1960 Stromberg-Carlson Signet 33 Guitar conversion!
- Thread starter Natima
- Start date
Yes and no. C8 is there to compensate for Miller effect. Tubes have an amount of internal capacitance between plate and grid. That figure times the gain of the stage is the effective capacitance. Add a resistor in front of that and you have an RC low-pass filter. The bigger the resistor, the lower the roll-off point is and the less HF signal you get out of the stage. Bridging a small value cap across the resistor passes some extra high frequency signal to make up for the roll-off. Basically it flattens out frequency response, but if you want more or less HF response that part can be adjusted or eliminated.
The tone circuit values work in a roughly similar manner, its all RC filter networks. Without getting too deep into particulars (and because I'm not nearly awake enough to bust out the math to do it) in a tube circuit, a resistor ahead of a tube tends to pass lows but not highs. Caps pass highs but not lows. Changing those values controls how high or how low a signal will pass. Resistor between parts and a cap to ground is a low pass filter, cap between parts and a resistor to ground is a high pass. Link them in the right way and you get a bandpass or notch filter, depending if you want all the stuff between two frequencies or none of the stuff between two frequencies.
The tone circuit values work in a roughly similar manner, its all RC filter networks. Without getting too deep into particulars (and because I'm not nearly awake enough to bust out the math to do it) in a tube circuit, a resistor ahead of a tube tends to pass lows but not highs. Caps pass highs but not lows. Changing those values controls how high or how low a signal will pass. Resistor between parts and a cap to ground is a low pass filter, cap between parts and a resistor to ground is a high pass. Link them in the right way and you get a bandpass or notch filter, depending if you want all the stuff between two frequencies or none of the stuff between two frequencies.
yes, but you'd lose a fair bit of gain. Possibly you wouldn't have enough to drive the output tubes to full power. The one big amp I have with a similar configuration is a Rauland-Borg 70w PA amp running EL34 outputs. It has half of a 12ax7 ahead of the usual 12ax7 cathodyne setup to get enough drive to make it go.
Natima
New Member
So.... I've just completed my Nth iteration of modding this amp, and I've REALLY this time found the magic combination.
For anyone out there converting one of these amps in the future I highly recommend the below attached schematic. It really broaches the early Marshall meets Fender sound.
Removing R12 and C8 seems pretty key to this. Beautiful cleans, amazing compression, and crunchy drive. This was my *hopefully* last set of mods on this amp and it just came to life.
https://photos.app.goo.gl/gGHWFAbAnRsNgR3W6
Video TBA... hopefully tomorrow.
I'd really love to try it through a 4x10.
For anyone out there converting one of these amps in the future I highly recommend the below attached schematic. It really broaches the early Marshall meets Fender sound.
Removing R12 and C8 seems pretty key to this. Beautiful cleans, amazing compression, and crunchy drive. This was my *hopefully* last set of mods on this amp and it just came to life.
https://photos.app.goo.gl/gGHWFAbAnRsNgR3W6
Video TBA... hopefully tomorrow.
I'd really love to try it through a 4x10.
Attachments
Natima
New Member
More generic like what? I did tonnes of googling and about the only good result was this: https://robrobinette.com/5e3_Modifications.htm#Add_Master_Volume
Which is what I did... Except multiple people insisted I need to leave the 470k grid leak in place (Including in this thread). I'm more convinced that it's just adding a baseline resistance of 470k and if I removed it, my master volume would work better.
Which is what I did... Except multiple people insisted I need to leave the 470k grid leak in place (Including in this thread). I'm more convinced that it's just adding a baseline resistance of 470k and if I removed it, my master volume would work better.
Are you using the volume control ahead of the phase splitter tube, or is your volume ahead of the pentode part of the 6U8? If you have it at the phase inverter, you'll need a coupling cap between the grid of the phase inverter and the wiper of the volume pot. The top of the volume pot goes to the coupling cap off the plate of the pentode, the bottom goes to ground. The 470k from grid to cathode of the 6u8 triode has to be where it is for biasing reasons.
You might be able to get away with using a 10uf or so cap from the bottom of the volume control to ground instead of changing how the volume pot is wired though.
You might be able to get away with using a 10uf or so cap from the bottom of the volume control to ground instead of changing how the volume pot is wired though.
Natima
New Member
Okay... why?
I've been told multiple different places, by different people that I both DO and DO NOT need that cap there. I had one in, and I was told is was rendering my grid stopper obsolete on another forum.
Look at the schematic I posted most recently... All my changes are on there. You can see where the master is. It's right after the pentode, exactly where it is on Rob Robinettes schematic. Except he doesn't deem it necessary for either the additional capacitor, or the grid leak resistor.
Regarding the master volume:
I have been told I DO and DO NOT need the cap. I asked in this thread whether it should be in series or parallel with the grid stopper. Either way, it seems fine without it.
I have been told I both DO and DO NOT need the 470k Grid Leak. There are plenty of schematics out there showing the exact same cathodyne master volume with NO grid leak (where the pot takes its place), and NO cap coming off the wiper.
I've been told multiple different places, by different people that I both DO and DO NOT need that cap there. I had one in, and I was told is was rendering my grid stopper obsolete on another forum.
Look at the schematic I posted most recently... All my changes are on there. You can see where the master is. It's right after the pentode, exactly where it is on Rob Robinettes schematic. Except he doesn't deem it necessary for either the additional capacitor, or the grid leak resistor.
Regarding the master volume:
I have been told I DO and DO NOT need the cap. I asked in this thread whether it should be in series or parallel with the grid stopper. Either way, it seems fine without it.
I have been told I both DO and DO NOT need the 470k Grid Leak. There are plenty of schematics out there showing the exact same cathodyne master volume with NO grid leak (where the pot takes its place), and NO cap coming off the wiper.
The 470k to the junction of R19 and R20 is there to bias the tube. Basically you're feeding cathode DC voltage into the grid. It should come out approx 2v lower than what you measure at pin 8. Without it, the inverter tube bias would be very wrong and you'd have extremely uneven grid drive to the output tubes. The resistor from grid to volume pot wiper doesn't need to be 470k though. Normally that would be about a 1k resistor. Given that the inverter runs at less than unity gain, you can probably live without the grid stopper entirely but its not going to hurt anything being there. It should be as close to the tube grid as possible for it to be most effective.
The reason it doesn't go to zero volume is because at minimum on the control, you're not at AC ground. Its still some 520k from AC ground. A cap from the bottom of the volume pot to ground would put the pot at AC ground to kill the signal, but not at DC ground so it won't upset the bias. Being on the other side of that 470k resistor, it won't have much effect on the grid drive of the pot either. Basically the cap would replace C11.
The other way to do it is to leave the 470k from the junction of R19 and R20 to pin 9. Use a 0.1uf or so cap from grid to the wiper of the volume pot. Top of the volume pot goes to C12, bottom of the pot to ground. That would not upset DC bias on the phase inverter, but it would shunt the AC audio signal to ground as needed. This is probably the better way of doing it since pots tend to get noisy and stupid with DC on them, though the amount of current flow into the grid is so close to nothing it probably doesn't matter. It would have the effect of not having to worry about 90-odd volts on the pot wiring though.
The reason it doesn't go to zero volume is because at minimum on the control, you're not at AC ground. Its still some 520k from AC ground. A cap from the bottom of the volume pot to ground would put the pot at AC ground to kill the signal, but not at DC ground so it won't upset the bias. Being on the other side of that 470k resistor, it won't have much effect on the grid drive of the pot either. Basically the cap would replace C11.
The other way to do it is to leave the 470k from the junction of R19 and R20 to pin 9. Use a 0.1uf or so cap from grid to the wiper of the volume pot. Top of the volume pot goes to C12, bottom of the pot to ground. That would not upset DC bias on the phase inverter, but it would shunt the AC audio signal to ground as needed. This is probably the better way of doing it since pots tend to get noisy and stupid with DC on them, though the amount of current flow into the grid is so close to nothing it probably doesn't matter. It would have the effect of not having to worry about 90-odd volts on the pot wiring though.
Natima
New Member
So why do all the other cathodyne pre-PI MV schematics out there forego the grid leak resistor and cap isolating the MV?
The reason I am using a 470K grid stopper is to prevent blocking distortion since the preamp has been re-worked for high-gain. Most amps like this have a recommended PI grid stopper of up to 1M.
The reason I am using a 470K grid stopper is to prevent blocking distortion since the preamp has been re-worked for high-gain. Most amps like this have a recommended PI grid stopper of up to 1M.
trobbins
Super Member
The original amp schematic shows C12 coupling cap going to cathodyne PI grid. The shunt capacitor C11 is there to load down V3A output at high frequencies to attenuate the open-loop gain sufficiently at high frequency for the negative feedback loop not to become unstable. You still have feedback connected, so C11 should be positioned back to where it was (directly to C12), unless your change to the feedback resistor has properly achieved stability with your new position of C11.
You can add a volume pot between any valve stage by having an input coupling capacitor, and an output coupling capacitor. That simple concept doesn't seem to be discussed in any reference I can quickly find - and not easy to find a schematic - see Fig 9.2 in https://www.scribd.com/document/365291177/GEC-audio-frequency-amp-design-pdf
Most designers don't want the hassle of using 2 coupling caps, so they typically work out a way to delete one. But in your case, if you want to add a simple volume control then that is one way to do it.
A grid stopper is the effective AC resistance between the previous stages anode and the V3A grid, and is typically used to soften the grid conduction characteristic if the PI is over-driven. It can also act as a treble filter (RC where C is the valve capacitance), that can reduce the level of distortion harmonics generated as grid conduction becomes noticeable. Most people find that at least 220k to 470k is needed as a grid stopper, to notice some taming of grid conduction distortion in a cathodyne. Some amps don't have a signal level high enough at that location to worry about taming PI distortion (ie. the output stage may start clipping before the PI starts to enter any noticeable grid conduction on peaks) - and alternatively, some amps aim to accentuate that form of distortion so do not include any stoppering.
If you then had your original circuit (and forget about C11 for the moment), then inserting a grid stopper to V3B pin 9 upsets the grid leak resistance. Eg. a 470k grid stopper in series with a 470k R18 gives an effective grid leak of 1Meg. A designer has a few options - just add the additional 470k grid stopper - change R18 to 220k and add a 220k grid stopper (this gives the same grid leak resistance).
In your latest circuit post #53, the effective value of grid leak and of grid stopper change significantly with the vol pot wiper position.
You can add a volume pot between any valve stage by having an input coupling capacitor, and an output coupling capacitor. That simple concept doesn't seem to be discussed in any reference I can quickly find - and not easy to find a schematic - see Fig 9.2 in https://www.scribd.com/document/365291177/GEC-audio-frequency-amp-design-pdf
Most designers don't want the hassle of using 2 coupling caps, so they typically work out a way to delete one. But in your case, if you want to add a simple volume control then that is one way to do it.
A grid stopper is the effective AC resistance between the previous stages anode and the V3A grid, and is typically used to soften the grid conduction characteristic if the PI is over-driven. It can also act as a treble filter (RC where C is the valve capacitance), that can reduce the level of distortion harmonics generated as grid conduction becomes noticeable. Most people find that at least 220k to 470k is needed as a grid stopper, to notice some taming of grid conduction distortion in a cathodyne. Some amps don't have a signal level high enough at that location to worry about taming PI distortion (ie. the output stage may start clipping before the PI starts to enter any noticeable grid conduction on peaks) - and alternatively, some amps aim to accentuate that form of distortion so do not include any stoppering.
If you then had your original circuit (and forget about C11 for the moment), then inserting a grid stopper to V3B pin 9 upsets the grid leak resistance. Eg. a 470k grid stopper in series with a 470k R18 gives an effective grid leak of 1Meg. A designer has a few options - just add the additional 470k grid stopper - change R18 to 220k and add a 220k grid stopper (this gives the same grid leak resistance).
In your latest circuit post #53, the effective value of grid leak and of grid stopper change significantly with the vol pot wiper position.
Capacitor instead of the 470k from grid to wiper. R18 would connect direct to the grid of the tube.
As to what guitar amps do, I really don't have a clue why they do some of the stuff they do. In the world I tend to operate in, you never have a volume control inside the feedback loop. It would be ahead of the grid of the pentode. All I can tell you is that if the bottom end of that volume pot doesn't have an AC ground, it will not be able to completely kill the signal getting to the grid of the phase inverter.
The stock design did not use a grid stop resistor at the grid of the inverter. It had C12 as the audio coupling cap, C11 as an HF stability cap, and R18 as the grid resistor tied to the cathode circuit for bias purposes.
As to what guitar amps do, I really don't have a clue why they do some of the stuff they do. In the world I tend to operate in, you never have a volume control inside the feedback loop. It would be ahead of the grid of the pentode. All I can tell you is that if the bottom end of that volume pot doesn't have an AC ground, it will not be able to completely kill the signal getting to the grid of the phase inverter.
The stock design did not use a grid stop resistor at the grid of the inverter. It had C12 as the audio coupling cap, C11 as an HF stability cap, and R18 as the grid resistor tied to the cathode circuit for bias purposes.
one interesting thing you would get from a control within the feedback loop is the potential for serious overdrive from the pentode. The gain of the pentode would increase as you dropped the volume control in an attempt to maintain the output level. Sometime before you hit minimum volume, you should hit maximum output from the pentode so it would get pretty distorted. Possibly thats the goal here?
