and circuits to go through before exiting the amplifier. You heard all the phrases, KISS, keep it simple stupid. or , less is more. All those audiophiles out there with single ended amps and high efficiency speakers must be hearing something that is good to their ears.

 

[Alan0354]

You are looking at this the wrong way.
Because of the apparent simplicity of a single ended amp each component used becomes an important part of the signal chain. There is no noise cancellation from signal splitting in the phase splitter and recombination in the output transformer. And at the same time the signal has not been altered. In this way the signal at the input has far fewer connections, tubes and circuits to go through before exiting the amplifier. You heard all the phrases, KISS, keep it simple stupid. or , less is more. All those audiophiles out there with single ended amps and high efficiency speakers must be hearing something that is good to their ears.

I think I look at it exactly the right way. The circuit is so simple that the quality is ONLY governed by the quality of the output transformer and the tube.

The thing that really bugs me is the output transformer is the biggest bottle neck with all the limitations. You have parasitic, flux leakage. You be lucky to get -3dB at or over 100KHz. You introduce a lot of phase shift in the signal chain. The equivalent circuit is not as simple as you think.

I can appreciate less is more, but the imperfection of the OT by itself is a lot more already.

As I said, I am not going to make my decision until I build a reasonable amp using Lindahl or Sowter transformer ( about $300 ea). I have to hear with my own ear to judge. But as I stress, horn and altec is absolutely out of the question as I hate the sound of horn. I cannot stress enough, this is 2017, not 50s. Life moves on.

BUT, if all the experience people here think there is no way for a SE amp to sound good unless it's driving a horn high efficiency speaker, then my venture into tube amp ends right here. I have a pretty good pair of speaker that the retail price was over $5000, I am not going to go spend another $5K+ for a high efficient speaker that I likely hate.

As for audiophile people's opinions, I think it works both ways. I read enough comments on SS amp that people die hard on SS. So I won't use that for judgement one way or the other. I can tell you from experience with my own designs, the sound really pops out, 3D, detail separation of all little sound. The biggnest of the sound is like listening in the concert hall. Actually feel echoie.

 

[maxhifi]

Hi Alan,

Sorry it took so long to get back to you, I sat down to draw out the load line and realized that UL curves for the EL34 are totally unavailable at the screen voltage you're running. Tried to use the adjustment factors in the RDH4 page 38, but things didn't add up, and I realized you really just need curves traced at this operating point.

Would be so simple if not for that, but all I could find are the 450V curves. Basically what I was getting at, is for a tube to be linear, the spacing between the grid voltage lines needs to be even, i.e. for a given change in grid voltage, the plate current should change roughly equally, at different parts on the curve. For the EL34 in ultralinear, this doesn't happen.. below the bias point there's much less delta Ip for a corresponding delta Vg than below it - hence, an asymmetrical waveform is guaranteed.

The use of cathode feedback is also old news.. look into how the Quad II is designed, or the Pye Mozart as two vintage examples. It also needs special consideration in circuit design, including a low distortion high output driver stage, and an amazing output transformer.

I think you're right on though, saying an SE amp is fundamentally not compatible with your speakers. I believe that spending $1000 - $2000 on a custom made amp, which will lack the power and low output impedance to properly drive your speakers, is more or less an experiment doomed to failure.

I'm a tube guy, always have been, likely always will be, but given your speakers, I would rule out single ended amplification as impractical. A push-pull tube amplifier could be made to sound better than the old Conrad Johnson you didn't enjoy, but again, this also could turn out to be worse with your speakers than your existing amplifier is.

I think it's a high effort activity with a low payoff, to seek alternatives to what you've got already. If I had what you have, I would probably be looking into a big screen 4k television, some new artwork for the house, perhaps make your wife happy and redo a bathroom.. I think you've picked the low hanging fruits already, and it's hard to see the value in investing more in amplifiers

Best Regards

Max

 

[Alan0354]

This is a technical discussion. Please don't come and say it's X-File and out there.

I am thinking out loud, what if I buy a pair of Sowter SE for 2 EL34 in parallel. Just replace the OT of the cheap amp, I can modify to have two EL34 by drilling extra hole for the driver tubes. I already gone to SS rectifier to gain like 25V out of the power transformer. So even if I put extra pair of EL34, it might just drag the +B down to like with the tube rectifier. Hell, I don't care the power transformer glows, as long as it last through the listening session one time, I am good. I can still experiment the local NFB and other circuits.


One thing puzzled me, I know the power tube and output transformer is everything, why nobody use SS as driver stage? Tubes has a lot of distortion particular even harmonics, BUT without going into detail, the even harmonics of the driver tubes will become odd harmonics after the second tube( power tube). A good cheap opamp can drive a whole hell of a lot better that ANY driver tubes, almost no THD ( in tube standard). So opamp will drive the power tube a lot better, distortion will be only from the power tube/output transformer.

I know good audio opamp can drive almost +/-20V output and over 20mA driving capability. I am surprised nobody talk about using cathode follower to drive the power tube like what I did. If you study the power tube, it is very important to be able to drive when the grid of the power tube starts to conduct current. I got a lot more linearity out of adding the cathode follower in the kit amp. Opamp output will make the cathode follower like a wimp.

This is one opamp I am using http://www.ti.com/lit/ds/symlink/opa604.pdf
http://www.ti.com/lit/ds/symlink/opa604.pdf
It can work with +/-24V supply and drive almost to +/-21V. That's more than enough to drive EL34.

 

[maxhifi]

This is a technical discussion. Please don't come and say it's X-File and out there.

I am thinking out loud, what if I buy a pair of Sowter SE for 2 EL34 in parallel. Just replace the OT of the cheap amp, I can modify to have two EL34 by drilling extra hole for the driver tubes. I already gone to SS rectifier to gain like 25V out of the power transformer. So even if I put extra pair of EL34, it might just drag the +B down to like with the tube rectifier. Hell, I don't care the power transformer glows, as long as it last through the listening session one time, I am good. I can still experiment the local NFB and other circuits.


One thing puzzled me, I know the power tube and output transformer is everything, why nobody use SS as driver stage? Tubes has a lot of distortion particular even harmonics, BUT without going into detail, the even harmonics of the driver tubes will become odd harmonics after the second tube( power tube). A good cheap opamp can drive a whole hell of a lot better that ANY driver tubes, almost no THD ( in tube standard). So opamp will drive the power tube a lot better, distortion will be only from the power tube/output transformer.

I know good audio opamp can drive almost +/-20V output and over 20mA driving capability. I am surprised nobody talk about using cathode follower to drive the power tube like what I did. If you study the power tube, it is very important to be able to drive when the grid of the power tube starts to conduct current. I got a lot more linearity out of adding the cathode follower in the kit amp. Opamp output will make the cathode follower like a wimp.

This is one opamp I am using http://www.ti.com/lit/ds/symlink/opa604.pdf
It can work with +/-24V supply and drive almost to +/-21V. That's more than enough to drive EL34.

1. Using a cathode follower to drive output tubes into class A2 or AB2 is nothing new. Look at amplifiers such as the Heathkit W6M, or the McIntosh MC60, and any number of others use this approach. It was especially common with public address amplifiers, a pair of 807 can generate 75W this way! Thing is, a properly specified output stage doesn't need to go into grid current to generate maximum output, this is kind of a brute force approach, and is often said to sound worse at low levels, since it adds another stage, and cathode followers in general aren't popular sound wise, so is seen less often in new equipment.

2. SS driver driving tube output stage has been done many times, - why would you not do it? Same reason you wouldn't drive the speakers directly with the op-amp, it isn't a tube amp when you do it that way, and sounds different. Why does it sound different? good question, falls into the category of why use tubes at all. One thing I do know though is the driver tubes do play a factor in the good or bad sound of a tube amp.

If you do go that way though, try and use a higher voltage op-amp to get more headroom, to make sure it absolutely never clips under any conditions. 21V P-P is not enough for your amp, which is biased at -22V. (needs 22Vp-p just to make it to grid current, but would hard clip once you reach it) I wouldn't be happy with less than 30VP-P available in that case. Do some research about hybrid amplifiers, and MOSFET drivers, it's not a new idea, lots of reading material out there.

There's nothing X-files about saying we don't really know WHY some people like tube amps better than SS. Lots of theories around, but none really nail it with an objective and impossible to argue reason. If it were that easy to quantify, the problem would be solved long ago.