Bassman aa864 Random noise and squeal when all preamp tubes and PI pulled

[Erik Aa]

Hi all. I have built a bassman aa864 circuit and it sounded ok. I just replaced all the preamp and phase inverter sockets to get better connections. After I replaced the sockets, I have a random noise popping squeal in the amp. With the preamp tubes and PI installed the noise is really loud. When I pull all the preamp tubes and PI the noise is still there but the noise  volume is much lower. The noise does not react to volume or tone controls. What on earth is going on? Filament wiring? Resistors? I have inspected and reflowed all solder joints on the new sockets to no avail. Thanks. 

[72Blazer]

Post your schematic and some high resolution pictures of your amp.  Since what you are describing/if indeed the noises started afteer you replaced the tube sockets-ensure all your connections to the the pre-amp tubes are correct and have good connections.  It's easy to make a mistake after replacing several tube sockets.


Vr
J

[AlNewman]

if you disconnect negative feedback, do you still get the squeal?

[Erik Aa]

I don’t use negative feedback, so it is disconnected. The noise is more dirty popping swooshing creaking light squealing electrical interference ish than a loud squeal. It is not a positive feedback squeal. Thanks.

[Esquirefreak]

Sounds like something might be microphonic. Possibly a solder joint or tubes. If you used solid core wire, those can become microphonic in some cases.

It could also be RFI. In that case, try to shield the amp by putting some aluminum foil/sheet metal over the amp chassis

A bit of chop-sticking might find the culprit.

[Erik Aa]

I have tried with a different set of 6l6s and the noise is still there. I have checked for dc on preamp tube grids, chop sticked, and checked resistors on the power tubes. The white filament wire on the preamp section is indeed solid core. Maybe I should just pull it out and try again. Could it be an issue with the artificial center tap on the filaments? I used 100 Ohm resistors, but when I measured them now they read about 52 ohms each? Thanks. Ps. The amp is a gutted peavey Classic 50.

[bmccowan]

The white filament wire on the preamp section is indeed solid core. Maybe I should just pull it out and try again.

I suggest that you do not start replacing additional things. You said the amp sounded "OK" before you changed the sockets. What does that mean? OK but not good? OK but noisy? what led you to replace the sockets? All my builds use solid core wire for the heaters.
If replacing the sockets introduced the noise, a mistake was made. Now you need to find and correct it.
Whenever I have a noise like you describe, it turns out to be a bad connection or a wrong connection. It my case it is usually at the phase inverter socket - maybe because there is usually more stuff going on there
If you move your hand or a tool near each socket - but not touching - does the noise change? Sometimes that can help to find the location.
You may want to follow a full start-up debugging procedure. Doug has one here on the site.
You have checked voltages? Including AC on the heaters?

[Erik Aa]

Thanks. The amp sounded ok in that there were no noise. Noise floor was also quite low. I changed the sockets because some pins were loose and not 100% reliable when swapping tubes. I put in new belton sockets. I also suspect the noise originates in the PI area. And a bad connection sounds probable. I have a hard time locating it though. I can check voltages again and report back. Thanks again.

[mresistor]

Could you please post an up closer view of the power supply end of the chassis and then the tube sockets. It would be helpful. I am concerned about the artificial center tap resistors measuring roughly half of 100 ohms. Are they visibly burned?  If so there has been too much current flow through them indicating a short on the filaming string. Also two 100 ohm resistors in parallel is 50 ohms. Are you sure you measure them correctly?

[sluckey]

Could it be an issue with the artificial center tap on the filaments? I used 100 Ohm resistors, but when I measured them now they read about 52 ohms each?

Not likely. Checking those 100Ω resistors in circuit should show about 50Ω since they are in parallel through the very low resistance of the filament winding.

[Erik Aa]

One of the 100 Ohm 1/2 watt center tap resistors were actually burned. Hard to spot. I have removed the white filament wires and will start over. Is it ok to use 1 watt resistors on the center tap?

[bmccowan]

I have removed the white filament wires and will start over.

Did you take those voltage readings? What leads you to believe those heater wires are the problem?

[sluckey]

One of the 100 Ohm 1/2 watt center tap resistors were actually burned. Hard to spot. I have removed the white filament wires and will start over. Is it ok to use 1 watt resistors on the center tap?

Those 100Ω resistors only dissipate 0.1W under normal operation. If one of those resistors is really burned they both should be burned and you need to investigate why. The most common way to burn those resistors is to have a short between pin 2 and 3 on a 6L6 socket. But your symptoms don't support that idea. Also, the fact that the resistors measure 52Ω in circuit indicates they are fine. I would not use 1 watt resistors. If so, you risk losing an OT and/or PT if the 6L6 ever develops a short between pins 2 and 3.

[Erik Aa]

I completely rewired the white filament wirings, I replaced 1/2 watt center tap resistors, I put in new screen resistors, and went over all solder joints on the tube sockets. There are no shorts either that I can detect with my meter. The swoshing, popping, crackling is still there. This is even with the preamp tubes and pi pulled. I am at a complete loss. I don’t have a scope. Any advice and/or suggestions are indeed very welcome. Thanks!

[bmccowan]

You already asked for advice and received it. And then you ignored it and starting replacing things. I do not get it. You need to troubleshoot the amp, following an established troubleshooting procedure.

[Erik Aa]

Bmccowan, here are some voltage measurements with no input signal. The 6l6s are installed. The 12ax7s are not because it makes a terrible racket. I had already removed the filament wirings when you suggested I don’t start removing things…

All heater voltages are about 6,5 volts ac. 3,2 v on pin to ground.

6l6 with tubes:
Pin 3: 485 v dc, 4,5 v ac
Pin 4: 489 v Dc, 0,15 v ac
Pin 5: - 48 v and - 40 v dc. 0,1 v ac.
Pin 8: 0 v.

PI without tubes:
Pin 1: 484 volt dc
Pin 2: 0,2 v ac, 2,5v dc(!)
Pin 3: 0,1 v dc, 0,02v ac
Pin 6: 485 dc, 0,3 v ac
Pin 7: 0,01 v ac, 0,7 v dc
Pin 8: 0 v

The other 12ax7s are similar.

Thanks!

[bmccowan]

If you built an AA864 - intended voltages are on the schematic and layout I believe. From a quick look I would say that you have issues with the B+ supply to your preamp tubes. I'll be offline, but I'm sure others can help. But I suggest that you carefully compare your build with the schematic. Best to use a printout and a highlighter and carefully follow the circuits.

[Erik Aa]

I thought maybe the dc on the grid of the pi might have something to do with the noise issue?

[bmccowan]

If your little table is right - you have 484 volts on the plates of the preamp tubes!! That's part of the reason I suggest you methodically go through your circuit. Doug has a procedure for doing that on his pages here.

[pdf64]

Pin 5: - 48 v and - 40 v dc. 0,1 v ac.
Pin 2: 0,2 v ac, 2,5v dc(!)
…
Pin 7: 0,01 v ac, 0,7 v dc
…

So this is ghe schematic? https://el34world.com/charts/Schematics/files/Fender/Fender_bassman_aa864_schematic.pdf

The pin 5 V DC should be the same, if not find out why, eg leaky caps, leaky board, bad valves, bad measurements.

The V DC on the V grids is weird. Find out where it’s coming from.

[Erik Aa]

Thanks, I will check more. The plate voltages on the 12ax7s were measured with the tubes pulled as mentioned. I measured the voltages with tubes and the plate voltages on the preamp tubes are about 290 volts. Pin 1 on the pi is about 315 and pin 6 is 460. The pi voltages are weird.

[sluckey]

Pin 1 on the pi is about 315 and pin 6 is 460. The pi voltages are weird.

Insure there is a solid connection on the jumper between pin 3 and pin 8. Insure there is a solid connection on the jumper between pin 4 and pin 5.

[Erik Aa]

Thanks Sluckey. The 3-8 and 4-5 connections are good. I measured voltages on the 6l6 plates with the 12ax7s pulled and no signal into the amp. I have 1,8- 1,9 volts AC fluctuating on both plates. Actually, this AC voltage is also present on the b+ coming out of the full wave bridge rectifier. Could this be part of the problem? Thanks again.

Also, the 500pf cap going to pin 2 on the pi has about 2,4-2,7 volts dc on it fluctuating between these values. Could this be part of the problem? The cap is brand new…I really appreciate your help guys.

[sluckey]

Actually, this AC voltage is also present on the b+ coming out of the full wave bridge rectifier. Could this be part of the problem?

Not a problem. It's common to measure several volts of AC ripple at the first filter cap node.

Also, the 500pf cap going to pin 2 on the pi has about 2,4-2,7 volts dc on it fluctuating between these values. Could this be part of the problem?

There should be about +70VDC on pin 2, but your meter loads it down and will probably show about 40 to 50 volts.

[Erik Aa]

I did a lot more chop sticking. Indeed it seems the noise varies with chop sticking on the wires to the plates of the phase inverter (pin 1 and 6). The soldering on the pins look good and I measure continuity from the pins to the respective turrets. But there is still something going on there. Is this a typical issue? What do you suggest? Replace the wiring from pin 1 and 6 to the turrets?  Thanks!

[sluckey]

Something went wrong when you replaced the sockets. So start by examining all the connections to the PI socket.

[Erik Aa]

Yes, Sluckey. This is indeed the case. Something went wrong. But I am at a loss what it could be. I have reflowed all connections to no avail. The wires to the pins (except filaments) are stranded so I don’t suspect broken wires. I have measured all the new belton sockets as well and there are no apparent issues with them (so shorts). Could it be bad soldering? Although I measure connectivity to all pins?

[sluckey]

Post some hi-rez pics of the PI socket and the PI circuit on the board. Need to see lots of detail. Not like that itty-bitty 640x480 pic you posted earlier.

Also tell us about any deviations from the stock schematic. The FWB is obvious. Other changes may be less obvious.

And do one more set of voltage readings for every pin on the PI as referenced to chassis. Also measure the AC voltage ***BETWEEN*** pins 4/5 and 9.

[72Blazer]

Maybe its my eyes, but in that picture you posted earlier it is hard to to tell if you wired the filaments on V4?

Vr
J

[72Blazer]

Are the heater wires even connected to the PI tube (v4)? 

See the pic posted in reply #5

[Erik Aa]

72Blazer, I am traveling now but will post more pictures and measurements as suggested by Sluckey later this week.

I am a bassist, so in my aa864 build I only built the bass channel. I omitted the normal channel. So I have two preamp tubes and the pi. The two tube sockets to the left in the picture are 6l6 sockets. The preamp and pi have white filament wires running from the second 6l6 socket. The filament wires between the 6l6 sockets are the original yellow wires from the gutted peavey Classic. Thanks.

[plexi50]

I would guess you have one of the new sockets wired wrong. You can look at the amp 100 times and not see anything different. Then all of a sudden , there it is right in front of you the whole time.

[Erik Aa]

Here are some upclose images of pi pins and circuit. I also made some more pi measurements as suggested by Sluckey.

The circuit has omitted the normal channel as well as one gain stage on the first preamp tube. Worked well before the socket switch.

The pi voltages are measured with pi tube installed and no input signal. 

Pin 1: 1,0-20 volts AC fluctuating wildly (!) and 464 v dc.

Pin 2: 0,1-0,2 v AC and 40 v dc.

Pin 3: 0,1-0,2 v AC and 70 v dc.

Pin 4: 3,2 AC to ground, 6,4 v between 4 and 9.

Pin 5: 3,2 AC to ground, 6,4 v between 5 and 9.

Pin 6: 0,5-18 v AC fluctuating and 298 dc

Pin 7: 0,01 AC and 30-400 v dc fluctuating wildly.

Pin 8: 0,1-0,2 v AC and 70 v dc.

These voltages, in particular the fluctuations look crazy and weird.

Also, when I try to insert a signal, I cannot hear it. There is only the noise. Bad tubes? Grounding problem (I cannot find any)? Other???

Thanks!!!

[dwinstonwood]

Am I reading the pins wrong?

Are 5 and 6 jumpered?

Is one side of the heater string connected to 3? Do you have the socket wired for series 12.6V using the green wire?

I apologize if I'm seeing it all wrong.

[Erik Aa]

Dwinstonwood, I believe the yellow wire is connected to pin 1? Or? The green wire is the connection between 3 and 8. That’s the intention at least.

Please see more detailed pictures.  Thanks.

[dwinstonwood]

Got it! Thanks. 

[pdf64]

Try probing the socket from outside the chassis.
Leave a probe tip touching the pin 9 female contact; is there 6.3V AC between 9 and both 4 and 5?

[Erik Aa]

Pdf64, with preamp tubes pulled and measuring from the outside, both preamp tubes and pi measure 6,55 v between pin 9 and pin 4 as well as 5 respectively. Thanks.

[pdf64]

So again, from outside the chassis, with the amp deenergised and isolated from the mains wall voltage, try measuring the resistance of female socket contacts 2,3,7,8 to chassis common.

[bmccowan]

You've got some smart folks helping you here - so not to interrupt; but in looking at your pics of the board, I suggest you clean up the board before putting a lot more time into chasing gremlins. There are stray component leads and such. Makes it hard to find problems and even after you do, something else is likely to bite you in the butt. 

[Erik Aa]

Pdf64, here are the resistances from socket pins on the outside to chassis.

Socket 1.
Pin 2: 5,08kOhm
Pin 3: 1,48kOhm
Pin 7: O.l (not used)
Pin 8: O.l (not used)

Socket 2.
Pin 2: O.l (not used)
Pin 3: O.l (not used)
Pin 7: 6,3Ohm
Pin 8: 1,488kOhm

PI.
Pin 2: 0,888mOhm
Pin 3: 22,41kOhm
Pin 7: 1,030mOhm
Pin 8: 22,39kOhm

Bmccowan, thanks for the good advice. I try to make a clean build. This is my second build so far. There are 4 unconnected wires in the chassis. 3 wires are the wires that I disconnected from the second gain stage on the first preamp tube and shrink wrapped. The 4th wire is the negative feedback loop wire that I have not connected and instead shrink wrapped for now. I have 1 component with stray component leads. This is a 500pf cap that I replaced as a test. I will of course clean up these leads. Replacing the cap made no difference.

Thanks again! I really appreciate your advice and assistance guys!

[sluckey]

Pins 7 and 8 certainly are used on socket 1.

PI pins 2 and 7 should read the same resistance and that should be about 1M plus 22K.

[Erik Aa]

Thanks Sluckey. I don’t know why pins 2 and 7 have different resistances on the pi. Bad soldering job?

I don’t use pins 7 and 8 on socket 1 because I omitted the second gain stage to reduce distortion. Worked well before socket switch.

Btw, did you see the pi voltage measurements and photos I posted earlier today as per your suggestion?

Thanks!

[Erik Aa]

Could the problem be the new Belton socket itself? I am considering replacing the pi belton socket with one of the old sockets to see whether the noise goes away. To do a proper test maybe I should replace the preamp sockets as well?

[pdf64]

I don’t know why pins 2 and 7 have different resistances on the pi. Bad soldering job?

I suppose that’s possible. But did you leave the meter long enough for the reading to settle to its final value?
As there’s massively different RC time constants on each grid.

Could the problem be the new Belton socket itself? … maybe I should replace the preamp sockets as well?

The socket could be the problem. Can you see the heater in both triode section glowing orange?

There’s a massive clear issue with the LTP phase splitter circuit, ie one triode is not conducting. It seems a bad idea to risk causing more problems by messing with other areas.

[Erik Aa]

I see two vertical heater wires glowing inside the pi tube.

I hold the meter in place for several seconds. The resistance appears to change a little when moving the probe in the female sockets.

[pdf64]

I see two vertical heater wires glowing inside the pi tube.

I hold the meter in place for several seconds. The resistance appears to change a little when moving the probe in the female sockets.

So you’re sure each triode section inside the valve envelope is getting heated? Photo if unsure?

So probing the PI socket from outside the chassis.
Leave the black probe tip touching the chassis metalwork. With the red probe tip, touch each female contact and note the V DC reading.
I’m expecting 1 and 6 to be high, and the others to be 0V, but it’s worth checking.

[Erik Aa]

IT WAS THE SOCKETS! Dang! I ripped out the new Belton sockets and reinstalled some old sockets from 60s radios I have gutted and now there is NO noise. Wow! What a wonderful feeling. I really started to doubt myself the last week. All the hours looking for an error I made and it was probably the sockets all along. I cannot see or understand what was wrong with the Beltons, but now the amp works again. Puh! Important learning lesson for me indeed…

[sluckey]

Glad you got it fixed but don't be so quick to blame the Belton socket. Could be that whatever error happened (like collateral damage) when you originally changed the sockets was corrected when you changed the socket again. I've never experienced or even heard of a new Belton socket that was bad.

[Erik Aa]

Indeed, you are right Sluckey. I cannot blame the Beltons per se. I don’t know what the fault was. But at least it was not a wiring error. I have wired the sockets the same all along. Maybe the filament wire is not as tightly twisted close to the sockets now as they were on the Beltons. That is the only difference. Thanks.