DC/transformer coupling

[Arthur Dent]

How feasible would it be to make an amp with mostly DC coupling, particularly in places where blocking distortion can be a significant issue? How could it be done, and what would be the pros and cons?

Also, I do not believe I have ever seen an amp that used interstage transformer coupling. Why is that?

[Chuck H]

Sure it's been done. Fender did it with the Bantam Bass and I think the 300/400 PS amps and Gibson did it on a couple of early models too.

The rub with the interstage coupling transformer in the PI position for guitar amps is threefold:

1) This is usually where the NFB is inserted on guitar amps and interstage coupling transformers create considerable phase shifts that can destabilize the circuit.
2) Interstage coupling transformers cause more of the bad kinds of distortion to occur than do tube stages.
3) Interstage coupling transformers do not overdrive and cannot be overdriven. So any attempt at generating additional good distortion by overdriving the PI becomes moot.

If used in other stages prior to the PI you still have issues 2 and 3 to deal with. That and in order to preserve any fidelity you need to use very high quality (expensive and bulky big) units.

With a little good design and tweaking you should be able to avoid blocking distortion.

I could see using an interstage coupling transformer for just the right application. But I haven't found it yet.

Chuck

[Steve Conner]

In the 300/400PS, Fender put NFB around the whole thing, including interstage and output transformers. That kind of scared me when I saw the schematic.

I have the interstage transformer from a 300PS, and it's pretty big and chunky. Much bigger than, say, a Champ OT. They probably had to make it go down to something like 5Hz, otherwise the phase shift in the two transformers would add up to 180 degrees and cause motorboating.

[Enzo]

I would also have to think a major reason for not using tramsformer coupling is cost. Transformers - even little ones - cost a hell of a lot more than tubes and caps and resistors.

Transformers can be used in place of a phase splitter tube stage, but what advantage would you predict for using a transformer between stages in the preamp? In other words, they don't use them because there is no reason to. In RF, they make tuned circuits convenient, but we don't use tuned circuits in guitar amps. RF transformer are tiny little light weight things too.

You could direct couple stages in an amp - a common example is a gain stage triode plate directly driving the grid of its companion cathode follower stage. But you can't put very many stages in a row before you run out of volts. A plate has to be at least a certain voltage to work well, which means the following grid starts that high too, and then the plate of that tube has to be even higher than the first, and so on. If each tube needs another 100-200 volts above the previous one, it adds up fast.

I suppose you could do some involved voltage division with resistors between stages, but that would also divide your signal, or so it seems to this old fellow.

[Rob Mercure]

To Chuck H,

Actually transformers produce desireable distortion artifacts, IMHO, folks have tried for years to emulate output stage distortion even to the extreme of adding in a smaller transformer stage that distorts used to push a higher powered output stage to allow output stage/transformer voicing at lower volumes. And there really isn't any reason that you can't overdrive an interstage or PI tranny - just push more current though until it saturates.
A while back we discussed SS amps and there was some consensus that some of the best sounding SS amps were those that used a transformer either at an output device or a driver.

Lastly one of the truly inspired bits of electronic engineering was the MacInstosh 50W-2 amplifier designed in the late 1940s which managed to get 50 W RMS out of a pair of 6L6s (not GBs or GCs) using both interstage (PP driven by 6J5s) transformer and an output tranny whose interwoven windings coupled the screen of one side to the plate of the other. The 50W-2 had truly impressive distortion figures for the day and was as ugly a bit of industrial engineering as you could want <grin> - I had a pair of these but sold them when I decided to live on the side of a mountain without running water and electricity for three years ('nother tale for another time) but now I wish I'd kept them. Look up the schemo if you want to seem the "uber result" of transformer coupling.

Rob

[Steve Conner]

Well, I kind of agree. Overdriving interstage transformers is going to create softer, lower-order harmonics than overdriving the usual RC coupled PI and power tube arrangement. That would be good for a bass amp, but maybe not for screaming lead guitar where you actually want aggressive high-end content.

Motivated by those discussions Rob mentioned, I've been experimenting with a hybrid amp that uses a pair of power transistors driven by an EL84 through an interstage transformer. After staring at RG's Vox Beatle schematics for hours, I figured out a circuit that lets me use a Champ OT, rather than getting a custom transformer made. I've got 50W out of it so far, and it sounds surprisingly good when overdriven.

Rob, I wish you'd kept your Mcintoshes too, or better still sold them to me I'm currently pestering a friend to sell me a busted Crown SXA270 as a restoration project.

Oh, and to try to get back on topic, check out the Loftin-White amplifier circuit.

[Rob Mercure]

Steve,

The sad part is that I sold the Macs to a good friend - a mad chemist and my "chemistry expert" when I need such - who really has too many interests and hobbies and such. He "stored" the amps in a semi-damp crawl space and one of the innerstage trannies shorted out. Around 1991-92 he let me rebuild the amps - as much as I could - which resulted in recapping, etc., the functioning one and attempting to unpot the defective one (these amps hall all of the trannies in a potted metal case with the tubes on the top). While I didn't further damage the defective one I realized after thoroughly covering myself with asphalt that I wasn't going to be able to get at the bad tranny with the means I have available. So he's got one amp that works well (getting close to time to recap again) and the other good PS chassis and bad amp chassis. If you were to offer him some token amount and pay the shipping to Scotland I suspect that he'd let you have it if you want to continue where I left off.
Originally these amps were used to drive the cutter head to make aluminum master records for a small recording studio in Bristol, Tennessee ("the Home of Country Music") during the days when musicians recorded direct to disk before tape masters.
Why don't you post the schemo for the amp you mentioned?
Back to the thread we've semi-hijacked: The Peavey "Special" Solo Series amp used an autotransformer to couple the output transistors to the speaker and produced a really nice sound. Had on come in years ago that got hit by lightning and was "totalled" as far as the customer's pocket went and he sold it to me for $10 - and I've been trying to think of a project for that OT ever since. Any ideas?

Rob

[loudthud]

Interstage transformers are favored in class AB2 power amps because the driving impedance to the output tubes can be kept very low. This keeps the distortion low and allows a lot more power to be produced from a given pair of tubes. The Fender PS series and the Mac are good examples.

[Arthur Dent]

Interstage transformers can drive AB2? That's sweet. I suppose that means blocking distortion won't be an issue, no matter how hard you drive the amp.

The main reason I started this thread was that, in my most recent build which uses a lot of pentodes and even a heptode, I cannot seem to quite get rid of significant blocking distortion in the preamp at any high gain levels, even though I'm using 100k grid reference resistors and 470p coupling caps most places. Bass response is already being effected negatively by cap size, and blocking is still a problem. I'll probably try Direct-coupling the PI and maybe something else, we'l see if that works.

Thanks for the help guys.

[defaced]

There's a thread in the archives that's dedicated to ways of running in AB2 and interstage transformers are discussed. It may be worth a look.

[Chuck H]

Actually transformers produce desireable distortion artifacts, IMHO, folks have tried for years to emulate output stage distortion even to the extreme of adding in a smaller transformer stage that distorts used to push a higher powered output stage to allow output stage/transformer voicing at lower volumes. And there really isn't any reason that you can't overdrive an interstage or PI tranny - just push more current though until it saturates. A while back we discussed SS amps and there was some consensus that some of the best sounding SS amps were those that used a transformer either at an output device or a driver.

I agree with this argument. But I also agree with Steve about the desirable harmonic content for the preamp. IMHO (and based on notion, not experimentation and testing) it could be that the distortion characteristics we like about transformer driven output stages has more to do with the way the tube is responding to the lower impedance circuit, rather than the artifacts from the transformer itself. AFAIK the transformer itself is offering nothing appreciable and is only an unavaoidable means of prefered tube operation.

Chuck

[Arthur Dent]

I suppose you could do some involved voltage division with resistors between stages, but that would also divide your signal, or so it seems to this old fellow.

I suppose it would work to use a 100k+ resistor in series with the signal, followed by a several-Henry choke to ground as the voltage divider. That would mean very little signal loss and a DC voltage close to ground.
Of course, you would then have the downsides of a large resistor in series with the signal, as well as having to use a choke.

Another way to do direct coupling would be to use a fairly large plate resistor on one stage, then set up the following stage a bit like a cathodyne splitter, but only take signal off of the plate and bypass the cathode resistor. Basically, use a massive cathode resistor and smaller-than-usual plate resistor such that the cathode is at a DC voltage slightly higher than the plate of the previous tube. See the SE 300B project on the How-To on the Angela Instruments site.

Some interstage transformers seem to me to be fairly reasonably priced. Most are less than a NOS 12ax7, anyway.

[R.G.]

Part of the MOSFET Heresies line of thinking was using MOSFET source followers to drive the grids of power tubes. This can drive them into AB2 DC coupled, eliminating blocking distortion completely. You need some modest power supply gymnastics to do this in an elegant manner, but it's not terribly complicated.