Transformer coupled electrostatic amps

[deepak]

Very cool! The amp looks great

[dsavitsk]

I need to put a kudo in here for Jack Elliano of Electra-Print. For 15 years he's built transformers for me. Almost everything I've asked for has been a non-standard, custom design. He's never let me down no matter how crazy my requests, and his prices are dirt cheap considering the custom engineering involved.

I'd just like to second this. Jack has made some tremendous things for me, now including a 32:32+32 OPT for use in a version of my solid state amp which is a real oddball.

I wonder if a choke phase splitter would work here? Something different.

[Voltron]

It doesn't need to be a World Premier, Frank, I just want you to bring it with you to Mayberry.

[FrankCooter]

It doesn't need to be a World Premier, Frank, I just want you to bring it with you to Mayberry.

I'll bring this amp, but you're right about the "Premier" thing. A true "Mayberry Premier" would require something much more indulgent, decadent, and "over the top". Feb. is still a long way off. I have some ideas. How big are your electrical mains?

[boomana]

I have some ideas. How big are your electrical mains?

I love Frank.

[Voltron]

I'll bring this amp, but you're right about the "Premier" thing. A true "Mayberry Premier" would require something much more indulgent, decadent, and "over the top". Feb. is still a long way off. I have some ideas. How big are your electrical mains?

do I check by climbing up the power pole and touch some wires or something?

[Dreadhead]

do I check by climbing up the power pole and touch some wires or something?

Main circuit breaker trip value will do the trick

[Pars]

^ And no Voltrons harmed in the process

[Voltron]

Main circuit breaker trip value will do the trick

^ And no Voltrons harmed in the process

Oh, that sounds much easier, and safer.

[spritzer]

Sounds like I'll have to make that Mayberry meet... Just checked the direct flights from Iceland and holy fuck they are expensive.

So how do you like the 7403's Frank? I have some 3d21a's too plus 807's and some other tubes Marc made me buy...

[Voltron]

Sounds like I'll have to make that Mayberry meet... Just checked the direct flights from Iceland and holy fuck they are expensive.

So how do you like the 7403's Frank? I have some 3d21a's too plus 807's and some other tubes Marc made me buy...

Of course you have to be there.

[n_maher]

Maybe team 307A will make the trip.

[Voltron]

That's hawt!

[spritzer]

Of course you have to be there.

[FrankCooter]

Playing with my new camera again. Here's a rough schematic of the amp. As you can see, it's brute simple. Plenty of opportunities to try different bias schemes, power regulation, and perhaps parafeed. Go ahead, rip it apart! I'll put up the power supply in the next post if this turns out OK.

[Dusty Chalk]

Turned out just fine.

[FrankCooter]

Now for the power supply. Here is where I expose myself for the hacker I am. It might have been more elegant to pull the 180v off the power transformer center tap, but I didn't have room for a second filter network. The cascaded 6ea7 regulators each drop 200v at about 40ma. Wastes some power, but who cares. I'm open to alternatives, as long as they're tube based.

[kevin gilmore]

Be careful frank, people will pick on you for your chicken scratchings...

They picked on me for years till i started doing pdf's of everything...

At least your chicken scratchings are far neater than mine were.

The bias resistor really does need to be at least 5 meg, and 600 volts

direct is a bit high.

[spritzer]

Certainly better drawn then anything I've ever posted.

I second a 5M resistor and a voltage divider to shave off a bit of the bias voltage.

[FrankCooter]

Since noboby shit on my schematics, I'll assume no major flaws, or at least a passive indifference.

The original prototype has been running without issue for about a year now. Gain and voltage swing are both about 1k. Beyond the basics, since I had ad-hoc output transformers, I haven't bothered with lab measurements. But like an any tube based, single ended, zero feedback design, lab measurements aren't what it's all about. You either like what you hear or don't. I think it sounds pretty good. Holds it's own against an SRD-7 and a solid state exStatA. Hopefully, I'll get a chance in the not too distant future to compare it to some of the "big boys".

One thing this amp isn't is cheap. I'm fortunate to have a big parts "junkbox". The only major expense was the output transformers, which ran about $260.00 with shipping. If everything was purchased new, the total parts content would have run near $1k. You could knock $200.00 off that by substituting a resistor or CCS for the d3A interstage transformers. At this amount of money, it needs to be a serious contender not to be a failure.

The amp in the photo will be finished shortly. Once it's stable and I'm familiar with the sound, I'll start trying to "improve" the design. The basic topology and tube compliment will stay the same (except for a possible DHT output). I've already experimented with most of the likely tube candidates. I tried the 12gn7, 417a, 6c45pi, 7788, 418a, 8233, as well as the d3A for drivers. My favorite was the Western Electric 418A, but it just didn't have enough gain. The d3A came in second. The 7788 was also good. Not a fan of the 6c45pi. All pentodes were run as triodes.

From the beginning, I assumed the output tube would be either an EL34 or a KT88. Not a lot of candidates that can take a 600v plate voltage, have decent gain, and are suitable for transformer loading. Tried the 6BX7 and the 6CK4, didn't like them. Settled on the KT88 over the EL34 because of the bigger power dissapation. I also am fortunate enough to own some original Genelex and Tung Sol. Somehow I missed the 7403/3D21. I'm using those now and liking them. Some people would be bothered by the plate caps, but I'm not. Spritzer asked about 807's. The 807 probably wouldn't be suitable due to it's relatively low screen rating and high rp.

I'm anxious to try a DHT in the output position. The ability to do this may be the biggest single advantage of this type of amp. Originally, I'd hoped to use an 801 (an uprated 10y) for the output position. The 801 has a gain of 8, can take 600v on the plate, and has a managable 1.25A filament. Unfortunately, it's 4.5k rp precludes precludes transformer operation in this application. ( a 9k primary in a 1:1-1 configuration makes a total 36k secondary. Not buidable in the real world without a huge high frequency roll off.) The 300B is a possibility, but an 845, run at a relatively low 600v looks good on paper. I'm a big 845 fan anyway.

The second major area of experimentation is the power supply. I'll substitute 816 mercury vapor rectifiers for the 6AU4 damper diodes. I also want to try shunt regulating the output tubes with mirror imaged power tubes.

Any other ideas or suggestions would be appreciated!

[Voltron]

I don't know what you are talking about, Frank, but I want to hear that thing! I will have a BHse built by Nate pretty soon and eventually the DIYT2 of his construction as well. We can definitely pit the bhse against yours the next time you are up here.

[dsavitsk]

Since noboby shit on my schematics, I'll assume no major flaws, or at least a passive indifference.

It looks like more or less the right way to do it. If it were me, I'd probably ultrapath the output tube with a CCS on the cathode. I might also parafeed the first stage. But, that's just my personal preference.

My favorite was the Western Electric 418A, but it just didn't have enough gain.

What about keeping the tube and using a 1:2 IPT, or a 5K:10K interstage?

[kevin gilmore]

I'm going to add a second opinion that the 3D21 is a good tube for this purpose.

[FrankCooter]

As usual, Doug has some interesting suggestions.

I've never tried a CCS/Ultrapath configuration for a cathode circuit. I can see where it might be a signifigant improvement over a simple resistor/bypass cap. Will definitely try this.

Not sure what the added complexity of parafeeding the d3A would accomplish. The additional plate choke or CCS plus capacitor would seem to make the interstage transformer redundant. Maybe I'm missing something here.

Changing the interstage transformer to accomodate a 418A probably isn't practical. I've asked Jack Elliano about high level step up transformers before and he's always been very negative. In the practical world, higher output impedance usually results in restricting the bandwidth.

One of Doug's ideas I'd like to try is his shunt regulated Shining Path/ Electric Avenue circuit on the output tubes. The late John Camille had an article about something similar years ago in "Sound Practices" magazine. He used a 211 to shunt regulate another 211 in the output position of an SET amp. I'd like to try something similar here.

Audiocat kindly sent me a pair of his custom Stax jacks. They look great and they fit much better than the standard Amphenols. There's a thread over in the DIY section in the "other forum". At least on this front, I'll no longer be an embarassment to the big guys.

Probably eventually substitute an 845 for the 3D21/7403. As many of you know, I'm a big fan of the Russian GM70 transmitting triode. Unfortunately, in order to keep the bandwidth up , the Electra-Print output transformers are a bit low in the primary for a GM70, 211, or 813. 845 should do fine however.

I'll never be able to afford any of the originals, but the cheap Chinese are actually pretty good. If any of you have any experiance with any of the premium types, I'd appreciate hearing about it.

[dsavitsk]

Not sure what the added complexity of parafeeding the d3A would accomplish. The additional plate choke or CCS plus capacitor would seem to make the interstage transformer redundant. Maybe I'm missing something here.

I guess I am not sure the benefits of the interstage at all -- there's no "coupling" cap, but you have a big PS cap in the signal current loop instead. Further, with single feed, the transformer has to be a large steel core with a big gap which seems far from ideal. Single feed only makes sense to me when there is a need to swing higher than B+ which doesn't seem to be the case here. Parafeed lets you use interleaved Permalloy, and a smaller core.

Only real advantages of the transformer over cap coupling are probably that you can use the transformer for current or voltage gain, and there is isolation from the grid of the output tube if it goes A2. And, parafeed seems better for boh of those? As I say, it's just a preference -- my ear likes nickel.

One of Doug's ideas I'd like to try is his shunt regulated Shining Path/ Electric Avenue circuit on the output tubes.

I think the best way to do this is likely to use the voltage regulator tube as a reference for a mosfet shunting element. The Zout from the vreg tubes are just too high (~300R) and to my ear led to constricted dynamics. In my amp, it was OK for low level stuff, but it seemed to fall apart with big swings.