Fender Model 30 Thermal Issue

[RadioComm]

Had this amp for years. Always had the same problem. Can play on it for about an hour, then I will get loss of volume with distortion. Happens on both channels, so pretty much sure it's in the power amp section. Just lived with it over the years, turning it off for a couple of minutes, then turning it back on which gives me another hour of play time, or I just switch amps. Here is some of the maintenance I've done over the years, but none of it fixed the problem:
    * Put new power tube sockets, and new resistors on the socket
    * New power tubes (V7/V8), biased
    * New phase inverter (12AT7 V6)
    * New filter caps

Recently, I decided to trace the signal. Plugged my old heathkit signal generator, perfect tone coming out of the speaker. Adjusted the oscilloscope and got a perfect sine wave post coupling cap (1000pf) at the junction of the coupling cap and the 1M resistor.

Putzed around the basement for a little over an hour, then heard the tone degrade (loss of power+distortion). Grabbed the oscilloscope probe, and the very instant that lead touched the solder point at that very same junction, there was an inaudible snap and the tone was restored with once again a perfect sine wave. Waited around another 20 minutes or so to see if the tone started degrading again, but no. Got impatient, didn't want to wait around another hour, so put the amp back together again.

I can't imagine there would be enough metal on that probe to act as a heat sink, but just in case, before putting the amp back together, I took out all the solder from that junction and replaced with new, but to no avail.

Not to sure how to proceed from here...?

Thanks

Just found a 1000pf cap in one of my shoe boxes, will replace it. Maybe I'll get lucky. Thanks again.

[tubeswell]

perfect sine wave post coupling cap (1000pf) at the junction of the coupling cap and the 1M resistor…
Just found a 1000pf cap in one of my shoe boxes, will replace it.

You mean the coupling cap going into the LTP? So, what’s it like after you replaced it?

[RadioComm]

Yep. the one going into v6A.  Wanted to start troubleshooting at the beginning of the signal flow into the power amp. But it fixed itself with the help of the probe... (lol) Anyway, just changed the cap now. Smoked tested it for a couple of minutes, all is still fine. Am breaking for lunch though. Will play this afternoon for an hour+ to see if I got lucky. Will give update then. Thanks and take care.

[RadioComm]

And nope. It was a shot in the dark. Played my usual hour, and then goodbye. So back to don't know where to go from here...  I did check the voltage points a couple of years ago, and all were within spec. But then again, I was playing guitar with it on the work bench until it started its thing and didn't have the tone while probing. So as far as I'm concerned, once I started probing, things might have gone back to normal. Don't know... Any thought or ideas would greatly be appreciated. Thanks.

[ac427v]

I wonder if it is a high frequency oscillation. Does it have any mods? If someone blackfaced the PI or power section, the amp may not be happy.Is this the schematic?

[RadioComm]

Thanks for the reply ac427v. No mods what so ever. I've had this amp forever. That is the schematic. Same one I posted with the pictures. Don't know anything about high frequency oscillation... Let look me it up, and see if those are the symptoms I'm seeing. Thanks again.

[ac427v]

I see your original post has a link to the schematic. I missed that. Sorry.

Looking at the photos, I do not see the 2000pf caps from pin 5 to ground on the power tubes. Can that be the culprit?

[Latole]

It is not a power supply issue ?
I would monitor the B+ with a voltmeter permanently connected. 413 volts on schematic

[octal]

I say connect the scope probe to the input to the LTP, then power up and start playing. Once you separate whether it's a preamp or poweramp issue, you can then divide and conquer further. 

[tubeswell]

Look for a heat sensitive part near a heat source


Resistance increases with heat, so maybe a flakey screen grid resistor or a power supply dropping resistor or a bias supply resistor. When the volume drops, what do all the PS and bias voltages look like?

[RadioComm]

Thanks for all your replies.
@tubeswell
I initially thought too a sensitive part near heat source, hence the new ceramic pa sockets and resistors. What really stunned (baffled) me though was the fact that the very instant that oscilloscope probe hit the junction into the input of the LTP (V6A), sound was restored. So even if the probe acted as a sink, that solder point would be too far from the screen grids and such. I'm thinking more along the line of a capacitor, that after a certain amount of time, is  just not being able to retain its capacitance. I've already measured the voltages at the points specified on the schematic, and they all checked out, but... The probing with the multi-meter might have "fixed" the problem, I don't know. I'm going to hook up the signal generator and then probe for voltages, this time a tone going through. If the probing "fixes" the problem again, I'll figure out how to proceed from there. Thanks.

[tubeswell]

When the volume drops, what do all the PS and bias voltages look like?

[Latole]

When the volume drops, what do all the PS and bias voltages look like?

That is what I ask 2 days ago, no answer!

[RadioComm]

When the volume drops, what do all the PS and bias voltages look like?

That is what I ask 2 days ago, no answer!

I've already measured the voltages at the points specified on the schematic, and they all checked out, but... The probing with the multi-meter might have "fixed" the problem, I don't know. I'm going to hook up the signal generator and then probe for voltages, this time a tone going through. If the probing "fixes" the problem again, I'll figure out how to proceed from there. Thanks.

This includes the PS, B+, and the bias voltage. Basically every voltage indicated on the schematic. But as I stated before, if the same thing happened while getting these voltages as with the oscilloscope probe, then I just measured a perfectly functional amp. Hence I'm going to redo the measurements but with a tone, so that way I can validate whether I got those measurements while the signal is still degraded. Hopefully I'll be able to get to it this week-end. I'll update the post then. Thanks again.

[Latole]

You must read voltage with voltmeter.

1- When amp play fine; you wrote you do and voltage are like schematic
And
2- When amp get loss of volume with distortion. Report
 

[octal]

I think the normal pop you get when you connect a probe is temporarily fixing things. (it creates a large signal which arcs through a bad connection or a bad component.) That's why I suggested connecting the scope before you power up and start running a signal through it. You could do the same thing with the meter. (Or if you own multiple meters, just clip 'em on a bunch of points.)

[RadioComm]

I think the normal pop you get when you connect a probe is temporarily fixing things. (it creates a large signal which arcs through a bad connection or a bad component.) That's why I suggested connecting the scope before you power up and start running a signal through it. You could do the same thing with the meter. (Or if you own multiple meters, just clip 'em on a bunch of points.)

Makes perfect sense. Thank you!
I'll hook up the signal generator to get tone (save me an hour of guitar playing which I can do on another amp that works (lol)). I'll hook up the oscilloscope on the input of the LPT just in case it is in the preamp (see note). I believe I have three multi-meters. I'll hook one up on the bias output (-46v on the schematic), and the other two on the output of V6A and V6B (+228V & +213V respectively on schematic). Since I've already recorded these voltages for the functioning amp, I know what the "error margin" is. Thanks again @octal for the diagnostic and the input. I like.

Note: This amp has a switching IC (syl-ecg-40938). There is stock switch up front which enables the ic, which enables the channel switching from the footswitch. This front switch can also disable the IC, which then brings the fender to the old school two channels (no switching). With this switch disabling the IC, I moved the jack from one channel to another and still had the degraded signal. So don't believe it's pre-amp, but agreed, cover all grounds, hence the oscilloscope probing the input of the LPT (V6A). Thanks again @octal.

[SILVERGUN]

I like throwing darts at stuff.
I haven't heard anyone mention the heater circuit so I'll do it.
There is a very slim chance that you have a turn to turn short in the heater supply winding. I haven't heard much of this in amp work but we used to see it all the time in old welding machines.
That's my $.02 guess.
Does the indicator lamp dim or flicker?
Monitor the heater voltage and just cross that off the list of potential problems.

[SILVERGUN]

Another dart:
That board looks a little juicy.
Maybe it's conductive but not obvious until you add heat. 

[RadioComm]

I like throwing darts at stuff.
I haven't heard anyone mention the heater circuit so I'll do it.
There is a very slim chance that you have a turn to turn short in the heater supply winding. I haven't heard much of this in amp work but we used to see it all the time in old welding machines.
That's my $.02 guess.
Does the indicator lamp dim or flicker?
Monitor the heater voltage and just cross that off the list of potential problems.

Another dart:
That board looks a little juicy.
Maybe it's conductive but not obvious until you add heat. 

Absolutely. Thanks. No stone left un-turned. I'll first visually inspect the windings. Occasionally it has happened that when soldering in tight places I've brushed wires with the iron... No lamp dim or flickering, but will also check voltage at end of winding. This amp has a hum balance though, which has nothing to do with the silverface hum balance. You'll see it on the schematic on the output of the PS 6.3V tap to the pilot light. Not sure how it affects the heaters... Any idea?

Juicy?... Not too sure what that means. To add heat... Is it safe to maybe use a hair dryer or hot air gun? Maybe not the hot-air gun... But could direct it on specific components....

@latole, haven't been too successful about researching symptoms and fixes of high frequency oscillation. Also, don't believe there is a missing 2000pf cap. Once again all is stock. This said, I know there's at least one of the two on the schematic, once I have it up on the bench, I'll look for the second one.

Thanks again for everyone's input.

[SILVERGUN]

I'll first visually inspect the windings.

Not sure you'll be able to see anything. The true tell-tale sign of failure will be if the heater voltage is dropping out after the amp warms up. Just wait for it to happen and measure the heater voltage.

Juicy?... Not too sure what that means. To add heat... Is it safe to maybe use a hair dryer or hot air gun? Maybe not the hot-air gun... But could direct it on specific components....

Sorry....the board looks like it has some moisture content. I have heard of people trying to bake the moisture out of a conductive board.
First, you'd have to confirm it. Clip one meter lead to ground and press the other lead tip into the board at different locations to see if you get a DCV reading off of the board material itself.

[RadioComm]

Got it all hooked up. Signal generator into reverb channel input jack. Oscilloscope probing at 1000pf/1M junction into V6A (12AT7) input. Nice sine wave. One MM on pin 4 of V7 (mislabeled V6A on schematic) ~384VDC, and another MM on pin 4 of V8 ~388VDC, and the last MM on output of bias pot ~45.8VDC. I let rip for 2 hours + this afternoon, but signal never degraded. Voltages and tone stayed put. Unfortunately ran out of time. This is well over an hour more than what I usually get. Will try again another day, but get an earlier start in the day. One thing different, is that I adjusted the hum balance a bit, but don't believe that had an impact on the testing. I imaging it's the MM's alligator clips that probably acted as heat sinks. lol, maybe that's solution, leave the alligator clips and put it back together... (just kidding)

Hope everyone had a great week-end!

[ac427v]

The schematic shows 2 2000pf caps. One on pin5 of each 6L6 power tube.

They do not show in your picture.

Supposedly Fender put them there to correct for oscillation caused by CBS changed wire routing. Internet lore says the caps are only needed on some of these amps-others have better lead dress. It is common for people to remove them hoping to create a more authentic fender sound. I suspect that has been done to your "stock" Fender build. and that your amp does need them to prevent the exact problem you are having.

[RadioComm]

The schematic shows 2 2000pf caps. One on pin5 of each 6L6 power tube.

They do not show in your picture.

Supposedly Fender put them there to correct for oscillation caused by CBS changed wire routing. Internet lore says the caps are only needed on some of these amps-others have better lead dress. It is common for people to remove them hoping to create a more authentic fender sound. I suspect that has been done to your "stock" Fender build. and that your amp does need them to prevent the exact problem you are having.

@ac427v. Once again, thanks for your responses. I do understand what you're talking about, but I've had this amp for well over 20 years, and it's only in the last 10 that this issue has been a problem. So I'm absolutely sure that the missing caps is not due to anyone's mods. This said, as you observed the Fender 30 schematic shows two 2000pf caps going from pin 5 of 6l6's to ground. The Fender 30 appears to be a "loosely" souped up Pro Reverb AA270, so I'm using it's layout diagram (aa270 layout attached) as a reference. The aa270 layout shows the grounded 2 caps, which the Fender 30 doesn't have. But... The Fender 30 has a cap between the two .1 caps on the board which the aa270 layout does not have.This unknown cap is not in the Fender 30 schematic either. One lead of this unknown cap is going to pin 1 of one 6l6, and the other lead goes to pin 1 of the other 6l6. I've attached two pictures to show the unknown cap and its position. It's quite common in industry that a schematic leaves the engineering dept to head to manufacturing, and then for production purposes the circuit gets modified but the schematic never gets updated. I'm not saying this is not a high frequency oscillation problem, just saying is not due to the caps that are shown in the schematic but not in the construction.

Taking the pictures, I did notice a pretty nasty looking 470 ohm 1w resistor though (picture attached). Don't see how I missed it. Did replace the sockets and resistors a few years ago, but maybe lightning struck twice. It can happen. Will replace resistor and test.

Take care everyone.

[pdf64]

Those LTP anode to anode caps aren’t used on any other other models that have a 12AT7 LTP.
The 12AT7 has a far lower anode impedance than the 12AX7 those caps are usually used with.

[ac427v]

Reading more reviews of this amp makes me want one. That does not distort and cut out after an hour. Here's my take:
The cracked 470 ohm resistor is not the problem but should be replaced.
The problem is an ultrasonic oscillation which increases with time. This is from poor lead dress in the original design, not a thermal-caused. You showed that when you instantly "fixed" the problem with your probe. The schematic shows that fender fixed the problem by draining ultrasonic signal to ground with those missing 2000pf caps. Someone (factory? or previous owner?) decided to change that plan by removing the caps and replacing with one anode-to-anode cap as seen in the tweed bassman circuit. I base that on your original post "Had this amp for years. Always had the same problem."

If your amp was on my bench, I would tack-solder in those two missing 2000pf caps and test it. If that fixed it, I would remove the suspect anode-to-anode cap. Then I would try reducing the value of the 2000pf caps until the oscillation came back. My goal would be to end up with the smallest value caps that still do the job. That would minimize their impact on desirable audible high frequency tones.

[RadioComm]

Reading more reviews of this amp makes me want one. That does not distort and cut out after an hour. Here's my take:
The cracked 470 ohm resistor is not the problem but should be replaced.
The problem is an ultrasonic oscillation which increases with time. This is from poor lead dress in the original design, not a thermal-caused. You showed that when you instantly "fixed" the problem with your probe. The schematic shows that fender fixed the problem by draining ultrasonic signal to ground with those missing 2000pf caps. Someone (factory? or previous owner?) decided to change that plan by removing the caps and replacing with one anode-to-anode cap as seen in the tweed bassman circuit. I base that on your original post "Had this amp for years. Always had the same problem."

If your amp was on my bench, I would tack-solder in those two missing 2000pf caps and test it. If that fixed it, I would remove the suspect anode-to-anode cap. Then I would try reducing the value of the 2000pf caps until the oscillation came back. My goal would be to end up with the smallest value caps that still do the job. That would minimize their impact on desirable audible high frequency tones.

@ac427v Makes sense. And certainly is easy to test and undo if necessary. Just ordered a cap kit from amazon:
https://www.amazon.com/gp/product/B0714C4PRT/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1
Will be getting it this Saturday.
Will start with the 2200pf 2kv for proof of concept, and then go from there. As you so well mentioned, I'll let my ears navigate me though this one. If this solves the issue:
1) remove a2a cap - Play and listen
2) if too many high frequencies cut - goto 1500pf - Play and listen
3) how low should I go?
4) bottom line - tinker till I find what I like following your approach...  :-)
5) If worst comes to worst tone-wise, undo the changes, and I'll just go back to my old MO. Play an hour and break for 5 minutes or switch amps. I only play with my buddies anyway in informal settings. So if the amp craps out in the middle of a song, oh well. Just love its current tone too much.

And if the 2200pf / a2a cap proof of concept doesn't solve the problem, skip steps 2-5 and well... I don't know
I do agree with you, I really don't believe its a thermal issue. I did think so until I touched that junction with the probe. As I stated earlier, no way that probe could have acted as a heat sink. And that amp is not getting that hot anyway. Even more so opened up on that bench.
Hopefully I'll get (make) time this week-end to work on it. Either way I'll provide an update. Thanks again ac427v for your help.

[RadioComm]

Had some free time this morning, and decided while I'm waiting for the caps, might as well change the 470 ohm resistor. Replaced the cap and started playing just for yucks, and played for well over two hours. First time in 10 years. *Really* don't think that it was the replaced resistor. I fully agree, not a thermal problem and not the resistor. So since the last probe was working solid for over two hours too, I started thinking what did I do that would make the difference. The only two things I did was adjusting the hum balance on the heater wires, and tied up the switching wire close the pre-amp/power amp coupling cap so that I could attach the oscilloscope. Definitely not the heater wire adjustment so started looking at the switching wire near the coupling cap. I noticed a kink in the wire where it used to lay on the coupling cap lead. Maybe the kink was caused by 41 years of rubbing/heat. But that's the only thing different that was done prior to the last probe.

So I'm thinking maybe some kind of interaction between the switching wire and the coupling cap (inductance of some type or even contact). And this was right at the junction where the first probe "fixed" the issue... Any thoughts?
I've attached a couple of pictures. The first shows where the switching wires come on from underneath the chassis along with the two original ties + the one that I added. The second shows the probed junction post coupling cap, along with the kinked wire. Thanks.

[sluckey]

That melted insulation on the shielded cable is suspect. Use your needle nose pliers to gently squeeze and flex that cable directly on the melted spot. Then I would cut the black insulation about 1/4" to 1/2" farther back from the end and then trim all the individual strands of shielding. This would remove any possible intermittent short between the shield and center conductor.

[RadioComm]

That melted insulation on the shielded cable is suspect. Use your needle nose pliers to gently squeeze and flex that cable directly on the melted spot. Then I would cut the black insulation about 1/4" to 1/2" farther back from the end and then trim all the individual strands of shielding. This would remove any possible intermittent short between the shield and center conductor.

Will do. Definitely a good idea. Will add further protection with some shrink tubing. Thanks sluckey.