I have tried different output tubes and replaced screen resistors. Plate volts and bias volts are fine. Here is a clip of the noise:
JCM800 noise - YouTube
JCM800 noise - YouTube
Check in this order:
1. Short the plates of the driver tube together (pins #1&6). If the noise goes away or is greatly diminished, the problem is in the power supply feeding the phase inverter stage.
2. Short pin 5 of the output tubes together. If the noise goes away or is greatly diminished, the problem is in the bias supply or coupling caps.
3. If ungrounded, short pin 8 of the output tubes together. If the noise goes away or is greatly diminished, the problem is in the cathode current monitoring resistors.
4. Try different output tubes.
5. Surely hope it is not arcing in the output transformer!
Dave
1. Short the plates of the driver tube together (pins #1&6). If the noise goes away or is greatly diminished, the problem is in the power supply feeding the phase inverter stage.
2. Short pin 5 of the output tubes together. If the noise goes away or is greatly diminished, the problem is in the bias supply or coupling caps.
3. If ungrounded, short pin 8 of the output tubes together. If the noise goes away or is greatly diminished, the problem is in the cathode current monitoring resistors.
4. Try different output tubes.
5. Surely hope it is not arcing in the output transformer!
Dave
Hello dcgillespie, I shorted pin 5 of the output tubes together and the noise is gone. Thank you so much for helping me but now I have to find the culprit. I am getting negative 45vdc at pin 5 for both tubes. The amp also has a master volume control with blocking caps on both sides. If a blocking cap was passing dc wouldn't the master volume stop it from passing to pin 5? I am guessing that my problem is in the bias supply but not sure. I am going to poke around in there tonight and check back here with more info.
Everything looks ok to me but I am not tech so I guess I will just start replacing things in the bias supply.
Famous -- leave the bias supply and grid return resistors in place, but disconnect the coupling caps into the output stage on the output stage side. If the noise is still there, it must originate from the bias supply. If it is gone, and the driver tube is removed, then it must be a coupling cap, driver plate resistor, or even the dropping resistor or power supply cap for the driver stage.
I take it this amp is using a dual control to function as a master volume control between the driver and the output stage? Disconnect the coupling caps into the output stage first, as if the noise is still there, then this master volume circuit has been completely eliminated from possibility.
Keep working the problem by a process of elimination, and the fault will usually show itself quickly enough.
Dave
I take it this amp is using a dual control to function as a master volume control between the driver and the output stage? Disconnect the coupling caps into the output stage first, as if the noise is still there, then this master volume circuit has been completely eliminated from possibility.
Keep working the problem by a process of elimination, and the fault will usually show itself quickly enough.
Dave
Hello Dave, yes the amp is using a dual control for the master volume. If you are interested here is a schematic:
http://www.drtube.com/schematics/marshall/4210.gif
I am going to lift the coupling caps on the output side of the master volume and see if the noise is still there.
http://www.drtube.com/schematics/marshall/4210.gif
I am going to lift the coupling caps on the output side of the master volume and see if the noise is still there.
Last edited:
This is very interesting. Any noise injected by the bias system itself is injected common mode fashion, so generally it should be non-audible. Based on what you are telling me then:
1. Lifting the coupling caps into the output stage does not stop the noise.
2. Shorting pins #5 together of the output tubes does stop the noise.
3. Different output tubes do not stop the noise.
The only thing left is the 220K grid resistors. If just one of those became noisy, then the noise created would not be a common mode signal, and would sound in the output. Check the 220K resistors by shorting across them one at a time, or both at once if shorting either one does not completely stop the noise. If shorting out one (or both) of them stops the noise, the one shorted when the noise stops is the culprit.
Dave
1. Lifting the coupling caps into the output stage does not stop the noise.
2. Shorting pins #5 together of the output tubes does stop the noise.
3. Different output tubes do not stop the noise.
The only thing left is the 220K grid resistors. If just one of those became noisy, then the noise created would not be a common mode signal, and would sound in the output. Check the 220K resistors by shorting across them one at a time, or both at once if shorting either one does not completely stop the noise. If shorting out one (or both) of them stops the noise, the one shorted when the noise stops is the culprit.
Dave
Hello Dave, reflowed some sketchy solder joints in the bias supply and replaced the 150k resistor just before the diode and the noise is gone. Maybe a cold solder joint arcing was the noise but I am not sure. The 220k resistors you speak of are 150k in the U.S. version that I own (the 220k are for EL34's). I am very interested in how you diagnosed this and also interested in what common mode is why it won't be audible. Any information you have is greatly appreciated as I really like to learn about this stuff.
As for the noise I am going to put in the preamp tubes and plug a guitar in and see if the noise comes back.
As for the noise I am going to put in the preamp tubes and plug a guitar in and see if the noise comes back.
Ok so the noise is gone with preamp tubes out. The noise comes back when the phase splitter is installed.
Famous -- In this case, common mode means any signal of identical amplitude and phase injected into both sides of the push-pull signal path.
Each side of the push-pull signal path is normally 180 degrees out of phase with the other side. This is in fact the primary purpose of the phase inverter stage: it takes an unbalanced audio signal, and turns it into two balanced audio signals (relative to ground) that are equal in every way, but opposite in phase with each other. These out of phase signals are ultimately re-combined in the push-pull output transformer to produce a standard unbalanced output on the secondary winding to drive the speakers with.
Since a push-pull output stage requires equal but opposite phase signals to be applied to it, then it follows that two identical signals of the same phase applied to this stage would simply cancel each other out in the output transformer, and produce no sound at all. Since any noise in the bias supply would appear as an identical signal injected into both sides of the push-pull output stage, the noise should therefore cancel itself out and not be heard.
I believe you likely hit on the real problem, with an arcing or very noisy bias dropping resistor or relevant solder joints, that was not being directly injected into the output stage, but simply being picked up by it, amplified, and reproduced in the output as static. I'm glad you found your culprit!
I hope this helps answer your question.
Dave
Each side of the push-pull signal path is normally 180 degrees out of phase with the other side. This is in fact the primary purpose of the phase inverter stage: it takes an unbalanced audio signal, and turns it into two balanced audio signals (relative to ground) that are equal in every way, but opposite in phase with each other. These out of phase signals are ultimately re-combined in the push-pull output transformer to produce a standard unbalanced output on the secondary winding to drive the speakers with.
Since a push-pull output stage requires equal but opposite phase signals to be applied to it, then it follows that two identical signals of the same phase applied to this stage would simply cancel each other out in the output transformer, and produce no sound at all. Since any noise in the bias supply would appear as an identical signal injected into both sides of the push-pull output stage, the noise should therefore cancel itself out and not be heard.
I believe you likely hit on the real problem, with an arcing or very noisy bias dropping resistor or relevant solder joints, that was not being directly injected into the output stage, but simply being picked up by it, amplified, and reproduced in the output as static. I'm glad you found your culprit!
I hope this helps answer your question.
Dave
If the noise has gone away with the replacement of the bias resistor and reflowing of the relevant solder joints with only the output tubes installed, then something else (likely associated with the bias supply) is still generating noise that the amplifier is picking up and reproducing as noise in the output.
As you start installing more of the small signal tubes, more amplification is taking place, so it is able to respond to smaller noise sources that can be output as noise. With the phase inverter tube installed, can you control the level of the noise with the master volume control? If your bias supply fix did in fact resolve the problem, then turning the master volume control full down should still produce no noise. Is that the case?
Dave
As you start installing more of the small signal tubes, more amplification is taking place, so it is able to respond to smaller noise sources that can be output as noise. With the phase inverter tube installed, can you control the level of the noise with the master volume control? If your bias supply fix did in fact resolve the problem, then turning the master volume control full down should still produce no noise. Is that the case?
Dave
Yes, the master volume on zero has no noise but when the master volume is turned up the noise returns. I must have had two problems going on at the same time, and I must have fixed the bias problem (hopefully).
And your answer helped me gain a better understanding to how all this stuff works. VERY INTERESTING STUFF! Thanks again and I really appreciate all the help.
And your answer helped me gain a better understanding to how all this stuff works. VERY INTERESTING STUFF! Thanks again and I really appreciate all the help.
Ok here are my PI voltages:
Before plate loaders 323vdc
after 82k 200vdc
after 100k 180vdc
cathodes 44vdc
both grids have 30vdc
The total cathode resistance is not 20,470 like the schematic says, it's 15,170. The last resistor on the tail is 4.7k not 10k. So doing the math the voltages look correct. But I don't think there is suppose to be any dc voltage at the grids of the phase inverter. I think I have an extra .022 cap laying around should I swap it out with the coupling cap from the previous stage or is voltage at the grids normal?
Before plate loaders 323vdc
after 82k 200vdc
after 100k 180vdc
cathodes 44vdc
both grids have 30vdc
The total cathode resistance is not 20,470 like the schematic says, it's 15,170. The last resistor on the tail is 4.7k not 10k. So doing the math the voltages look correct. But I don't think there is suppose to be any dc voltage at the grids of the phase inverter. I think I have an extra .022 cap laying around should I swap it out with the coupling cap from the previous stage or is voltage at the grids normal?
voltage drop across the 100k resistor is 143v so
143/100,000 = .0014 or 1.4ma
voltage drop across the 82k is 123v so
123/82,000 = .0015 or 1.5ma
Seems like a lot of current looking at the 12ax7 datasheet. I wonder if the positive voltage at the grid is throwing the bias off and causing my noise issue?
143/100,000 = .0014 or 1.4ma
voltage drop across the 82k is 123v so
123/82,000 = .0015 or 1.5ma
Seems like a lot of current looking at the 12ax7 datasheet. I wonder if the positive voltage at the grid is throwing the bias off and causing my noise issue?
The current draw of the stage is normal. As a phase inverter/driver, it is designed to draw more current since more output is needed from it to drive the output tubes.
The voltage on the grids should be positive since they are not returned to ground, but to the junction of the 470 cathode resistor, and the 10K tail. They will be positive by very close to the amount measured at that point. However, that you are measuring 44 volts on the cathodes, but only 30 on the grids begs the question as to what type of meter you are using for your measurements. If it does not have a high input impedance (such as about 10 megohms), it will "load down" the voltage at the point of measurement, since it is being supplied through 1 meg resistors. That will cause significant error in the reading at certain places, such as you are getting.
If you have an extra .022 cap, try connecting it between ground, and the grid of V5 that that is driven from V3 by a .022 cap (this will be either pin 2 or 7, but is not designated on the schematic). This will effectively short that grid to ground with respect to any AC signal presented there. If the noise stops in that condition, then the phase inverter stage is fine, but picking up the noise signal and merely amplifying it. If the noise is still present, then it is being generated in the phase inverter stage, or the power supply to it. Have you tried a new tube in that position?
Dave
The voltage on the grids should be positive since they are not returned to ground, but to the junction of the 470 cathode resistor, and the 10K tail. They will be positive by very close to the amount measured at that point. However, that you are measuring 44 volts on the cathodes, but only 30 on the grids begs the question as to what type of meter you are using for your measurements. If it does not have a high input impedance (such as about 10 megohms), it will "load down" the voltage at the point of measurement, since it is being supplied through 1 meg resistors. That will cause significant error in the reading at certain places, such as you are getting.
If you have an extra .022 cap, try connecting it between ground, and the grid of V5 that that is driven from V3 by a .022 cap (this will be either pin 2 or 7, but is not designated on the schematic). This will effectively short that grid to ground with respect to any AC signal presented there. If the noise stops in that condition, then the phase inverter stage is fine, but picking up the noise signal and merely amplifying it. If the noise is still present, then it is being generated in the phase inverter stage, or the power supply to it. Have you tried a new tube in that position?
Dave
Yes I tried a couple of tubes in that position and the noise is still present. I will short the input from the previous stage through the .022 cap to ground and let you know what happens.
Ok so I shorted Pin 2 to ground through a .022 cap and the noise is still there but at about 75% intensity it was before. During this test the tubes in the amp were the PI and the output tubes only, tubes 1-4 were removed.
My meter I am using is an Ideal 61-360. I am not sure about the input impedence and I don't have another meter to check it.
I don't know anything about that particular meter, but your readings would indicate that its input impedance is rather low. Is it a digital or analog type?
Since reducing the volume control eliminates the noise, but placing an AC short at the input of the PI stage does not eliminate it, that would indicate that the noise is most likely being generated in the PI stage. The most likely candidates are the plate load resistors but caps can cause noise as well.
Next test would be to separately ground the top of each master volume control section to see if the noise is predominantly coming through one side of the push-pull signal or the other, or both.
Dave
Since reducing the volume control eliminates the noise, but placing an AC short at the input of the PI stage does not eliminate it, that would indicate that the noise is most likely being generated in the PI stage. The most likely candidates are the plate load resistors but caps can cause noise as well.
Next test would be to separately ground the top of each master volume control section to see if the noise is predominantly coming through one side of the push-pull signal or the other, or both.
Dave
- Status
- Not open for further replies.