This video is a prime example of how remote/distance learning is the future. I sure wish the instructors I had in college were this clear and lucid. Would have learned a lot more and better. I have a couple Scott tube amps in the basement that are dormant and need restoration. One of these days. When I get around to it. One thing this video brought up which I never thought about - at least consciously - is the question of input voltage. I am old enough to remember when people stated house voltage as 110. Way back when. Our whole house ran of two 20/30 amp fuses. Then house voltage stealthily crept up to 120. Like the boiled frog. It happens and you don't notice. Just checked the voltage - same house - and it reads 121/122. Something to remember when working on this stuff designed for 110.
Welcome back Mark! Would you consider changing the first resistor in the CRC network after the Rectifier tube to drop the voltage at that point? That would have the added benefit of bringing all of the voltages in the power supply chain down to specs and relieve some of the overvoltage that the filter capacitors see.
This would be my choice, to put all the voltages in the amp back to their design values, in particular the the ratio of the output tubes' plate to screen voltages.
This is about sub-optimal design of this amplifier in terms of power output and power consumption. The problem was how to obtain relatively stable screen voltage when screen current increases (from from 0.8 mA to 12.5 mA per tube, or 15-fold increase) with increasing signal level. Scott's solution was to shunt screen supply with that stick power resistor, so that now zero signal total screen circuit current is 44 mA and max signal current is 90 mA, or only 2-fold difference. The price paid for such screen voltage "stabilization" is not only heat, but waste of precious 45 watts of power from a tube-rectified power supply.
Thanks for the video Mark great explanation. A handy piece of equipment to use on old amps and tube grear would be a step down transformer then don't need to modify circuit :)
According to the Sylvania datasheet for the 7189/7189A the Screen (G2) max dissipation is from 2.2 to 4.4 watts from no signal to full signal per pair of tubes in push-pull. That's an additional current draw for the total amplifier of 14 ma at quiescent state to near 30 at maximum signal assuming only 300 VDC on the G2s. Scott may not have been driving the tubes all the way to design maximum, but at 345 volts they are probably close to it and there is a significant current demand by the screens of the output tubes that would need to be considered in calculating the voltage division at the bottom of the 1.2k resistor and its available power dissipation.
222c real power output is about 10 watts/channel. The power is limited by voltage sag in the RC-filtered power supply, and current capability of the 5AR4 rectifier. 299c is also 10 watts, for exactly the same reason, despite higher B+ and 7591 tubes.
Yeah, I don’t know what HH Scott were thinking on the (at least 7) variants of the 299c. Very high power transformer voltages on my particular one. The schematic matched closer to a 233.
Boy, if only every product had a) schematics, and b) test point voltages, troubleshooting would be much easier. Good video ... love troubleshooting videos.
Mark, Adjusting the screen voltage makes sense. However, the plate voltage should also be significantly higher, due to the higher input from the transformer. Will the the increased difference in plate-to-screen voltage have any significance? Should the plate voltage also be adjusted to correct for the higher line input voltage of today?
Thanks a lot .I have a collection of HHscott amps 222,299, etc .I was thinking to replace all those resistors with a big voltage transistor regulator ..
That would work. If screen voltage is stabilized at 300 V, there would be no need in that bleeder resistor. A separate screen supply will make it possible to realize 24 W power output (with the original power supply the amp clips at 10W).
First I don't know Jack Crap but need to ask... Wouldn't it be better to adjust the resistor in the C-R-C to establish the voltage across the total network of the circuit? Maybe do the same for the first resistor after the full wave rectifier and for the filaments? Would it matter? What problems could it cause?
No, that won't work. Screen current of output tubes dramatically changes with signal, so screen voltage needs some kind of stabilization. This is done by adding bleeder resistor parallel to screens.
not at DC it isn't an option - you can't make a dc voltage divider with capacitors - capacitors look like an open at dc. Remember for capacitors, impedance = 1 / jwc, when w=0, impedance is infinite. What they could've done is scaled all the resistor values in the divider up by some factor, and the voltage divider would draw less current, but the drive strength in the middle (the 345 v rail in this case) would be weaker, and could wander. I have no idea how much current the screens of tubes draw, but if they truly don't draw much current, you could probably double the resistor values in the divider (2.4k on top, and 16k on the bottom) and end up with 1/4 of the power being dissipated. What determines what values you should use in the divider is the equivalent resistance of the 4 tube's screens to ground in parallel with the bottom resistor in the divider. This is all very basic "circuits 101" analysis. Edited to add: I was reading another comment below that the screens can draw up to 30mA for those tubes x 4 tubes in parallel = 120mA - that's starting to be significant, so you probably couldn't scale up the resistors in the divider - it would change the screen voltage too much as the load changes.
gorak9000 Where does it say the screen has to be at a constant DC voltage? Common practice in preamp tubes to decouple the screen with a capacitor. Or part of the output transformer come to think of it.
Why do vintage guitar amps use capacitor values such as 20uf or 10uf in their power filtering? Is that because, unlike hi-fi, lower headroom and sag aren't always a bad thing? What would be the result if a modern cap is used such as a 220uf in place of a 20uf?
As long as it is not the first capacitor in tube-rectified power supply, a 220 uF capacitor can be used in place of 20 uF. But it won't make the supply any better. Push-pull circuit does not need low level of ripple: ripple is common mode for it.
Aren't ALL of the voltages coming off the power supply likely to be "higher" than normal? Wouldn't it be better to drop some more voltage on R205 (the first resistor after the rectifier tube)? That way all of the power rails would be closer to spec?
I have a tube amp designed for 110 volts and I have about 123 volts at the receptacle. To remedy the situation I bought a small 120 to 12 volt xmfr and wired it in buck configuration. Since only about 10% of the voltage needed to be bucked the xmfr was pretty small. The upshot is that the amp is getting the 110 volts it wants so I didn't have to fiddle with the circuit. Eric
I wouldn't worry about increased B+. Output tubes can tolerate significantly more than on the 222c schematic because at zero signal plate and screen currents are very low. With signal, these voltages will go down because of voltage drop in transformer secondary, rectifier, and 80 ohm resistor.
So in one of XRayTonyB's videos he talked about a similar circuit with 2 power resistors that were just bleeding current to ground, and what that was about. He wasn't sure why they made that design choice, but it was also on an HH Scott unit. ruclips.net/video/aoiHCxcSjuA/видео.html
The I (45 mA) that you used does not account for the current taped off at the 1.2k-8k junction feeding the screen grids of the 7189(s) and what ever the pentodes are above them . A better way to calculate would be P=E^2/R=345^2/8200= 14.5 Watts for the 8k resistor.
Good learning session. Would wonder if putting transformer inbetween house and unit at proper power output required and 110vac output would be best instead of altering amp, possibly causing other problems.
So glad to see you back Mark. really missed your videos.
This video is a prime example of how remote/distance learning is the future. I sure wish the instructors I had in college were this clear and lucid. Would have learned a lot more and better. I have a couple Scott tube amps in the basement that are dormant and need restoration. One of these days. When I get around to it. One thing this video brought up which I never thought about - at least consciously - is the question of input voltage. I am old enough to remember when people stated house voltage as 110. Way back when. Our whole house ran of two 20/30 amp fuses. Then house voltage stealthily crept up to 120. Like the boiled frog. It happens and you don't notice. Just checked the voltage - same house - and it reads 121/122. Something to remember when working on this stuff designed for 110.
Great video Mark, clean and clear technical discussion, thank you. Hope your business ventures are going well. Be safe.
Welcome back Mark! Would you consider changing the first resistor in the CRC network after the Rectifier tube to drop the voltage at that point? That would have the added benefit of bringing all of the voltages in the power supply chain down to specs and relieve some of the overvoltage that the filter capacitors see.
Yes that works too if all the other voltages are high. For sure
This would be my choice, to put all the voltages in the amp back to their design values, in particular the the ratio of the output tubes' plate to screen voltages.
This is about sub-optimal design of this amplifier in terms of power output and power consumption. The problem was how to obtain relatively stable screen voltage when screen current increases (from from 0.8 mA to 12.5 mA per tube, or 15-fold increase) with increasing signal level. Scott's solution was to shunt screen supply with that stick power resistor, so that now zero signal total screen circuit current is 44 mA and max signal current is 90 mA, or only 2-fold difference. The price paid for such screen voltage "stabilization" is not only heat, but waste of precious 45 watts of power from a tube-rectified power supply.
Great to see you back, Mark!
Ditto!
I got a lot out of this. Thanks for taking the time.
That was clear. Thanks. Tube amps put out a lot of heat and I noticed a similar resistor in my Fisher that was concerning, now I understand.
Thanks for the video Mark great explanation. A handy piece of equipment to use on old amps and tube grear would be a step down transformer then don't need to modify circuit :)
According to the Sylvania datasheet for the 7189/7189A the Screen (G2) max dissipation is from 2.2 to 4.4 watts from no signal to full signal per pair of tubes in push-pull. That's an additional current draw for the total amplifier of 14 ma at quiescent state to near 30 at maximum signal assuming only 300 VDC on the G2s. Scott may not have been driving the tubes all the way to design maximum, but at 345 volts they are probably close to it and there is a significant current demand by the screens of the output tubes that would need to be considered in calculating the voltage division at the bottom of the 1.2k resistor and its available power dissipation.
222c real power output is about 10 watts/channel. The power is limited by voltage sag in the RC-filtered power supply, and current capability of the 5AR4 rectifier. 299c is also 10 watts, for exactly the same reason, despite higher B+ and 7591 tubes.
Yeah, I don’t know what HH Scott were thinking on the (at least 7) variants of the 299c. Very high power transformer voltages on my particular one. The schematic matched closer to a 233.
Boy, if only every product had a) schematics, and b) test point voltages, troubleshooting would be much easier. Good video ... love troubleshooting videos.
Excellent info! Thank you.
Mark,
Adjusting the screen voltage makes sense. However, the plate voltage should also be significantly higher, due to the higher input from the transformer. Will the the increased difference in plate-to-screen voltage have any significance? Should the plate voltage also be adjusted to correct for the higher line input voltage of today?
He should get the plate voltage in check too :-)
Thanks a lot .I have a collection of HHscott amps 222,299, etc .I was thinking to replace all those resistors with a big voltage transistor regulator ..
That would work. If screen voltage is stabilized at 300 V, there would be no need in that bleeder resistor. A separate screen supply will make it possible to realize 24 W power output (with the original power supply the amp clips at 10W).
First I don't know Jack Crap but need to ask... Wouldn't it be better to adjust the resistor in the C-R-C to establish the voltage across the total network of the circuit? Maybe do the same for the first resistor after the full wave rectifier and for the filaments? Would it matter? What problems could it cause?
No, that won't work. Screen current of output tubes dramatically changes with signal, so screen voltage needs some kind of stabilization. This is done by adding bleeder resistor parallel to screens.
If this was a tuner IF stage that resistor would be a capacitor. Wondering whether that isn't an option here?
not at DC it isn't an option - you can't make a dc voltage divider with capacitors - capacitors look like an open at dc. Remember for capacitors, impedance = 1 / jwc, when w=0, impedance is infinite. What they could've done is scaled all the resistor values in the divider up by some factor, and the voltage divider would draw less current, but the drive strength in the middle (the 345 v rail in this case) would be weaker, and could wander. I have no idea how much current the screens of tubes draw, but if they truly don't draw much current, you could probably double the resistor values in the divider (2.4k on top, and 16k on the bottom) and end up with 1/4 of the power being dissipated. What determines what values you should use in the divider is the equivalent resistance of the 4 tube's screens to ground in parallel with the bottom resistor in the divider. This is all very basic "circuits 101" analysis. Edited to add: I was reading another comment below that the screens can draw up to 30mA for those tubes x 4 tubes in parallel = 120mA - that's starting to be significant, so you probably couldn't scale up the resistors in the divider - it would change the screen voltage too much as the load changes.
gorak9000 Where does it say the screen has to be at a constant DC voltage? Common practice in preamp tubes to decouple the screen with a capacitor. Or part of the output transformer come to think of it.
Why do vintage guitar amps use capacitor values such as 20uf or 10uf in their power filtering? Is that because, unlike hi-fi, lower headroom and sag aren't always a bad thing? What would be the result if a modern cap is used such as a 220uf in place of a 20uf?
As long as it is not the first capacitor in tube-rectified power supply, a 220 uF capacitor can be used in place of 20 uF. But it won't make the supply any better. Push-pull circuit does not need low level of ripple: ripple is common mode for it.
Aren't ALL of the voltages coming off the power supply likely to be "higher" than normal? Wouldn't it be better to drop some more voltage on R205 (the first resistor after the rectifier tube)? That way all of the power rails would be closer to spec?
Yes, this could be done too. Without having his unit in hand, its hard to tell.
I have a tube amp designed for 110 volts and I have about 123 volts at the receptacle. To remedy the situation I bought a small 120 to 12 volt xmfr and wired it in buck configuration. Since only about 10% of the voltage needed to be bucked the xmfr was pretty small. The upshot is that the amp is getting the 110 volts it wants so I didn't have to fiddle with the circuit.
Eric
I wouldn't worry about increased B+. Output tubes can tolerate significantly more than on the 222c schematic because at zero signal plate and screen currents are very low. With signal, these voltages will go down because of voltage drop in transformer secondary, rectifier, and 80 ohm resistor.
So in one of XRayTonyB's videos he talked about a similar circuit with 2 power resistors that were just bleeding current to ground, and what that was about. He wasn't sure why they made that design choice, but it was also on an HH Scott unit.
ruclips.net/video/aoiHCxcSjuA/видео.html
The purpose is to stabilize screen voltage.
The I (45 mA) that you used does not account for the current taped off at the 1.2k-8k junction feeding the screen grids of the 7189(s) and what ever the pentodes are above them . A better way to calculate would be P=E^2/R=345^2/8200= 14.5 Watts for the 8k resistor.
By the same token the 1.2 k resistor would be dissipating (420v-345v)^2/1200 ohms=4.68 watt.
Surely it would've been better to have corrected the complete power supply
I've always thought H. H. Scott made some odd design choices. IMO, Fisher did a better job.
Agree. 10 W real output of 222c and 299c is pathetic.
Good learning session. Would wonder if putting transformer inbetween house and unit at proper power output required and 110vac output would be best instead of altering amp, possibly causing other problems.
It is not necessary. 117 vs. 125 VAC is about 7% difference, not important.
@@anatolygrishin4234 I hear you.