Interesting all pentode octal guitar amp - design analysis and mods

[guitarrock04]

I recently picked up a Montgomery Ward Standard Model guitar amp, 25-JDR-2484 at a local guitar show. It is a very neat looking and sounding amp that I picked up for a reasonable price. I've been unable to locate any information specific to this model, but this appears to have been built by Danelectro, and I've found a very similar schetic here:

http://danguitars.com/uploads/dano_special.pdf

Mine is the same, minus the tremolo circuit, and has a tone stack prior to V1, more like this example:

http://www.webphix.com/schematic%20heaven/www.schematicheaven.com/bargainbin/airline_wards_gdr-8513a.pdf

I am overall pleased with the tone of this amp, but would simply like more of it. It stays clean no matter what, I can only just get it to start to break up with a Les Paul and controls dimed.

I have a few questions, to make sure I understand the schematic, and am also open to any opportunities for improvement.

First and foremost, I interpret V2 as a triode-strapped cathodyne phase inverter. Why bother? Is there any technical benefit to this arrangement? Perhaps they found it more economical to use another 6SJ7 since it was already on the shelf?

V1 is grid leak biased, yeah? 15M is a sizeable resistor value! I'm not familiar with this bias strategy, and am trying to wrap my head around how this affects the gain and response of the stage. This arrangement still permits cathode-current feedback, correct? Perhaps I'm better off to update this with typical bypassed cathode bias for higher gain from V1.

V1 screen grid bias is a rather nifty means, pulling voltage from the cathodyne cathode bias resistor. I'm not sure I understand quite how this affects tone, though. Typical screen grid bias is take from the high tension supply through a dropping resistor. Assuming a similar nominal voltage, would the layout here provide a similar response? I'm inclined that the cathodyne bias method here provides some unique means of negative feedback.

Any thoughts, corrections, observations are greatly appreciated.

Thanks in advance

 

[PRR]

> It stays clean no matter what

That's all there is??

No wonder it won't dirt-up.

Idiot's Guide: you need TWO gain stages to thrash output bottle(s) with guitar.

This has ONE gain stage. (As you say, T2 is unity-gain; 6SJ7 was going out of style and may have been cheap per crate-full.) 

6V6 this way needs 20V peak grid swing. This 6SJ7 condition is way off the RC-amp table, but I'll guess gain is 300. 20V/300= 66mV peak or 50mV RMS. You "should" be able to touch clipping with a good whang. However a silver-face will clip at 20mV and modern grind-masters clip with less than 1mV.

There's a reason the first Fender Champ (one pentode gain-stage) was soon changed to a 12AX7 (two gain stage) design.

Put a simple booster, such as LPB, in front. It should be easier to overdrive without arm-strain.

Although first you ought to take some voltage readings. Resistors and caps were the cheapest stuff when new and are far from new now. Target values are unknown and probably not critical. But 6V6 cathodes should be 15V-20V, cathodyne cathode should be well above 20V (more like 40V; and the "15K" there is surely more like 1-point-5 K), both cathodyne 220K resistors should have equal drop, gain-stage plate should not be stuck way high or way low.

Grid-leak works because a healthy small tube's grid current through a huge resistor will run the grid negative. (Short grid to ground, plate voltage should drop; else the grid cap is leaking). This also "tends" to a maximum gain condition.

Pentode voltage gain is a lot about G2 voltage. Low G2 is high gain. But dropping-down from a high B+ is uncertain. Voltage-divider is good but an extra resistor (and this is a cheap amp). Finding a low voltage supply is a neat hack. Output stage cathode has been used. Cathodyne butt is less common but ought to work. Drawback: if G2 bypass cap is sick (open), NFB reduces gain. (And if short, zero G2 voltage kills gain.)