Marshall 8080.... hot resistors

[Garydean]

I've got a Marshall 8080 combo I'm working on. It's a Valvestate type dating back to around 1992. The complaint was very noisy and low static type noise with no signal applied. I opened it up, pulled the board, cleaned pots, jacks and inspected the board for bad joints and burn marks...etc. I then turned it on to test for voltages. After about 90 seconds the 7w dropping resistors R112 and R113 got really hot along with R102 and R103 both 5w resistors. Hot to the point where they were almost too hot to touch. I never made it to the output transistors to test, as I had to shut it down. I tried to post a schematic. here's the link.

http://www.drtube.com/schematics/marshall/pc0689p.pdf

Everything else on the board was cool. How, at this point, would I be able to test the circuit without burning up those resistors? I have a variac. Will the readings say at 50vac be directly proportional to the readings at 120vac? Any thoughts from those who have encountered similar problems with this amp would be greatly appreciated. I'm not too familiar with OP-Amps but I have a basic knowledge.

Gary

[bbartilson]

How, at this point, would I be able to test the circuit without burning up those resistors?


Well, first you've got to verify that there's actually a problem with the power supply or not. If you're concerned about the heat the resistors are producing, verify that they are operating within their limits or not.

R101, 102, 112, and 113 are indeed dropping resistors. Two are for the +/- 15v supplies for the IC's. The others drop the raw AC off the power supply transformer secondary to provide a heater supply for the 12AX7.

Being that the tube wants 150ma or so for it's heaters, it's easy to calculate the power dissipation with Ohm's law. 150 mils through 180 ohms = 27volts dropped across each of those resistors. 27 volts times .150 = 4.05 watts.

Generally speaking you'll find that engineers design these sort of power supply resistors to be running at about half their rated power, so our guesstimate figure is reasonable.

By the same token, 4 watts is a LOT of power, and I would expect all of the resistors in question to get hot. That's their job.

How do you verify this is the case without burning them up? Simple. Measure the voltage across them, and use Ohm's law like I did to see if the dissipation is within reasonable bounds. It's possible that one of the zener's has failed, or some other such component is dragging one of the supplies down, and pulling more current through one of those resistors than it should. But if each pair is more or less equal to one another, and each is at half of it's rated power or less, I would say that you are troubleshooting something that isn't a problem at all.

Once you've verified that the supplies are okay (and that those resistors get surprisingly hot in normal operation) you can move on to finding out what is actually wrong.

As for troubleshooting the IC's:

Measure the DC voltages at the power supply pins, inputs, and outputs. The power supply pins should all be at +/- 15v (or something close). Except in very rare situations, there should be NO measurable DC voltage at any of the input or output pins of the IC's. I didn't look at the schematic of the preamp to verify, which the is the only reason I even mention it.

If there is DC on any one of the IC's audio pins, remove it. If it's in a socket, that's too simple. If not, unsolder it. Then verify that the IC was the CAUSE of the DC by measuring the voltage at that pin without the IC in place. You can safely power the thing without the IC in the circut, with the exception of the one in the schematic you posted. That's part of the power amp, and it might act funny without that part. In that case, merely replace it for testing.

Again, if the IC's are in sockets, you can swap like ones around and see if the problem moves with the particular IC as well.

As for your noise? You just gotta find it. Do what you can to isolate what part of the amp it's coming from. (before or after the effects loop if it has one, the volume turns it up or down, or it doesn't, it happens on both channels or it doesn't.

Then report back.

Best,
Bill

[52 Bill]

I'd expect those resistors to get hot, they're all dropping big voltages in there.

The power transformer is putting out maybe 60vac ct? I'm guessing that this is being dropped to 12vac by R112 and R113 for the preamp heater circuit.

The DC supply is probably around plus and minus 45v? R101 and R102 drop this down to plus and minus 15V, so all of that lost energy has to go somewhere. Heat.

I'd guess that this is normal, and if the board hasn't been charred or the solder joints loosened, test all of the dc voltages. Make sure that the tube is ok and that the 15 volt supplies are ok. Is the transformer getting warm/hot?

[Garydean]

WOW, thanks Bill and Bill, once again.

The PT is not even warm. What made me touch the resistors was the fact that I started to smell heat from R101 & R102, you know the smell. I may have measured a 30v+ drop each at R102 and R101. All kinda went too fast. I may have seen smoke or it could have been a floater...LOL.

The tube is fine I checked it in another amp after checking it with a tester.

The board around R112 & R113 is slightly discolored, not enough to suspect trouble there. I'm going to fire it up with 90vac from a variac just in case just to see if the voltages are odd. If I saw 40vdc on one side of the resistor with 120vac then I should see 30vdc with line voltage at 90vac? OR will the circuit not be fully operating until it sees 120vac?

thanks again guys, for your guidance,

Gary

[bbartilson]

I'm going to fire it up with 90vac from a variac just in case just to see if the voltages are odd. If I saw 40vdc on one side of the resistor with 120vac then I should see 30vdc with line voltage at 90vac? OR will the circuit not be fully operating until it sees 120vac?

Sure, more or less. The ratios work out. I don't see a problem with firing it up at line voltage for the time you'll need to make measurements.

I really don't think there's an actual problem with the resistors, which is why I gave you the long explanation regarding calculating their power dissipation.

Even if you go slow, it shouldn't take more than 30 seconds to measure the voltage across each of the four resistors in question.

Turn the thing on and measure each of them, then turn it off and do your calculations. Or cycle the power four times with time to cool off in between if it eases your mind.

There is the possibility that there is something wrong, and if it involves one of those resistors, it will likely read well differently from it's partner. If each pair measure about equal, and the voltage dropped across them calculates to a reasonable dissipation, they are just hot, and you can go about the actual troubleshooting.

Some discoloration of the circuit board is also expected and not at all unusual. Sometimes the heat leads to failures in solder and tracework later on, so it's worth a decent look, but with that amp's age, my guts tell me your hot resistors have nothing to do with the problem. It's easily verified with four quick measurements.

-B

[Garydean]

Thank you Bill.

I certainly appreciate your advice. I'll be in my shop tonight testing it. Results to follow.

Gary

[Garydean]

Last night I fired up the amp long enough to take voltage readings across the two sets of dropping resistors. I measured the watt dissipation to be slightly more than half of what the design calls for. I'm getting 48V on one side and 15.5 on the other side of each. The resistors just before the heaters were both 36vac dropping down to 6.8. Yet, resistors 101 and 102 started to get too warm. The board got yellow and I saw them starting to smoke, not sure which one.

Testing Zener's...... when I use my DMM on diode function, Zener diodes should read what? Mine were taking a charge and climbing past 1.5v. I haven't tested enough of these to get the hang of it. Bill, you mentioned that these may be bad. I should replace them anyway as they have been exposed to some high temps. Any 15v 3w Zener will work?

Thanks again,

Gary

[52 Bill]

If either R101 or 102 is overheating, I'd look for a short somewhere in the 15 volt busses. What happened to the plus and minus 15 volts when the thing started to overheat? Did either side start to drop?

The Zener should test almost like any silicon diode. Open in one direction and conducting in the other. Unless one is thermally reactive, if you got 15 volts in your test, the zeners are probably ok.

I'd also check C40 and 41 for ESR shorts.

[Garydean]

Thanks Bill.
I'll monitor the voltages again as the resistors begin to overheat.

I tested ZD1 and ZD2 and both are conducting in both directions. D1 and D2 are also conducting in both directions. I'm testing them in circuit though, should I take them out and test them. I also tested C40 & C41 with an ESR meter and they test fine.

When I test the output transistors, do I set the DMM on diode function and test across all legs? Should I take them out to test them? I tested them separately in circuit and found all three to be conducting in every combination.

Thanks,
G

[52 Bill]

I tested ZD1 and ZD2 and both are conducting in both directions. D1 and D2 are also conducting in both directions. I'm testing them in circuit though, should I take them out and test them.
The only way to know for sure is to test them out of circuit. Sometimes, but not always, you can get reliable in circuit results. If they conduct in both directions then pull one leg and retest.

Remember that as you test the zeners, your meter will be charging C40 and C41, and when you reverse the leads your meter will at first be reading the small charge on the caps. If you hold the leads in place and allow the caps to discharge the reading may change,

When I test the output transistors, do I set the DMM on diode function and test across all legs? Should I take them out to test them?
I think that those outputs are plastic cased Darlingtons(?), they will test like any other power transistor, but there will also be the equivalent of a diode connection between the emitter and the collector.

Place the red lead on the base contact (from the front, the leftmost lead). Take the second lead and touch it to the center lead which is the collector. Depending upon what the transisitor polarity is, and the polarity of your meter, you will get either a high or low reading. Now touch the second lead to the righthand lead, which is the emitter. The reading should be similar to the first b-c reading, meaning if the b-c reading was high, the b-e reading should be high as well. Now reverse the meter leads, and you should get the opposite readings. If the original readings from b-c and b-e were high, they should now be low.

Test across the collector and emitter terminals, this should test like a diode, high in one direction and low in the other.

With all of that being said, they are probably not the source of your problem, as they are supplied from the other side of the power supply resistors, not the 15 volt source.

[Garydean]

Great info Bill... I will apply my new knowledge this evening.

Thank you!!

Gary

[Garydean]

I tested the Zener's out of circuit and as you thought they are fine. I took more measurements after R101 $ R102 at full line voltage and saw the 15vdc get to 15.9vdc and -15.9vdc and shut it down as I saw smoke. The voltages before those resistors remained at 48.4vdc the whole time. So something is happening on the other side where 15v should be. C41 & C40 could be a problem even though they tested fine with an ESR meter? What should the vdc be coming out of the rectifier, it doesn't say on the schematic. I'm not sure where to look now. Thanks for your help.

Gary

[52 Bill]

What started to smoke?

Plus and minus 15.9 vdc isn't out of line there, neither is the plus and minus 48.4 vdc.
If one of the resisitors is smoking, it may be a bad resistor.

[Garydean]

The smoldering looked like it came from the resistor(s). It measured correct but I guess maybe it's lost it's 5w status. I ordered up a couple replacements today.



thanks you for the lead,

Gary

[bbartilson]

I'm not sure where to look now.

Please post your math for me. It's not that I disbelieve you, I just want to know what numbers you got and how you got them. In order to direct you further, there isn't any wiggle room - you've got to be VERY specific regarding the information you post.

For instance, there's a post which says you measured equal + and - supplies at 15.5v. Then you recently post saying that those voltages are climbing, and when they reached 15.9, you saw smoke and shut the thing off.

So which is it? 15.5 or 15.9? Over what period of time is it climbing, and are you sure they are climbing symmetrically? Or is one higher? This is key information.

In my original post, I directed you to find whether those supplies were equal or not. Components don't usually fail in pairs, and so it's MUCH more likely that one or the other of those supplies can help you find the real problem.

It IS possible that you simply have worn out resistors. They could heat up and increase in resistance over a period of a minute or so. As they heat up and resistance increases, the supplies would climb. But as I said, it would be highly unusual for both to fail in the same way at the same time, so that is a red flag to me.

While the smoking resistor is an interesting symptom, it may or may not be indicative of the problem. If you see blood dripping on the floor, you might be curious as to where its coming from, until you notice the gunshot wound on the guy from whom the blood is dripping. Don't let the smoke ruin your troubleshooting. It's a clue, but it's easily eliminated by the instructions I gave you before. To reiterate:

1. Measure each of the resistors in question. Do they measure within tolerance? One or both may actually be bad. Just because it's unlikely does not mean it's impossible. They measure close to tolerance, or they are very different from one another, or both are high. Report your finding.

2. Measure the voltage ACROSS the resistors, not from chassis to either point. Report your finding.

3. The most likely failure still remains with one of the IC's. We will eliminate those before we go further. For sure it's something tied to the supply that's smoking. Are the IC's in sockets? If not, take a moment and put your finger on each after powering the thing. WATCH OUT! If they are bad, they can get quite hot. If you find a hot one, remove it, re-power and see if the smoking problem goes away.

Finish each step in order before moving on. If the resistors measure funny, it may still be as a result of a failed IC. Which is why I haven't suggested replacing the resistors as a matter of course. I'd much rather you burnt up the old one (or ones) in the process of finding the actual problem. From the sound of it, they are suspect and ought to be replaced anyway - but not until you've found the reason for the excess current draw, or confirmed that the resistors are the actual, and only, problem.

-B

[bbartilson]

I just happened to think:

With the power off, measure resistance from each of those supplies to ground. Likely as not you'll get a climbing value as you charge the filter caps. Go ahead and let it climb.

If something is dragging down one or both of those supplies, it SHOULD show up as a low resistance. If it is one of the IC's it's possible that it doesn't exhibit the symptom unless it's under power. But there's a chance. Ohms to ground from each supply.

Thanks,
-B

[Garydean]

Thanks B,

I got a bit brave this morning after reading your post and left the amp on for almost 10 minutes....figuring, it it burns...whateva. (east coast thing).

EDIT:......:: does it matter if the reverb tank is unplugged or not??:::.....

I measured

R101 out of circuit...... 334.72 ohms
R102 out of circuit...... 340.66 ohms

101 resistance to ground 5.95M
102 resistance to ground 6.0M

R101 VDC across -32
R102 VDC across 32

at 2 minutes full line AC

R101... -16.1v
R102... 16.1v

at 4 minutes

R101... -16.2v
R102... 15.9v

at 8 minutes

R101... -16.3v
R102... 15.9v

IC9 warm
IC10 med/hot

The very slight smoking stopped after 2 minutes.

thank you for your help,

Gary

[52 Bill]

Did you power up the amp and listen for the problem before doing anything?

Is it possible that something got spilled on the smoking resistor when you sprayed out the pots etc.?

I would investigate IC9 and 10, I don't have the schematic, but probably should not be getting hot.

[Garydean]

I won't rule out the spill theory even though the spraying was very contained.
I haven't run a signal through it yet. I opened it up to do a visual first and take some readings and to see how the board looked (that was my first encounter with the smoke). If that checked out I was then going to turn it on and listen for problems.

The complaint was "crackling" and "strange noises".

Looking at IC9 and IC10 I see the board area has yellowed underneath from heat. Do these go bad on their own or do they start to fail when another component looses it's original value? I'm going to install a few sockets then pop in the new op amps. Short of lifting a leg from every nearby component, how do I narrow it down?

thank you for the education,

Gary

[52 Bill]

I'm not sure what those IC's do, but in general the fastest test is to read the voltages to ground at all pins.

What you are looking for is any voltage on the input or output pins, as this will show possible shorted chips. In a dual voltage power supply design, there will be one pin with +15v, one pin with -15v, and the rest should be near zero.

[bbartilson]

I opened it up to do a visual first and take some readings and to see how the board looked

The complaint was "crackling" and "strange noises".

Looking at IC9 and IC10 I see the board area has yellowed underneath from heat. Do these go bad on their own

Well done. Now we're getting somewhere.

As for "I opened it first." I'd consider this to be a mistake. It isn't necessarily, and I can't say that I never open something without having turned it on first. But, as a general rule, it's a very good idea to turn any unit on and see what kind of symptom it displays. When you run a shop that produces any volume of units, about 1 out of 10 don't have anything wrong with them.

Yes, the 'crackling' symptom is what made me think IC's. That's a common failure mode.

If IC's got hot enough to yellow the board, I wouldn't trust them. It's quite possible that they are interconnected, and whatever went wrong with one of them is causing the other to heat up. Regardless, I'd just replace them both and see what happens.

Make sure you look for DC on the signal pins of those IC's BEFORE you install the new ones. Yes, the IC's can and do fail all by themselves. Most of them that I see have some reason for having failed - the ones that connect with the outside (input, effects loop, etc) are the ones where someone can plug something creative into the hole and ruin the part at the other end of the jack. Simple static electricity can do it as well.

Whatever the case, replacing those resistors can't hurt if you already ordered them. Leaving the old ones is fine too. They aren't your problem.

Replace those IC's, and let us know of your success, or if there's more to troubleshoot. Either way is good with me. :)

Kudos for being brave enough to let it run a while.

Best,
Bill

[Garydean]

Thanks Bill and Bill for sticking with me on this one. I've ordered the new IC's, (I went to my local electronics store and they looked at me like I had 4 eyes when I asked for a TL072 Op Amp, or even an 8 pin dip socket). Mouser, if anyone wants to compare, sells them for around a half a buck each.

I did check for DC and nothing struck me as too high to write it down. IC3 was quite warm and IC7 read only 12.8v on pin 1. Everything else looked good. I ordered a complete set for this amp....5$, plus a bunch of spares for my next encounter. After I install the new parts I'll post the results.



Thanks again for your guidance,

Gary

[52 Bill]

IC3 was quite warm and IC7 read only 12.8v on pin 1.
Pin 1 on a TL072 is an output. Did you mean pin 8?

[Garydean]

Hi Bill,

IC3 is an M5201, common to Korg circuits. Looks like it's a switch in the distortion section just before the 12AX7. A 2120D replaces that, I've learned, (also $4.50 cheaper). I measured it in the clean mode. Does this get a few extra volts when the distortion button is activated?
My new Tl072's will be here Wednesday.

thanks for your interest.

Gary

[Garydean]

Finally the new IC's came in and I replaced them all, with sockets. There's something else wrong as IC9 and IC10 are still very hot. Pin one of both IC9 and IC10 have over 400mV on them. I'm running out of ideas.

Gary

[52 Bill]

OK, I downloaded the schematic of the pre-amp. IC9 and 10 are both part of the reverb drive and recovery circuit. They've paralled 3 opamps together to increase the drive current to the tank, this is where I'd look first. Is the tank and wiring ok? Check all of the components related to the drive circuit. Is it possible that C56 is shorted or leaky?

[Garydean]

While testing the circuit, is it imperative that the reverb tank be connected?

thank you.

[52 Bill]

I doubt it. I would think that without the tank the load on IC 9 and 10 would be lessened.

[Garydean]

Thanks Bill, I'm not at the amp this moment. But my notes indicate that it did check ok.. I'll be back at it later today and keep searching.

Thanks again,

Gary

[Garydean]

Going through every component in the reverb section, I find no excessive DC. The rails are at 16 and -16. No alarming DC either on the input or outputs of
IC9 and IC10. Except 340mV on pin one of IC10, which is the highest.

IC's are still hot

[bbartilson]

Going through every component in the reverb section, I find no excessive DC. The rails are at 16 and -16. No alarming DC either on the input or outputs of
IC9 and IC10. Except 340mV on pin one of IC10, which is the highest.

IC's are still hot

Okay, so after some searching I found a schematic that I can ALMOST read. Yes, it seems that we have three sections of opamp sharing the load to come up with enough current to excite the reverb pan.

For the most part, this reverb drive circuit is utterly simple. 100k resistor setting input impedance. 47 ohm and 4k7 voltage divider for a gain of 100. The bypass cap acts as a high pass filter to ground, increasing the effective distance of the low side of R107 from ground as frequency decreases, and thus reducing the gain of the amp at lower frequencies. It also means that any DC bias on the output pins wouldn't really matter, as there is no DC path to ground to form a circuit.

So yes, replacing (or at least measuring) C56 may yield a clue or fix the immediate problem. Were it leaky, a DC path to ground could cause the IC's to heat. But unless those IC's are bad, they shouldn't have any appreciable DC on the input or output pins. And 340mv is plenty appreciable in my book.

So with IC's 9 and 10 removed, is there DC in the input pins? If so, is it also on the output of IC6b? The coupling cap C44 may have failed short in this case. Really an unusual scenario, and not likely, but we need to know the DC measurements at the input and output pins of the reverb driver without the chips present. Then we'll know which way to go.

It's quite possible that you're still troubleshooting nothing.

Asking those little IC's to provide that kind of current is kind of extreme. It could well explain the yellowing of the board, and one being hotter than the other would simply be caused by the difference between one half and the whole of the IC being asked to drive extra current. (IC9 vs IC10)

After years in service, one of the old IC's could have failed and started to make crackle noises. You replace them, and though they are hot, just like the resistors, they end up not being a problem. Re-installing the old ones and getting the crackling back would be a good thing at this point.

With IC9 and IC10 installed, measure the DC on each of the output pins of those respective drivers. Report each reading at the IC, and also the reading at the other side of those 10 ohm resistors where they are tied together.

Like I said, this might be nothing at all, and normal operation. Especially if the reverb works normally, and the crackling is gone.

-Bill

[Garydean]

Thanks Bill,
C56 measures -1.5mV on one side and .000 on the ground side.

With IC9 and IC10 removed from the circuit.

C56 measures -220.mV and 0 on the ground side.
C44 -5mV on one side and 0 on the other.

IC9 and IC10 pulled
IC9
pin 1 -220mV
pin 3 0
pin 5 0
pin 7 -220mV

IC10
pin 1 -220mV
pin 3 0
pin 5 0
pin 7 4.8mV

Both IC's in
C56 -1.5mV and 0.0mv
R104 -163mV and -1.2mV
R105 -163mV and -1.2mV
R106 321mV and -1.1mV
R107 -1.5mV and -1.5mV

my findings were changing slightly as I was taking the measurements. I seemed like the readings would drop slightly as the amp warmed up.

thanks for your advice.

[bbartilson]

Sounds to me like a non-issue. The 300mv at the output of the one op-amp is merely trying to offset the 160mv from each of the other two- and it's doing a pretty good job of it. You could try swapping op amps around to see if any pair of them generates lower voltages, and therefore less idle current. My guess is that in actual operation, the current asked for is quite likely a lot higher than what you're seeing at idle, and that's what it making them yellow the board.

Do we still have the original symptom? Your shotgunning the IC's may have fixed the problem as well as teaching some valuable lessons.

-Bill

[Garydean]

Thanks Bill,

The amp is out the door and working great. Certainly a great lesson for me working with IC's. Thanks for sticking with me on this one. You along with everyone else helped me greatly. I really appreciate the experience.

Gary

[bbartilson]

Thanks Bill,
The amp is out the door and working great.
Gary

You're quite welcome.

Glad to hear it's working.

-B