Peavey Artist VT Overheating

[ozymandias]

I’m hoping one of you kind souls on here can help me out with a problem with this amp. My experience of working on amps is very limited, but so is my budget so I’ve decided to have a go at this myself.

A couple of months ago in the middle of some low volume practice this amp suddenly made a loud humming noise before dying. It was very hot. I found the 3.15A fuse had blown, so I replaced it, allowed the amp to cool and turned it back on. It worked still, but was maybe a bit quieter than usual. As I had a couple of gigs coming up I decided to limp along with a hot amp rather than be without it. A couple of times it made the loud hum, but I managed to get to it and switch it off before it died; let it cool down and it recovered fine.

It's fine in 'Standby'. It only overheats when ‘On’, and can take about an hour to get to the 'big hum & dying' stage.

Having got those gigs out the way, and fresh from my success of Big Muff building I decided to have a go at a fix myself. Firstly equipping myself with some knowledge about the dangers of capacitors and the working of the triode I waded in.

It might save time if I say here that I think it may be a problem with the Screen Bias Voltage – it’s only about 4V below B+, but according to the schematic it should be 10 below; it’s quite likely that I’m wrong about this as I have no experience in these matters. Anyway, to continue with the full gory details …

First thing I found was that IC17 had fried (it was completely brown) and had left scorch marks around it. I replaced IC17 and (as a precaution) IC15. ICR11 and ICR10 checked out ok. IR12 was over-spec (113ohms), so I replace that. All the other big wirewounds (IR11, R29, R35) were fine, but I replaced them anyway as I had a pack of 5.

Main Board : looked like its had a ‘cap’ job at some point - all the 2uF caps (C3, C23, C30, C32, C46, C54) have been replaced with 2.2uF – the 25uF ones (IC18, IC21, IC24) have been replaced with 22uF; I’ve now put these back to 25uF. Couldn’t find any 2uF’s on the market, so the 2.2’s have stayed.

Tube Socket Assembly board : Removed board. Removed, tightened and replaced each socket pin. Removed a mess of solder. Checked the 47k resistors and the diodes. Replaced the big white resistors (R29 R35). Repaired a couple of damaged tracks with copper wire (looks like V1 had fried at some point, the tracks to pins 3 & 4 having been half blown away and repaired with some big blobs of solder). Cut the track from around each pin 8, and installed a 1ohm 1W resistor between each pin 8 (cathode) and pin 1 (ground ), so I could measure the cathode current on each tube.

Power Supply board : Resistors R2, R3, R4, R5 all over-spec (276k, 310k, 279k, 286k), so replaced them with new 220k ones. R1 ok (399ohm). Electrolytics Capacitors all seem ok – I’ve only got a multimeter to test them with, but none of them are shorted and they all charge. Ceramics are all ok (9.5nF, 8.8nF, 9.5nF - spec=10nF).

AC Voltages : All the AC voltages coming in are slightly over-spec – Red = 384VAC (spec = 370VAC), Violet = 57.9 (spec = 56VAC), Orange = 41.7VAC (spec = 40VAC), Yellow = 6.55 VAC (spec = 6.3VAC).

Tube Cathode Current : These were all over the place – 14.3, 16.2, 8.2, 11.3. I found in another post that Enzo said these were fine at around 11, so I set about getting these below that. First though I dug out some old sets of tubes to try and get a more closely matched set. That done, I installed a 47k Pot after IR9, (and with the wiper tag soldered to the ‘out’ tag so it maxes-out to 110k if the wiper fails). So, I’ve now got Bias Voltage adjustment from -61.7V to -66.5V. Setting it at -65V gives Cathode currents of 7.0, 8.3, 7.5, 7.6, which gets the temperature down to bearably hot.

However, I suspect the Bias Voltage may not be the root cause of the problem. The B+ voltage is 514V and the Screen Grid voltage is 510V (spec = 500 & 490). I read somewhere that a too-high screen voltage can result in too much current, and therefore, heat, so I am wondering if this is the root cause. But, I can’t see why there is only a 4v difference rather than the specified 10v. The resistors R1-5 are all on-spec. I’ve also checked IR6 & IR7, and they are also (just) within spec (24.9k and 24.3k – spec=22k).

Sorry for the information overload, but I don't know which bits are going to be relevant so I chucked everything in. Hoping you can help. Thanks.

[Steve Conner]

You probably had a worn-out set of power tubes that were going into thermal runaway. A new set would have fixed it.

There was no need to change the 22uF caps back to 25uF. Electrolytic caps have a wide tolerance: the actual value you get can easily be 20% different to the nominal value. So, you could easily buy a batch of "22uf" that measured 25uF, or vice versa. The same argument applies to the little 2.2uF ones.

The high voltages could be explained if your mains voltage is towards the high end of the range the unit was designed for. In the situation I'm familiar with, European mains voltage is nominally 230, but in some parts of the UK, the mains voltage can be over 250V.

[soundguruman]

well um it probably is the bias voltage or some failure of the bias to reach the tubes.
But there is no IC 17 that I can see on this posted schematic.
The thing about these amps is the video driver transistors Q 101 and 102
It's 101 and 102 on this posted schematic, there may be a difference between your amp and the schematic that is posted.
If the thing is overheating, I immediately would suspect the transistors. Just from past experience, they blow a lot.
I would just replace them anyhow, cause I don't like the look of 'em.
I would order up some new ones from Peavey...
In fact I would replace the diodes and other little parts in that driver circuit (with better parts).

As far as the output tubes, you may have a bad one or more...
Start by running 2 output tubes, one on each side of the transformer.
Test that for a clean sine wave, and for hum...
then try another 2, same test.
then the last two tubes.
By process of elimination you can find a bad output tube.
You can run this amp with 4 output tubes instead of 6...etc...

Well, one thing that "might" help is to call the Peavey techs 601 483 5365.
Ask to speak to a tech, then ask if there are any service notes (by serial number).
The service notes will document the parts that have been failing.
It is advisable to replace the parts that have been causing repeated problems.
They are really nice and helpful, and you can give them a call.
Also, Peavey charges fair prices, and gives free shipping, at least they did for me.
Anyhow I would at least call and find out if you have the right schematic by serial number,
a schematic with IC 17...

[ozymandias]

Thanks for the replies guys. Given me something to chew over.
I'm a bit confused about soundguruman's comments about the wrong schematic, though; think there may have been some mistake there as the one I posted does have an IC17 (page two top left). Thanks very much for the advice anyway - I hadn't gone anywhere near the transistors but I'll have a close look at them now and see if I can get some replacements.

[soundguruman]

Thanks for the replies guys. Given me something to chew over.
I'm a bit confused about soundguruman's comments about the wrong schematic, though; think there may have been some mistake there as the one I posted does have an IC17 (page two top left). Thanks very much for the advice anyway - I hadn't gone anywhere near the transistors but I'll have a close look at them now and see if I can get some replacements.

Yes they are video drivers, but Peavey uses them for audio...
These blow and screw up the bias.
Just because a part does not look "blown" does not mean it's good.
So test those transistors...

[J M Fahey]

IC17 exists and is visible both on page 2 top right, between ICR11 and IC15 and on page 3 , center 1/3 left.
It's a .1 Cap (probably ceramic), filtering the -15V rail.

I fail to see how a failure on the High Voltage drivers, Q101/102 can affect power tube bias, since said tubes have their own bias supply and are separated from Q101/102 collectors by .1 capacitors rated 600 Volts. (IC22/23).

Ok, ok, just *suppose* said capacitors are leaky.
Anyway, it would qualify as a *capacitor* problem, not a *transistor* one.

Also fail to see how a couple little transistors , even failing, can overheat a big guitar head powered by 6 x 6L6.
When working properly, they dissipate 232 milliWatts each.
When either open or shorted, they dissipate less by definition.

I also fail to see how anybody can repair an amp (or anything else) when bombarded with inaccurate (let's be kind) information.
Hard to separate the chaff from the wheat.

[ozymandias]

Yes, point taken. I thought I might be chucking in too much 'chaff'. I can focus on some 'wheat', then ...

Can anyone suggest please why the Screen Grid Voltage is only 4v below the HT Voltage, when it should be about 10v below (see bottom-left of page 3 of attached schematic - +500v/+490v)? I realise this may be entirely academic given the tube/transistor suggestions above, but I'd be interested to find out how this voltage drop is supposed to be achieved and whether it can be changed. In other words, which component(s) set the voltage drop? Is it just R1 (400) - is there some voltage-divider thing going on around R4/R5/R6/R7? Sorry if I'm being a bit stupid, but I am.

Thanks.

[J M Fahey]

Those 10V specified may be some "typical" value but yours does not seem to show some defect.
Besides, Peavey recommendations are always very conservative (the secret behind their reliability and longevity) and , among other things, they tend to nias *everything*, both Tubes and Transistorts, somewhat cold.
None of that Forum invented nonsense of 70% biasing or 60% or whatever.
What you originally has was meager 8W dissipation in your hottest tube (500Vx16 mA) which is already quite cold.
You lowered this to 7.5 mA average, which amounts to less than 4W.
And the colder you bias it, the less screen current you'll have and the less voltage drop across R1.
So your results are normal and consistent with what you did.
If you like the sound, fine, and the amp will last longer.
If you find it somewhat "hard" (which includes a subjective factor), go by the Peavey recommendation.
Can't go wrong
PS: yes, there is also added consumption to feed the PI transistors, but it's quite light.
And "there's no stupid question"

[ozymandias]

Those 10V specified may be some "typical" value but yours does not seem to show some defect.
Besides, Peavey recommendations are always very conservative (the secret behind their reliability and longevity) and , among other things, they tend to nias *everything*, both Tubes and Transistorts, somewhat cold.
None of that Forum invented nonsense of 70% biasing or 60% or whatever.
What you originally has was meager 8W dissipation in your hottest tube (500Vx16 mA) which is already quite cold.
You lowered this to 7.5 mA average, which amounts to less than 4W.
And the colder you bias it, the less screen current you'll have and the less voltage drop across R1.
So your results are normal and consistent with what you did.
If you like the sound, fine, and the amp will last longer.
If you find it somewhat "hard" (which includes a subjective factor), go by the Peavey recommendation.
Can't go wrong
PS: yes, there is also added consumption to feed the PI transistors, but it's quite light.
And "there's no stupid question"

Thanks Juan. Right, so the voltage drop is proportional to the current. Yes, that joins up with something I was reading at the weekend, just didn't follow the thought through. And probably never would have done without your help.
So it looks most likely that the overheating problem was due to some 'thermal runaway' of my rather mature tubes. I'd better get some replacements as soon as I can afford it, meanwhile I'll run it cool and 'hard'.

[Enzo]

First, never lose sight of the fact this is a guitar amp, not a precision rocket surgery piece. VOltages in the schematic are only approximate. If your 490v is 494v, no one cares, least of all the amp. The difference between 2uf and 2.2uf is irrelevant, same with 22 and 25. Back then, odd values were more common. These days part makers stick a lot closer to standard values. Today 2 is not a standard value, 2.2 is. Same in other amps where an old amp had a 0.02uf cap, and today we see a 0.022uf cap. Old 20uf filters now become 22uf filters. 50uf caps are now 47uf caps. etc.

The components on the power suppy board have confusing numbers. For example IC17. It is NOT "IC17", it is "1C17". That is "ONE_CEE" 17 not "EYE_CEE" The parts on that board have their own number sequence, so to "prevent confusion" they added a 1 before all part designators on that board.


Tubes are BY FAR the most common cause of problems in the amps. You have made great effort looking at everything EXCEPT the tubes. I;d bet my lunch money you need a new set of power tubes.


The colder you set your bias, the higher your B+ will climb. Fact if life. It is normal for tubes to radiate a lot of heat. What you don;t want is for the inner metal structure to get red hot. We call that red-plating. Your amp came from the factory with random tubes, no one matches tubes on production lines. We may prefer matched tubes, but the amp doesn't care. If I said amps might be happy at 11ma - and I see 5150s that low and still happy - it was not a target, just a note that it is not necessary to run a 40ma. So struggling to get down to 7ma doesn;t help you.

Tubes can fail without going up in smoke, just as grandpa's heart can have "episodes" now and then without him croaking altogether. I think you need new tubes.

[Steve Conner]

The screen current drawn by a tube is a small portion of the plate current. In beam tetrode tubes like the 6L6, the actual amount of screen current is quite poorly defined. The better they aligned the grid wires at the factory, the lower it will be. So that's one reason why your screen voltage drop could be lower than the schematic specifies. The other reason is as Enzo says, you turned down the plate current so the screen current will have gone down too.

And no, I don't believe the PI transistors are faulty.

[ozymandias]

Thanks guys. Sounds like excellent advice. I'd better turn the bias voltage back up before I fry the innards and blow the dust of my 'new tube' money. Dammit, I'm gonna miss it though, we've been together for so long.