PA -> Champ-ish guitar amp

[Rhodesplyr]

A few months ago, I disassembled a ~1948 Thordarson T-31W08 I inherited from my grandfather in preparation for converting it to a guitar amplifier. Thordarson was a transformer manufacturer, and the output transformer is large for 8W of output; based on some of the documentation I've found, I believe the OPT was rated for 25W and also came in a center-tapped version for push-pull. In general, the transformers and choke look better made than the ones I've seen on Masco PA amps.

Though it was in working condition when last checked, this amp one of those things that had been worked on to the point that it was a bit of a mess, so I decided to do a fresh build. Then, of course, I started thinking about it--always the way to delay completion of a project. It's a very typical PA of the era with two pentode gain stages (6J7 and 6SJ7) and single 6L6 output. I've heard that the 6SJ7 Champ really didn't really have enough preamp gain, but I'm calculating that, with an extra gain stage, this one shouldn't have that problem. I had toyed with the idea of, perhaps, changing preamp tubes to one pentode and one dual triode (6SL7) for different overdrive characteristics, but I'm now thinking that I'll just rebuild the stock circuit and see how it sounds.

I will update it, of course, with better grounding, hum reduction, and little things like an AC power fuse. I'll increase power supply caps a bit within sane limits and may play with cathode bypass values. The output tube's 20uF cathode bypass is definitely a high-pass filter based on my math. The power switch will move from the tone control pot to the back.

I like the fact that it has 4/8/15 Ohm transformer taps, and I was thinking that I'd use a three way toggle switch on the back to select among them. The original octal output socket is not very useful, and I could mount the 3-way switch in its place on a plate.

Question for more experienced builders: is there anything on the schematic that sends up any red flags or that could easily be improved upon for use with guitar? As I understand it, the low screen voltages on the preamp pentodes limit plate current, but increase input sensitivity. I will admit that like many people, I understand triodes much better than I understand pentode operating points.

Thordarson T-31W08.pdf

[Steve Conner]

The original circuit seems like it would work reasonably well with guitar. It all depends on what sort of tone you want from it.

A big OPT in a single-ended amp is always a good thing.

The optimum screen voltage for a pentode is different for every tube type. Most people just use the datasheet recommended voltage.

One thing you might like to try, if you want a bit more gain, is connect the pentodes as triodes and then cascade the two stages, the way the two triodes are connected in a Champ.

If you cascaded them as pentodes, it might be too gainy.

[Rhodesplyr]

If you cascaded them as pentodes, it might be too gainy.

I was hoping the volume control between the stages would deal with that, but I was thinking along the same lines: wiring one of the two pentodes as a triode. As always, there are trade-offs both ways. A first stage pentode avoids the Miller capacitance of a triode, but you get more noise.


It's a simple enough amp--I might try both approaches and see what sounds best.

[Chuck H]

Keep in mind (whatever you decide to build with the preamp) that too much OD with a single ended amp just sounds bad. Unlike AB1 amps where all the power tube OD is saturation, with class A you have one end cutting off and the waveform can get VERY assymetrical (a little seems good, a lot not so much). Point is with a high gainer whatever you generate in the preamp will need to be made compatible with the drive requirements of a class A amp by attenuation. This can get a little tricky when you intent to use more than one channel and/or multiple gain options. It's a good idea to include channel specific attenuation (master volumes) for higher gain channels to help keep drive voltages under control. Otherwise the power tube could just squash and "duck" and otherwise cut out and perform poorly. I found this out for myself when rebuilding a Vibro Champ into a two channel thing for a friend. So I'm just passing it along.

Chuck

[Rhodesplyr]

Chuck,

I think I see what you mean. Doing a rough calculation, it seems to me that, based on the stock design, the input signal to the 6SJ7 would end up overdriving the 6L6 before it would overdrive the 6SJ7. Are we thinking along the same lines?

I wonder if gain control was the idea behind the small 6L6 cathode bypass capacitor in the original design. I've seen in Class AB1 push-pull amps, in which the output tubes share a single cathode resistor, that the size of the cathode bypass capacitor (if there is one) can be used to tune how the output stage distorts. A large cap extends clean power, but when you hit distortion, you get a nasty, blatty tone.

And this is unrelated, but I just noticed on the original Champ 5C1 schematic that there's no NFB and the 6SJ7 is grid-biased with a 5M resistor, so Thordarson's circuit is really less similar to the Champ than it seemed to me at first glance.

[Rhodesplyr]

Steve, I see what you mean by "too gainy." I've been doing some reading on pentode gain stages; fortunately, the 6SJ7 is often used as an example. One discussion uses basically the same values as in the original Thordarson circuit (470k plate resistor) and claims a gain of 164--neglecting the load, of course. Most popular 6SJ7 input stages for guitar use 220k or 100k plate resistor values.

It makes me wonder why Thordarson designed their circuit for such high gain. One of the inputs is marked "Phono," which, in 1948, would have been the relatively high output from a ceramic/crystal cartridge, and the other is for Microphone. I'm not sure what typical mic output levels of the period would have been.