1960's Supro Thunderbolt scratch build copy

[Chuck H]

I really don't think you should get nutty with the screen grid resistors. Something like 2.2k would be about as high as I ever see in a design. That wouldn't decrease dissipation enough for the correction you need.

Since you already have a cathode resistor the 1 ohm resistor is sort of redundant. You can use the cathode resistor to figure the shared current. Provided that resistor is on spec your shared current is 183mA!!! That's 91.5mA per tube!! If you subtract the cathode voltage from your plate voltage we get a working voltage of 406.4Vp. If we multiply the current (per tube) by the working voltage it shows your amp is dissipating 37 watts per tube!!! I suspect either your cathode resistor or your 1 ohm resistors spec low, but even if we use your figure of roughly 67mA per tube that's still 27 watts per tube. You need to get that down brother! I don't know what you have on hand but I think a 330 ohm cathode resistor would be about right. Or...

If you have another 200 ohm resistor you could put it on the HV winding center tap before ground. That would simulate a rectifier tube somewhat adding sag and softening the attack envelope. Way better idea than giant screen resistors. This will also drop the Vp to about 400. Which gives you a working voltage of about 365. In this case a 250 ohm cathode resistor should be better. What values of big resistors do you have on hand?

And you never answered my Q about the PT spec for that circuit.?. I'm in serious doubt that the original design spec'd a 200 ohm cathode resistor with over 440Vp.

[soundmasterg]

If you look at the original schematic that Sean Weatherford compiled with my help and others help, (S6420.pdf above) the B+ with the solid state rectified version is right around 425V. Your voltage is a bit higher, so the proper cathode resistor value should be a bit higher too. With the stock amp, a 200 ohm resistor puts the bias up around 80% to 90% of the max dissipation for the 6L6GC. Sovtek 5881's actually sound great in these amps because they like to be biased hot. In any case, with a stock amp and a 200 ohm cathode resistor on the power tubes, tube life is shorter than many fixed bias amps such as your typical fender. With a stock amp, changing to a 240 ohm cathode resistor cooled the tube bias down, but made the sound noticeably less sweet. You will probably have to experiment around to get to a good resistor value.

A couple other things....

- the originals used ceramic coupling caps throughout the amp. Using film caps is generally considered to be a better route to go, but that may get rid of some of the character of the original design.
- You will notice if you look at vintage amps that they either don't use a screen grid resistor, or use a 470 ohm, or a 1k, but almost never much larger than that. Read the attached document, specifically section 8 and they talk about optimum screen grid resistor sizes. The main thing to keep in mind is that you want some protection for the screens, but not so much that you start to lose power.
- You can take a look at a stock Supro Tbolt in the attached pics and the way they made the amp.

Greg

[mort]

If you look at the original schematic that Sean Weatherford compiled with my help and others help, (S6420.pdf above) the B+ with the solid state rectified version is right around 425V. Your voltage is a bit higher, so the proper cathode resistor value should be a bit higher too. With the stock amp, a 200 ohm resistor puts the bias up around 80% to 90% of the max dissipation for the 6L6GC. Sovtek 5881's actually sound great in these amps because they like to be biased hot. In any case, with a stock amp and a 200 ohm cathode resistor on the power tubes, tube life is shorter than many fixed bias amps such as your typical fender. With a stock amp, changing to a 240 ohm cathode resistor cooled the tube bias down, but made the sound noticeably less sweet. You will probably have to experiment around to get to a good resistor value.

A couple other things....

- the originals used ceramic coupling caps throughout the amp. Using film caps is generally considered to be a better route to go, but that may get rid of some of the character of the original design.
- You will notice if you look at vintage amps that they either don't use a screen grid resistor, or use a 470 ohm, or a 1k, but almost never much larger than that. Read the attached document, specifically section 8 and they talk about optimum screen grid resistor sizes. The main thing to keep in mind is that you want some protection for the screens, but not so much that you start to lose power.
- You can take a look at a stock Supro Tbolt in the attached pics and the way they made the amp.

Greg

Thanks for the info ! It's amazing that I would get some advice from one of the very people who's schematic I found and copied.

I'll play around with the cathode resistor value and see how it goes.
\

And Chuck I really appreciate your help too. I'll have time and parts to implement a couple more adjustments by the weekend and I will definitely post up the results.,


Thanks guys !!

[mort]

Ok I'm getting back around to working on this and was planning on upping the value of the cathode resistor to decrease plate current a bit and realized that I had actually installed a 270 ohm 10W, instead of a 200 ohm like my schematic showed. I had read a note somewhere about using anywhere from 200-330 and used a 270 but forgot to reflect that on my drawing.

The above comment about losing tone above 240 has me wondering what the amp would sound like if I dropped the value down to 200 or 220(sounds great now but maybe it could sound even better?). I expect lowering the value would increase plate current so now I need another strategy of decreasing current.

I guess one way would be to raise the grid resistor values but as mentioned already that may not be a great idea. ... how should I go about this?

[Chuck H]

I guess one way would be to raise the grid resistor values but as mentioned already that may not be a great idea. ... how should I go about this?

Bad idea. Don't do it. 47k is already high and one side has a 470k bias resistor, which is also high. If you look at the spec'd grid circuit impedance you'll see that 517k exceeds the maximum grid circuit resistance already. Besides, the current limiting wouldn't be enough to matter much unless the tube were trying to draw grid current. About the only way to keep the cathode resistor low and the tubes cool would be to get tubes that tend to run cooler (the net difference in tone should be about the same) or reduce plate voltage, which really complicates matters.

Incidentally, your mistake on the cathode resistor value explains how you and I got different bias calculations.

I wouldn't try decreasing the cathode resistor value if I were you. No matter what you read. That's running the tubes dangerously hot and any possible tonal benefit would be short lived at best. Momentary at worst.

[mort]

Remember we found the error on the 470k resistors [should have been 470 ohm] and I put in 1K. All of the last voltages are with the 1K in place and the 270 ohm cathode resistor.


Here's an updated drawing of what is actually in the amp.

[mort]

Ok I tried a 240 ohm 5W resistor just for a test. Here's what voltages I got.

V4 - 71mv across 1 ohm resistor
p3- 442v
p4- 435v
p5- .111
p8- 35.93


V3 - 68mV across 1 ohm resistor
p3- 442v
p4- 435v
p5- .163
p8- 35.95

I kinda like the tone better with the 270 ohm resistor in, so I'm going to switch back. But I have a question about wattage calculation. Chuck you had mentioned "Measure the cathode voltage and divide by the cathode resistance. This is your current. Multiply that number by your "plate to cathode" voltage to get your "watts at idle" figure"

How do I figure the cathode resistance? I entered in 240 ohms and came up with around 60w per tube. If I use the voltage across the 1 ohm resistor method for calculating current, then I get around 28W per tube. Where was my error?

[g1]

Your error was that you didn't divide the 60W by two. Both tubes current runs through the 240 ohm resistor.
So 36V divided by 240 ohm = .15A or 150mA. That is for 2 tubes, so current is about 75mA each tube (if they are matched).

[Chuck H]

Remember we found the error on the 470k resistors [should have been 470 ohm] and I put in 1K.

That was the screen grid resistors. I'm talking about the G1 (input grid) resistance. It's 470k with a 47k grid stop resistor. That's 517k on a circuit with a max value of 500k in an amp that will probably be cranked as often as not. Bad idea.

[soundmasterg]

Ok I'm getting back around to working on this and was planning on upping the value of the cathode resistor to decrease plate current a bit and realized that I had actually installed a 270 ohm 10W, instead of a 200 ohm like my schematic showed. I had read a note somewhere about using anywhere from 200-330 and used a 270 but forgot to reflect that on my drawing.

The above comment about losing tone above 240 has me wondering what the amp would sound like if I dropped the value down to 200 or 220(sounds great now but maybe it could sound even better?). I expect lowering the value would increase plate current so now I need another strategy of decreasing current.

I guess one way would be to raise the grid resistor values but as mentioned already that may not be a great idea. ... how should I go about this?

Remember that a stock amp using a 200 ohm cathode resistor to bias the power tubes is running a lower B+ than your amp and that changes the bias calculation. The stock amp is biased hot AB and probably somewhere around 80-90% of the max plate dissipation. It sounds great, and works fine...the only downside is that the power tubes wear out sooner than if it was biased cooler. A proper cathode resistor to get your amp in the same range can be calculated or you can sub in different types and figure out what the dissipation is, and see if you like the tone or not. The comment on the schematic regarding using a 240 ohm if desired was for people who didn't want their tubes to wear out so quickly. Another thing to remember is that since your amp is cathode biased, you need to remember to subtract the cathode voltage from the B+ before you calculate your dissipation.

Greg

[mort]

Your error was that you didn't divide the 60W by two. Both tubes current runs through the 240 ohm resistor.
So 36V divided by 240 ohm = .15A or 150mA. That is for 2 tubes, so current is about 75mA each tube (if they are matched).

ok yeah ... not sure why I didn't see that... sometimes I miss the obvious


OK so with the 270 ohm back in the circuit here are my numbers. And point taken about the higher B+ voltage vs the B+ in the pdf

V4 68mA(444 - 36.62) = 27.7W
p3- 444
p4- 437
p8- 36.62


V3 64mA(443 - 36.62) = 26.0W
p3- 443
p4- 437
p8- 36.62

So it looks like I'm running around 90% at idle... I may try a 300 ohm cathode resistor just to see if it still sounds good but according to what soundmasterg has mentioned it may lose something...


Oh hey one more question... would I raise or lower one input resistor if I wanted to make the inputs high/low ?? Lower one value to make it a high input??

[Chuck H]

would I raise or lower one input resistor if I wanted to make the inputs high/low ?? Lower one value to make it a high input??

That is more effectively done with switching jacks that use the grid resistors placement in the circuit to create different voltage divisions. Look at any Fender or Marshall schematic to learn how it's done.

[mikepukmel]

Hi Amp Gurus, This thread is getting a bit old in internet terms, awesome info on Mort's build. its exactly what I was hoping to do (once I get all the kitchen and bathroom repairs done, and the yard cleaned up and the bad tire replaced ... ).

In case Mort isn't monitoring this thread anymore, anyone have any idea where he got the eyelet board from? Is it a custom made eyelet board, or maybe one of those standard N hole boards with a few eyelets added?

Thanks!
Mike

[klimbo]

If you look at the original schematic that Sean Weatherford compiled with my help and others help, (S6420.pdf above) the B+ with the solid state rectified version is right around 425V. Your voltage is a bit higher, so the proper cathode resistor value should be a bit higher too. With the stock amp, a 200 ohm resistor puts the bias up around 80% to 90% of the max dissipation for the 6L6GC. Sovtek 5881's actually sound great in these amps because they like to be biased hot. In any case, with a stock amp and a 200 ohm cathode resistor on the power tubes, tube life is shorter than many fixed bias amps such as your typical fender. With a stock amp, changing to a 240 ohm cathode resistor cooled the tube bias down, but made the sound noticeably less sweet. Tutuapp 9apps ShowboxYou will probably have to experiment around to get to a good resistor value.

A couple other things....

- the originals used ceramic coupling caps throughout the amp. Using film caps is generally considered to be a better route to go, but that may get rid of some of the character of the original design.
- You will notice if you look at vintage amps that they either don't use a screen grid resistor, or use a 470 ohm, or a 1k, but almost never much larger than that. Read the attached document, specifically section 8 and they talk about optimum screen grid resistor sizes. The main thing to keep in mind is that you want some protection for the screens, but not so much that you start to lose power.
- You can take a look at a stock Supro Tbolt in the attached pics and the way they made the amp.

Greg

Thanks for this helpful informations

[soundmasterg]

Thanks for this helpful informations

You're welcome!

Greg