Thumping from amp on startup

[bal704]

Regarding the grid leak resistor, are you saying run a 470k resistor (for example) from PI input (PI pin 2) to ground?

I changed the NFB scheme, removing the separate lines from each preamp to the 22k resistor, and connected the 22k resistor from pins 3 & 8 on the PI to the 8 ohm tap on the OT. Didn't stop the thumping, and didn't seem to make much difference on the sound.

I measured the voltages at each of the 5 nodes on my power supply. Note that the first node (called B+) isn't used. Here's what I see:

B+, 465 VDC, 1 VAC
B+1, 441 VDC, 0.8 VAC
B+2, 438 VDC, 0.6 VAC
B+3, 405 VDC, 0.5 VAC
B+4, 400 VDC, 0.5 VAC

I'm getting a bit of AC on the B+ measurements, but it's not much. Don't know if it will make a difference.

Also, the voltage between pins 3 and 8 on the power tubes is 400 VDC, and the voltage between pin 8 and ground is 39 V.

Also, after messing with this thing this afternoon, I need to clarify the thumping sound. About 15 seconds after I turn on the standby switch (turn on the B+), the thumping starts. It lasts for about 15 seconds.

I see a couple of problems with the way the NFB is employed here. ...

To add to the above, the volume and tone controls are inside the NFB loop as drawn - which will cause much weirdness (maybe not your thumping, though). Another reason to move it to the PI.

Also, I don't see a grid leak resistor on the top half of V2 - this could be related to your thumping. The cross-coupling of all the V1B cathodes could cause funny things to happen, especially if all the V1s are fed from the same B+ node, but overall, I tend to concur with the thought that the power supply filtering is insufficient.

Andy

[bal704]

I may have found the problem. One of my power tubes has a small blue flicker that corresponds exactly to the thumping. I swapped the two tubes, and the problem follows that particular tube.

I'll order a couple of more tubes and report back.

Thanks for the help so far.

[Firestorm]

I may have found the problem. One of my power tubes has a small blue flicker that corresponds exactly to the thumping. I swapped the two tubes, and the problem follows that particular tube.

I'll order a couple of more tubes and report back.

Thanks for the help so far.

Probably a response rather than a cause (sorry). Many Fenders (some late tweeds, but especially hi-power Silverfaces) have a reverberating thump, but it lasts only about a second. That's why I asked about where the s/by switch was wired in the p/s circuit. In Fender circuits with a choke, the choke damps the p/s ringing after one thump (more or less.)

[bal704]

The Standby switch is between the rectifier diodes and the first P/S cap.

I may have found the problem. One of my power tubes has a small blue flicker that corresponds exactly to the thumping. I swapped the two tubes, and the problem follows that particular tube.

I'll order a couple of more tubes and report back.

Thanks for the help so far.

Probably a response rather than a cause (sorry). Many Fenders (some late tweeds, but especially hi-power Silverfaces) have a reverberating thump, but it lasts only about a second. That's why I asked about where the s/by switch was wired in the p/s circuit. In Fender circuits with a choke, the choke damps the p/s ringing after one thump (more or less.)

[Phil_S]

Move the standby so that it sits after the first filter cap. The idea is to allow that cap to properly charge before you close the s/b switch and bring the rest of the circuit on line. I am even more convinced the thump is a cap problem.

Actually, you might simply eliminate the standby. It's not needed.

[Tillydog]

Regarding the grid leak resistor, are you saying run a 470k resistor (for example) from PI input (PI pin 2) to ground?

I would have suggested 1M, but yes, from pin 2. to the bottom of your bias resistor R6 (which is currently grounded). Otherwise, how does the grid "know" what DC voltage it should be at?

I changed the NFB scheme...and connected the 22k resistor from pins 3 & 8 on the PI to the 8 ohm tap on the OT. Didn't stop the thumping, and didn't seem to make much difference on the sound.

Regarding the NFB. Probably unrelated to the thumping, you will have the same problem as before, in that your NFB signal will be shorted to ground via the cathode bypass cap (C9). Arguably, the bypass cap serves no purpose here, as the 2 cathodes that share the bias resistor will have equivalent, but out of phase currents. Removing it would allow your NFB to function.

See Merlin's page on paraphase PIs:

http://www.freewebs.com/valvewizard/paraphase.html

You've already had good advice on PSU filtering - I think this is the root of your thumping. Compare the voltage drops on your supply nodes to some established designs and see if it gives you any clues.

Andy

[Firestorm]

I may have found the problem. One of my power tubes has a small blue flicker that corresponds exactly to the thumping. I swapped the two tubes, and the problem follows that particular tube.

I'll order a couple of more tubes and report back.

Thanks for the help so far.

Probably a response rather than a cause (sorry). Many Fenders (some late tweeds, but especially hi-power Silverfaces) have a reverberating thump, but it lasts only about a second. That's why I asked about where the s/by switch was wired in the p/s circuit. In Fender circuits with a choke, the choke damps the p/s ringing after one thump (more or less.)

[bal704]

OK. Here's what I did tonight, with results.

1) I borrowed another set of tubes from one of my other amps, and dropped them in. THUMPING GONE.

2) I added a 470k resistor from pin 2 on the PI to ground.

3) In order to install the NFB at the PI level, I clipped C9 out of the circuit. This is the cap on pins 3&8 of the PI. Big problems. All I got was a huge, house shaking howl with the master volume at 0. Huge. When I re-installed C9, all quiet again.

4) I added a 22k resistor from PI 3-8 to the 8 ohm tap. Didn't really seem to help, as I was getting mucho gain on all channels. Feedback inducing gain. The only thing I've done so far to solve this is add a 22k resistor from the cathode on each preamp tube to the 8 ohm tap. The PI level NFB doesnt seem to work.

[Structo]

Sounds like you are making progress.

Check this schematic out.

It is a D'lite power amp.

But what I want you to look at is the NFB.

The 6K2 resistor controls the amount of feedback and depending on which power tubes you use, changing this value will determine the level of NFB.

This scheme is for 6V6 so with 6L6 it would probably be better at 4K7.

Less in value will allow more feedback more will add rawness to the tone.

[Tillydog]

OK. Here's what I did tonight, with results.

1) I borrowed another set of tubes from one of my other amps, and dropped them in. THUMPING GONE.

2) I added a 470k resistor from pin 2 on the PI to ground.

OK - Out of curiosity, do you still get the thumping with the old tubes after doing (2) ?

I'm assuming that you left this in place:

I changed the NFB scheme, removing the separate lines from each preamp to the 22k resistor, and connected the 22k resistor from pins 3 & 8 on the PI to the 8 ohm tap on the OT.

before doing this:

3) In order to install the NFB at the PI level, I clipped C9 out of the circuit. This is the cap on pins 3&8 of the PI. Big problems. All I got was a huge, house shaking howl with the master volume at 0. Huge. When I re-installed C9, all quiet again.

OK, so your feedback tap now works, but you need to to flip the OT primaries (OR secondaries, OR output tube grid connections) to get the phase right.

4) I added a 22k resistor from PI 3-8 to the 8 ohm tap.

How is that different from what you did above?

Didn't really seem to help, as I was getting mucho gain on all channels. Feedback inducing gain. The only thing I've done so far to solve this is add a 22k resistor from the cathode on each preamp tube to the 8 ohm tap.

The signal is inverted by your preamp stage, so if you have things configured for NFB on the preamp, and move the connection to the PI without flipping the OT primaries (for example) you will end up with positive feedback as you (and you neighbours??) heard.

The PI level NFB doesnt seem to work.

Not yet....

Andy

[bal704]

I did step (2) before I changed out the tubes. Still had thumping. I had the PI NFB installed when I cut out C9, and it was howling. Removed it while C9 was out, still howling. Added C9 back in, no howling with or without the PI NFB.

When I had the PI NFB installed, when I turned the gain on any channel to about 8 the amp started squealing. From about 1-8, the amp sounded fine. It only squealed when I really pushed the gain. When I installed the NFB at each preamp cathode to the OT tap, I don't get squealing at any gain level. I'll try using the PI NFB and swap the OT primaries.

This started out as a separate preamp/power amp project, with the goal to be able to have a reliable power amp side to match to any preamp I might build. I started with the ax84 20w pp as the model, but I didn't want to mess with the biasing on the power side. I eventually used the Fender 5C5 power amp side as the model, which doesn't have any biasing or NFB.

I used the weber schematic as the template:

http://taweber.powweb.com/store/5c5_schem.jpg

Some of the values don't match exactly because I had already started on the ax84 path before I decided to swap. But that is how I got where I'm at, since it appears I've now got a seriously bastardized amp.

Thanks for the help so far.

OK. Here's what I did tonight, with results.

1) I borrowed another set of tubes from one of my other amps, and dropped them in. THUMPING GONE.

2) I added a 470k resistor from pin 2 on the PI to ground.

OK - Out of curiosity, do you still get the thumping with the old tubes after doing (2) ?

I'm assuming that you left this in place:

I changed the NFB scheme, removing the separate lines from each preamp to the 22k resistor, and connected the 22k resistor from pins 3 & 8 on the PI to the 8 ohm tap on the OT.

before doing this:

3) In order to install the NFB at the PI level, I clipped C9 out of the circuit. This is the cap on pins 3&8 of the PI. Big problems. All I got was a huge, house shaking howl with the master volume at 0. Huge. When I re-installed C9, all quiet again.

OK, so your feedback tap now works, but you need to to flip the OT primaries (OR secondaries, OR output tube grid connections) to get the phase right.

4) I added a 22k resistor from PI 3-8 to the 8 ohm tap.

How is that different from what you did above?

Didn't really seem to help, as I was getting mucho gain on all channels. Feedback inducing gain. The only thing I've done so far to solve this is add a 22k resistor from the cathode on each preamp tube to the 8 ohm tap.

The signal is inverted by your preamp stage, so if you have things configured for NFB on the preamp, and move the connection to the PI without flipping the OT primaries (for example) you will end up with positive feedback as you (and you neighbours??) heard.

The PI level NFB doesnt seem to work.

Not yet....

Andy

[Structo]

I guess what I don't understand is why you are providing global negative feedback for a preamp tube?
That is generally reserved for a long tail phase inverter.

You can use local negative feedback on a preamp tube, but usually only the first or second stage.

[bal704]

I have swapped the OT primaries. So, I now have:
1) C9 removed
2) a 22k resistor between PI 3 & 8 and the 8 ohm OT tap
3) a 470k resistor between PI pin 2 and ground
4) no NFB resistors on the preamp tubes.

Upon power up, I get no major excitement. There is no howling or screaming. There is a tremendous amount of gain on all three channels, even with the gain on each channel in the 3-4 range. Taking the gain up to 8-10 causes squealing on all channels. Also, when I strum my guitar hard, I get a high frequency thump on all channels. When playing normally, this doesn't happen. I assume this is some kind of clipping caused by the already hard driven preamp tubes getting a really hot signal from the guitar?

PI 3 and 8 are currently tied together. Do I need to give them their own resistor to ground?

I guess what I don't understand is why you are providing global negative feedback for a preamp tube?
That is generally reserved for a long tail phase inverter.

You can use local negative feedback on a preamp tube, but usually only the first or second stage.

[Tillydog]

Taking the gain up to 8-10 causes squealing on all channels.

Likely to be a lead dress / layout issue IMHO, but that's part of the fun of 'rolling your own'.

Don't forget you've got a fair amount of gain built into that PI.

You could lose some gain not bypassing one or both of the preamp cathodes, or by more NFB - reduce the value of the NFB resistor and see if you like it. (Actually, it might make things simpler to disconnect the NFB completely until the other issues are resolved).

Also, when I strum my guitar hard, I get a high frequency thump on all channels.

It could be many things (e.g. blocking distortion), but your description makes me think it could be instability in the power supply. I would be looking for a much bigger voltage difference between your B+ 3 & 4, and would look to get a more conventional profile there first (YMMV).

I don't know how the paraphase PI responds to being overdriven, but someone else might.

It's handy to have a 'scope in these circumstances. Second best is to model the circuit in something like LTSpice to get an idea of what might be happening.

PI 3 and 8 are currently tied together. Do I need to give them their own resistor to ground?

No - see Merlin's page that I linked to in a previous post.

Andy

[bal704]

I'm going to rebuild the power filtering layout on this amp.

I was going to use the 5E3 power filtering schematic as the template:

http://taweber.powweb.com/store/5e3_schem.jpg

Does this sound like a good plan?

Also, on the 5E3 schematic, I don't see the NFB. Does it not have any?