Solid State Guitar Amp Forum | DIY Guitar Amplifiers

I am new here and wondered if anybody could help with the following problem?

I have a Peavey 112 studio pro which sounded great - especially clean channel with lots of reverb but then...

I built a silicon fuzzface (BC108) and was testing it out yesterday through the amp. I got sound - albiet a little quiet. There a couple of crackle and then suddenly nothing but a loud low hum - sounds like mains hum. This does not vary in volume or frequency with changes in the amp's controls but is constant and loud.

I am assuming I have knackered the power stage - does this assumption seem correct? If I have would the home made pedal being plugged in have contributed to this and if so how?

Thank you for your help in advance.

It sounds like your power amp blew and is sending Dc to the speaker= BAD THING.
Disconnect that speaker at once and measure whether you have DC across the wires which feed it. I bet you do.

Why this happened? Probably your distortion pedal was oscillating like crazy at some inaudible frequency, and power transistors overheated and couldn't take it.

Thanks for that. DC across speaker wires - looks like one of the output transistors have gone I guess.

Why would the fuzzface go into that kind of oscillation and why would it cause the power transistors to overheat?

The Fuzz Face is a very crude preamplifier but with very high gain.
More modern distortion pedals use some small capacitors to kill high frequencies above 10KHz or so, the FF did not.
Now, the original ones used Germanium transistors which anyway "didn't have highs" so it was never a problem ... but silicon transistors have gain well into the Mega Hertz range.
Nothing a 47pF or 100pF capacitor can't cure ....but to most it's "heresy" .... what will Saint Jimi think?
Oh well.
Couple that to a handmade construction when these problems appear often.
Mind you, it appears also in a Factory setting ... but then they make a lot of them, they get tested and corrected if necessary.
They have the time and money to test anything a lot.
And even so ... in a couple of Jimi's recordings, notably at Isle of Wight, a strong radio interference can be heard when he steps onto his FF.

Failed transistors?
They don't like switching rail to trail at tens of KHz and overheat *way* beyond normal, which is what probably happened.
When you heard that "low power" probably they were struggling with the oscillation.

Google your Peavey schematic and post it here so we can suggest measurements and start building a lamp bulb limiter here in SSGuitar, you'll need it.
Don't worry , you'll repair your amp.

I ended up contacting Peavey in UK for a schematic - Corby is the head office here. The schematic is for the EFX model - mine is the non EFX. I have attached the power amp schematic that they provided - I am assuming that it is similar. The other versions that were available online were the non-transtube which I believe is a different circuit - particularly in the power amp.

Thanks so much for your detailed info on the fuzz face - I am not averse to putting in caps to tame that hf. The authenticity doesn't worry me - I like the sound, the way I get it is irrelevant!

Once again thanks for your help - it is nice encountering such a helpful, friendly forum.

Forgot to attach the power amp schematic! Here we go. Will look up the lightbulb limited.

Or it could be total coincidence that the output blew up right then, and have nothing to do with the Fuzz.  We often never know why a part picks its exact moment to die.  "My amp blew a tube right in the middle of Sweet Home Alabama.  WHy would that song kill my tubes?"  It didn't.

I have at least six different Peavey Studio Pro schematics, and I think there are even more versions than that.   This is like cars, a 1969 Chevy Impala is nothing like a 2009 Chevy Impala.   What EXACT model do you have?   DO you have a "Studio Pro" with a 1x12 speaker?  Or do you have a "Studio Pro 112"?   If you asked for a "studio Pro 112"   they probably sent you THAT schematic.  What does it say on your serial number plate? Get that exact information and the serial number and call PV back.  If it says STudio Pro, then don;t call it a Studio Pro 112, at least not when ordering schematics. If it says Studio Pro 112 '04, don't leave out the 04 part.  Did you mention it was transtube too? Then it usually has TT on the serial plate.  Then get ahold of Peavey again and ask for the proper schematic.   All those versions in my files, and none of them are transtube versions.

Absolutely Enzo - n=1 does not equal a data set. However, I thought it useful to ask the question so I could see if that is a potential problem and  and if there is find a way to prevent this happening again if it is.

With regard to the exact model I emailed Peavey to point out that it was a different model to mine and they have requested I send details of the front and the back plate so that they can determine which schematic is relevant. I will post when they have replied.

I believe that the attached is now the correct schematic.

Sweet Home Alabama not withstanding, my money is on JM's diagnosis; the "not very loud" is a good clue to supersonic drive.  I'll bet the pilot light went a bit dim at the same time.

Edit to add: what will kill supersonic oscillation dead won't make any difference to the top end response from a guitaring POV.

Could be a lot of things.  My experience with the Peavey line tells me they are seldom unstable, but if you have RF going on, look for an open R122, look near the speaker connections on the schemo.

Loud hum and DC?  OK, could be a shorted output, though that often results in blown fuses.   I'd first look for an open R125, in the return leg of the speaker.  Look right below R122.  I'd then suspect the op amp.

A reason I don;t suspect the fuzz facce is that even if it went all RF on you, that RF signal would have to get past C15, and then ALL THE WAY THROUGH the preamp to even get to the power amp, let alone cause it to fail.

Your power amp has of course two main supplies, make sure both are present and clean, but the power amp also centers with help ffrom +15 and -15.  Not only does the op amp run off those rails, but your four diode junction bias string also centers between +15 and -15.  If one of those is missing, your output cannot center, plus your op amp would wind up way off center, sending DC offset instructions to the finals.

Remember too, this is a flying rails or grounded output form of amp, the main power supplies are referenced to the speaker NOT to ground, so if you find that instead of +35 and -35 (or whatever) you have instead +50 and -20, you still have the 70v rail spread, but it is being yanked offcenter by the circuits.   If so, remove the op amp and see if the power stage centers itself.

I don't think anyone is suggesting the amp itself was unstable, but I have seen situations where a gross supersonic (not RF) input signal caused considerable distress downstream, and we could be getting 9 volts worth of supersonics at the input.  Looking at the circuit there isn't a lot to prevent a high audio frequency passing through, certainly up to 10kHz - I'm thinking of something between the speaker cutoff and the amp cutoff, say gross overdrive between 5 and 10kHz.

Looking at it another way, based on your experience, what would you expect this amp to do if presented with say 10 volts of 10kH and cranked right up (because the audible part of the signal sounded "weak")?  What's the Ft of these output 73100/83100 transistors (I can't find a datasheet for them)?

I have not calculated the freq, but aside from C15 at the input, we also have C55,71,68 across op amp feedback loops.

What would the amp do?   I honestly think it would shrug it off.   I was recently in a discussion - I think on a different forum, maybe MEF - wherein I was convinced by others that feeding RF to a speaker (feeding signal higher in freq than its upper response) would not burn up the speaker.  And that implies the amp was not sourcing any current into the load.   The output devices will crank out an unloaded waveform of any size all day if it doesn;t have to provide any current.  Voltage swings don;'t get them hot, current does.

I am sure the transistors have MORE than enough bandwidth to cause trouble if freq is a concern.   70473100 and 70483100 are just house numbers for MJ15016 and MJ15015  Motorola/ OnSemi.   If I have space I will attach the PV cross ref guide.


These are just my thoughts on the amp, I am not demanding everyone agree with me, I could surely be wrong on this.

Peavey Master Semiconductor Cross Reference

Thanks Enzo.  Well that IS a surprise!  When you peel it all away we are actually looking at a 2N3055/2955 pair hiding under an in-house number!

http://pdf1.alldatasheet.com/datasheet-pdf/view/11469/ONSEMI/MJ15016.html (http://pdf1.alldatasheet.com/datasheet-pdf/view/11469/ONSEMI/MJ15016.html)

These are notorious for their poor bandwidth, being cooked by ultrasonics, and in fact exactly what I had in mind when I said "distress downstream", above; I just didn't expect to find them in a modern amp.  These are the very fellas I've seen destroyed by full swing ultrasonics because their f_T is runcible compared to modern devices, the NPN version has a min f_T of only 800kHz (the PNP being a more reasonable 2.2MHz, but still not wonderful).

So I'm going to put a buck on the NPN device having suffered Second Breakdown as a result of ultrasonic drive.  The device killer here, as I understand it, is hole storage time within the device, resulting in high voltage across, and high current through, the device at the same time, taking it outside its SOAR curve.  This is what was once known as "purple plague" after the colour the die went after failure.  I wonder what the resistor in the Zobel network looks like - thermally distressed I wonder?

Oh it could be so many things, but now we venture into the land of how would we ever know.   SHort of cutting the bad part open AND knowing what we were looking at, we are left with a failed circuit needing repair.

I tend to stick with the simplest explanations - the old when you see hoofprints, think horses, not zebras thing.

Thanks all for your informative replies - some went a little into the esoteric for a newbie, but interesting and appreciated nonetheless.

So, what is the way forward with this? Do I just find the relative transistor and replace? Anything to watch out for?

In general lines, yes, but there may be other bad parts, that's why a proper schematic is essential so we all "see" the same.
Like many others here, if I had it on my bench *maybe* wouldn't need one, because the actual PCB would be before me, in plain sight.
And even in that case, the schemo helps a lot, at least saves time.

When I started helping people on line about ten years ago I thought that I may be able to help out by e-mail perhaps one-in-ten.  I'm still amazed that it has turned out more like nine out of ten when you can't directly observe the gear, much less the fault.

I have been dispensing advice online in this biz for maybe 15+ years, but I have been training techs in one industry or another since the 1970s, and have in a lot of that time been in field service.  SO I have had many opportunities to fix stuff with someone else's hands.   I have come to be fairly confident we can fix most things within the owners capability.

I learned early on to establish a level of communication, one lesson came when I suggested a guy start by checking his plate voltage, to which he relpied, "Is the plate that metal thing the amp bolts to?"  Oh, can;t assume we all share the same vocabulary, can we.

One thing I say to new techs a lot is that MOST repairs are simple - if you can solder, you can fix 90% of stuff.  Broken or loose jacks or controls, cracked solder to main filters or maybe large resistors, cracked solder on all the panel controls, bad tube, etc...

The truly fun part is when you try to explain the function of some circuit - that bias green/red LED circuit on an SVT for example - so someone might understand it, and I find out that up to that point I had never thought my way through that circuit, so this is the first time even I understood it.   Or some circuit I had a working assumption about but never anaylzed and find out my working assumption of long time was in fact not correct.   Learn something every day.

Oh yeah, I've certainly found that to be true, forced to think about something I just taken for granted before.  Trainees certainly keep you on your toes - if they are any good they ask seriously curly and perceptive questions.

Only today I was dealing with SWMBO's car which has had a baffling intermittent pause that has had several mechanics (and me) scratching their heads for weeks.  Today my son (who is also a mechanic, but who learned diagnostics from his dad) seems to have located a dirty connector on the air flow sensor.  Time will tell, but he seems to be the first person to actually "touch" the fault and change the cars behaviour (very much for the better).

I've done a lot of 'puters over the years and a good 90% of those problems come down to dirty connections (the other 10% being software configuration or virus problems that take 90% of the time  ::) )


Communications: was doing a 'puter help over the phone with a German fella.
"So where are you?" (with the system)
"Sitting in front of the computer".
...
"Okay, what's on the screen?" (at the command line)
Proceeded to read the entire boot screen from top to bottom.


Reporting a damaged public telephone.
Operator: "What state are you in?"
Me: "Wide awake and stone cold sober."

(5 seconds baffled silence)

Operator: "No, no; what *State* are you in?"
Me: "(sigh) I think you mean 'WHICH State am I in?' - I'm in Victoria."
{so much for centralised call centres}

I know that this is an old thread but reading through it I have a question that is universal.

Let's assume that the amp was killed by the ultrasonic output of the Fuzz Face. 

If an amp can be killed by a 10KHz oscillation signal, why didn't Peavey put the the 47pF or 100pF capacitor in the preamp input to prevent it from ever getting to the power stage?  Wouldn't that be good design practice to protect the amp?

Thanks!

  If the amp could be killed by a 10Khz signal, then it can be killed by a high note.
Putting a cap in to kill 10khz is going to hurt the treble response of the amp.

QuoteLet's assume that the amp was killed by the ultrasonic output of the Fuzz Face.

Well, let's not, because it won't.

QuoteIf an amp can be killed by a 10KHz oscillation signal

It can't.

Quotewhy didn't Peavey put the the 47pF or 100pF capacitor in the preamp input to prevent it from ever getting to the power stage?

You are assuming it would get that far through the amp to cause trouble.  The speaker won't respond to that high a freq anyway.  A zillion hifi amps go through life passing 10kHz.

Don't confuse a signal with oscillation.  If 10kHz appears at the input to the amp, the amp has no idea whether it came from something oscillating or from a high note.  To the amp it is just a signal.  IN this context, oscillation in a power amp - or any other section really - refers to the circuit becoming unstable and essentially feeding back on itself.  THAT can damage an amp, and they all DO have a circuit or two to prevent it.  Look up "zobel network".

Quote from: aoresteenLet's assume that the amp was killed by the ultrasonic output of the Fuzz Face.

That is actually a whole bunch of assumptions, many of which are debatable.


My homebrew Twin-50 stage/synth amp is getting pretty long in the tooth these days so it still uses 2N3055's in the output stage.  A particular oddity of the good ol' '3055 was that, for a transistor, it has a very low f_T or transition frequency.  In English this means that its frequency response is ratshit and that just about any modern transistor would eat it for breakfast.

But it makes a handy example.

The result of running a transistor up close to its f_T is that it get quite lossy, so its heat dissipation is a function of the operating frequency.  Despite being the poor kid on the block frequency wise the amp will handle in excess of 10kHz with no problems, just the heatsinks run a bit warmer.

Without going too deeply into transistor physics a primary reason for this is that the '3055 has a very long hole storage time, or in English it takes a significant time to turn the transistor from "on" to "off".  The result is that both transistors in the output totempole can be conducting at once, even with no load at all connected.

Talking of which, the typical loudspeaker has a rising impedance and being highly inductive a 12-inch guitar speaker may be effectively open circuit above about 5kHz, never mind 10kHz.  Not so piezo tweeters which are almost entirely capacitive and will be quickly destroyed by application of continuous high or supersonic frequencies.

My Twin-50 has no problems because it was designed for continuous power output with BIG heatsinks, as all stage amps should be, but domestic stereos are not designed for continuous output because even replaying Thrash Metal the actual duty cycle is a fair bit lower, and for typical programme only about 30%.  To design an output stage for 30% duty is a hell of a lot smaller and cheaper than for 100%, so stereo amps are at risk if asked to deliver continuous power, as more than a few poor young guitarists have discovered.

Subject to an oscillation the entire signal chain is driven to the limit, and if the signal is out of the range of hearing, supersonic or above what the speakers can reproduce, the power stage may well cook because you are not aware of what is happening.  This shouldn't be an issue with a guitar/stage amp, but then not all are created equal (as this "heatsink" shown in the Peavey Escort thread (http://www.ssguitar.com/index.php?topic=3692.msg28570#msg28570) illustrates).


So after that quick whizz around the ultrasonic block we come back to the original observation/diagnosis that "10kHz from a fuzzface killed an amp".  On the face if it that is dubious and deserves a deeper dig to find out what really happened.


(ed: typo)

Thanks all!  A lot to digest here.  I've got a lot to learn about ss amps.