JCM 900 frying speakers

[Teleman55]

A customer brought in this amp because he has fryed 3 Celestion vintage 30's.The amp is a 100wt combo,1X12.He said after a tube failure LED came on he replaced the output tubes,and started frying speakers.So I checked the screen grid resistors(2.2K)-1 measured 2.0k the others are 2.1k-The solder joints looked as thought they had been getting real hot,looked grainy.I suspect the OT,having read that Marshall OTs can be somewhat delicate.As an initial test I plugged in to the speaker output and it tests as shorted,but so do working amps.A resistance reading is .0-I have yet to try the current limiter to test for a short,and have found a test method using a 6.3v transfromer.
This amp belongs to a touring muscian and is a part of his signature sound.He has opened for ZZ Top,played with Mark Farner,Rick Derringer,Matt "Guitar" Murphy,this is a hard working amp.If a replacement is needed do you recommend an original replacement ,or maybe a Mercury Magnetics for improved sound and reliability.I have no experience with them and want to do right by this guy.

[stokes]

The current production Celestion vintage 30 is a 60watt speaker.A 100watt combo into a 60 watt spkr.is surely going to blow the speaker.Do it enough and the OT is likely to go next.I have found Marshall OT's to be less tolerant than others.I think the OT should be disconnected and tested thoroughly.I would also recomend not trying to force the 60watt speakers to handle the 100watts that amp is capable of putting out.

[Teleman55]

The owner told me that on the road he uses 4X12 extension cabs,the speaker that is in ther now is a Mesa Black Shadow Celestion-90wt.By the way sorry for posting here I meant to be in the repair forum.

[stokes]

Your original post only mentioned 1X12 Celestion Vintage 30.Where exactly is it blowing the speakers?If it is amtter of him using a severe impedance mismatch any damage would more likely occur in the amp,rather than blowing a speaker.You are going to have to give more specific details as to what he is using and what is blowing.

[Teleman55]

Thanks Stokes; I spoke to the owner tonight here is what I know.Has play the amp for years with the Vintage 30 that it came with.During a gig the amp died and he noticed a tube fault LED was lit and he replaced the tubes with Mesa EL34s and the dead speaker with the Mesa 90wt.(he is sponsored by them).He played the amp for a day with no problems,went to a gig and the speaker died and he brought it to me.
I did some more testing,on both my Peavey DB210 and Hot Rod Deluxe I get a resistance measurement at the speaker jack.The Marshall measures .0 ohms.Using a light bulb current limiter shows no short,first with no power
tubes and then with them installed.I checked some voltages(OT plates-490,bias_-46v) and when checking the filament supply I rested my probe on the chassis and found 231mv of AC,checked the speaker leads and found the same thing.That is as far as I have gotten today.

[Enzo]

If the OT speaker secondary winding were shorted, it wouldn't be blowing speakers, it would be putting less out to the speaker. If the OT primary shorted the high voltage to the secondary, it would blow the fuse right now. I really doubt the secondary winding is involved in any speakers blowing.

Ohm readings tell you little. Most output windings measure less than an ohm. Your PV amp is most likely a 16 ohm winding which will show more than a 4 ohm, but still not much. There are relatively few windings on that secondary and the wire is heavy. That is a formula for low resistance. If someone used more turns or lighter wire, then the reading will be higher. It is hard to damage the secondary winding.

Have you seen any of the blown speakers? Cut one open and look at the voice coil. DO all of them if they are available. If the coil is warped, darkened, burnt, then it was exposed to exces power over time. If the coil looks like new, then some impulse more likely killed it.

He may have got away with that 60w speaker and the 100w amp, but did anything change in the band or his rig? Maybe he is turning up more now?

Look for instability in the amp. If the amp goes into RF oscillation, it will be cranking 100 watts of RF out to the load as hard as it can. You wouldn't hear anything, and the speaker can't move that fast. SO all that energy just goes into heating the voice coil.

What model amp is this? Does it have the little 22pf disc cap between pins 3 and 4 of the end power tube - the one closest to the power tranny? See if that is broken, missing, damaged.

Does the guy run straight or does he use effects either in front or in the loop?

[stokes]

Have you tried the amp with a known to be good speaker?As Enzo pointed out,I would check those blown speakers,if for nothing else to be sure the speakers are actually blown,and there isnt a problem in the amp.Check the output with a moderate signal going into the amp set at a moderate level.If you see an unusually high output,this could indicate an oscillation as Enzo described.If everything in the amp is okay,I would have to say it is just a case of a 60watt or 90watt speaker not being able to handle the 100watts the amp is putting out.Using that amp with one speaker that is under rated for the output of the amp in a situation with a drummer and bass player and whatever else in the band you are certainly going to blow the speaker.It is possible that blowing the speaker a couple of times did damage the OT,but I have to agree with Enzo that a damaged OT is not going to blow a speaker.Check the amp thoroughly and when you are satisfied the amp is okay,tell him to get a 150watt speaker if he is going to use the amp as a combo with full power going into one speaker.

[Teleman55]

Thanks for the replys-Tonite I plugged in a JBL e120 and played at about 1/4 volume in both High and Low power modes with no problems for about 15-20 minutes.No problems,the amp functioned fine.I will see if he will let me do a speaker autopsy,I have the Celestion 90 in the cab.Ther is no cap on the power tubes.The amp is Model 4101-serial #Y00010-The owner says it came with a parchment attesting to its being the tenth amp manufactured in this line.
I have a Hitachi V-212 20mhz O-scope and a Heath/Schlumberger Audio generator to test with but I have limited experience with them.I could use some guidance.What about the 231mv of AC on the chassis does this indicate a leaking cap?

[stokes]

How are you reading this ac voltage on the chassis?Where are you connecting the DMM leads?Obviosly one to the chassis,but where is the other lead?I dont think you need the scope or signal generator to check anything.If the amp is okay the problem is running the amp into a speaker that is not rated at least 100watts.To put out 100watts the amp doesnt have to be maxed out.If you have a 150watt dummy load you can put a 1k signal at between .15 and .25 volts into the input and measure the RMS ac volts on the dummy load and you will see where it reaches 100watts.He may have been lucky using the speakers you mentioned but sooner or later putting 100watts into a 60 or 90watt speaker its gonna blow.In that amp I would use a 150watt speaker,at least.At full bore its going to put out more than 100watts.If you dont have a dummy load you could run the test I described with a speaker but it will be quite uncomfortable.Just read the volts across your speaker terminals.If you are using a 16ohm speaker,when you see 40ac volts you have reached 100watts.You must have a True RMS voltmeter,and I dont know if you will be able to stand a 1k signal at 100 watts.I have done it with a 30watt amp and it is brutal.

[Teleman55]

Stokes-I had my ground connected to a transformer bolt and measured the AC on the center of the chassis and on the speaker leads.I will try your recommendations tomorrow.

[stokes]

When you connect your lead to the transformer bolt and the other to the center of the chassis you are connected to the same source-both are at ground you are not getting a true reading.I suspect when you touch the speaker lead you are touching the ground lead,again both leads are at the same source.A leaky cap is not going to leak ac voltage,they pass ac and block dc.That .231vac reading you see is a ghost,forget it.

[Enzo]

Well, they are both places chassis ground, but the chassis is not a zero ohm impedance. That is why we make star ground wiring arrangements. The small resistance in the chassis is still enough to cause a voltage drop at high enough currents. It has nothing to do with grounding caps, but I can believe a quarter volt from one end to the other. After all, ground loops happen when everything is connected to the same ground.

This would be a normal condition.

[stokes]

Yeah,I see where you are coming from,Enzo,makes sense,but wouldnt that current be more likely to show as dc?But like you say it aint nothing to be concerned with.

[Teleman55]

I will post again after I get to check the speaker voice coils.I played the amp tonight after letting it warm up and then running a 1k signal into it for about 10 mins..Of course with the family home I can't really crank it up.

[Enzo]

You need a dummy load, and I don't mean the wife...

[stokes]

But if she dont put up too much resistance.......

[Teleman55]

I spoke with the owner (Dan Lawson) and he told me that he has been playing bigger venues and playing louder.His amp setting are-Clean channel,preamp volume at 12 oclock and master volume at about 3 oclock or 3/4s.By the way the vintage 30 he was using ended up with the paper cone tearing to shreds.He gave permission to examine the 90s voice coil,I will look at that tomorrow.Dan also told me that he has spoken to his contact at Mesa Boogie who told him that these Marshalls can have oscillation problems and agreed that a 150w speaker is the way to go.I had been looking through my TUT books (1&3)and he mentions that when Marshall went to smaller cabinets the crowded parts layout led to oscillation problems.
The only dummy loads I have are low wattage,until I get a better one is there a way I can identify oscillation? Enzo you mentioned a cap on the last output tube,that would be a snubber yes? Do I need to identify the oscillating frequency to choose the proper cap value or can you suggest a suitable value that will not alter the tone?I hope I am making sense and not being a PIA.

[Enzo]

Smaller cabinets and more crowded? These 100 watt heads I have sitting around all seem to be pretty close in size.

I looked through the various models in the JCM900 line - really a model number is far more useful than JCM900 - and none of them seem to have the little cap. it appears on some later models.

You can see oscillation on a scope. The flat line of the trace expands into a wide band of light on the screen. If you then speed up the trace it wil resolve into a high freq repeating waveform.

The dummy load is just so you don't have to listen to all the noise. Oscillation will happen with all manner of loads.

[Teleman55]

This is a 4101 112 combo amp,it is smaller than my Fender 75 112 combo.Do I need to identify the frequency of the RF to properly choose the cap value?Could the fact that he installed the tubes with out rebiasing have caused oscillation problems.I am going to bias the amp tomorrow before I do any testing.This is my first Marshall 100w tube amp,so far I have only worked on 50w and Valvestate Marshalls.

[stokes]

I really dont see any reason to suspect an oscillation causing the speaker to blow.You have a 100watt amp,you are using a 60watt or 90watt speaker.An amps stated rating is the watts it puts out before clipping.I have yet to meet a guitar player who plays a Marshall "clean".If there are any malfunctions in the amp I would suspect they were caused by the blown speaker,not the opposite.An oscillation would be indicated by other problems.Lets assume I am wrong and your 60watt speaker can handle 100watts.If you had an oscillation problem it would cause the tubes to work harder,but the tubes plates are only capable of putting out x amount of watts,in this case 100,so as the oscillation causes the tubes to try to put out enough watts to blow the speaker the tubes wont,the tube plates will glow red and get hotter and hotter till the tube fries,not the speaker.

[Enzo]

The idea with RF is that if the amp puts out the energy, it has to go somewhere. A speaker that can handle 100 watts does so by turning it into mechanical energy - moving air. It doesn't just dissipate it in the voice coil like a light bulb filament.

ANy energy that doesn't get turned into motion has few options other than to become heat in the voice coil. The speaker cannot physically respond to the RF - it cannot move that fast. So ALL the energy in the RF from the amp goes into becoming heat. So the speaker sits there in silence burning up. It can't turn any power into motion, and the lack of motion means it cannot cool itself like it would producing sound.

That is how RF can damage the speaker, it is not from overpowering it as you might with audio.

That said, I agree, it is far more likely we have someone trying to cram 100 watts into a 60 watt speaker. oscillation is just covering the basses.

AS to what value cap to use in an amp to stabilize it, that all depends upon where it will be in the circuit. Stability caps are not aimed at a particular freq, they simply roll off the high end response of the amp. You set the rolloff point above the highest freq you want coming through.

I wouldn't be slapping stability caps on the thing unless I knew there WAS oscillation going on.

[Teleman55]

I see what you mean and agree but still want to be sure that when returned the amp is road ready.I have not had a customer who travels,plays and records like this guy and don't want to over look anything.Having never had to diagnose an oscillation problem I want to learn as much as possible.

[stokes]

That would be one hell of a RF,I would think a RF oscillation that high would be beyond the frequency response of the OT and not make it to the speaker.I just dont see any thing to indicate this problem is related to anything other than the numbers- 100 dont go into 60.I aint saying that what you describe in your RF theory is impossible,but I think it is extreme and unlikely.If that were the case that the oscillation was so extreme the speaker would burn out with the amp idling.

[Enzo]

I don't disagree. I also think it is simple overpowered speaker.

I went into the explanation only to describe how the speaker burns up under those conditions - from heat, not from excess power. I was countering your argument that 100 watt speaker would be ok with 100 watts of RF. It wouldn't. The 100 watt speaker will burn up on a 50 watt amp even if all it does is turn into a space heater. The power rating of a speaker is handling audio, not simple dissipation, and it is assumed most of the energy goes into moving the cone. I don't know how efficient the OT might be at 30kGz or 50kHz. When we say RF, we just mean any oscillation above audible. SOlid state amps drive the speaker directly, and it is indeed a problem when it happens there.

If the amp is oscillating, the audio trying to fight its way through is essentially irrelevant. The amp would be cranking all by itself.

[Teleman55]

I went through the amp today and I installed 1ohm resistors to bias it.I was getting the most odd readings 2 tubes at 35ma,1 at 19ma and 1 at 95ma as I tried to make adjustments it would go from these reading to all most 0.I started tapping on the bias pot and the voltages would change.I guess the pot must have burned where it was set when the other tubes went out so I replaced it with a 15 turn cermet but the best I could get was 3 tubes that matched and 1- 8ma lower, I figure its the tubes because the voltages moved with the tubes when I shuffled them around.These are Mesa EL34s.I recommended JJ tubes,I have been very happy with those.
I want to thank you for all your help most of my repairs are physical damage so this seemed like it might be a bit tricky and I neede expert guidance,still have much to learn.

[stokes]

If you saw that 8ma move with the tube then it is surely the tube that is mismatched.Enzo,I could see that oscillation being more likely in a SS amp,but in a tube amp there are other symptoms that would most likely show before the speaker blows.To me RF means just that- radio frquency,I would refer to other oscillations as high frquency oscillation or low frequency,just semantics,I know,but when I hear RF I assume RF,dont blame me,I'm just a product of my environment.One of the most prominent symptoms of oscillation at the speaker is lower volume of the audio signal fighting to get thru while the oscillation uses up the power amplifying an unaudible signal.Most notable with HF oscill.as low freq.will cause motorboating of the audio signal as well.

[Enzo]

We are not really at odds. Once the topic came up I felt the need to discuss it, that's all. It may be sloppy talk, but all my professional life we have referred to oscillating above audio as "going RF." You are right in that 50kHz would be considered VLF or even ULF if it were considered RF at all, and most of is would not call it RF in general, but it is so much easier just to say "RF" than to use multi-syllable terms. I shouldn't do resistor is at proper value and check the resistors involved with the control grid.

[stokes]

I didnt think we are at odds at all.Like you said its just discussion,which is exactly what a forum like this is all about.I see now you use the term RF to simplify things,and I just read it too literally.I totally respect your knowledge on amps,Enzo,and it is always good to bounce opinions with you here,if I seemed argumentative at all I appologize,was not my intention.Back to Teleman,if that old bias pot was a normal volume type pot,they dont like dc on them and the trace will burn out over time,the cermet is a much better choice.As well as being better with the dc volts,you get a smoother adjustment with it,as you probably noticed.

[Teleman55]

Yeah I'm glad I had one on hand I don't really have a lot of parts in stock,this is a part time,at home gig.Dan has been talking to his contact at Mesa-Tim McKee and he is impressed with the questions I've been asking and thoroughness as well as agreeing with the answers I came up with.I owe it all to you folks and this forum.I really love doing this and experience,availability of good books and guidance from people like you makes the learning less frustrating and frightening.Hell I don't want to blow up someones amp due to my stupidity.Oh and it doesn't hurt to be stubborn and obbsesive at times,If I can not figure out what a problem is I know I've overlooked something or need to learn some more and thats when it time to seek help.