Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 29, 2024, 06:59:28 AM

Login with username, password and session length

Recent Posts

 

carvin x-amp

Started by ilyaa, August 03, 2014, 06:45:37 PM

Previous topic - Next topic

ilyaa

heres a weird hybrid one -

R132 was (literally) burnt out - i replaced it and the amp is working fine, but im trying to understand what would make that resistor fry -

i checked the diodes in the bridge that feed it and they seem fine - where does the V- that its filtering/providing even go? i assume to the opamps for a +/-15V swing but i cant quite find evidence of that on the schematic -

what could make that particular resistor burn up? its in a lowish voltage part of the amp so im a bit confused.

note: R132 is in the lower right part of the schematic - off the transformer, through the lower voltage bridge -

Roly

Yes these +/- supplies are implicitly for all the op-amps.

Well you are certainly asking the right question, burnt, yes, but why?

Most likely a short to ground.

Check the board for any signs of work, fresh solder, particularly on op-amps.

No loose metallic objects inside, screw, nut, bottle top, ring pull...?
If you say theory and practice don't agree you haven't applied enough theory.

ilyaa

there have definitely been some modifications made but all the solder work looks pretty good - nothing obvious and all the joints around the op-amps look very clean -

anything other than a short to ground possibly responsible?

the person who gave me this amp to look at claims that he has a pedal that does poorly with amps - an ibanez sonic distortion "modified" pedal or whatever - ive never seen it and am not familiar - but lets say worst case scenario a pedal somehow dumps 9V into the input jack of an amp - could that wreak any havoc? and wouldnt op-amps be the first to go in this case?

Roly

Well if you have evidence of a short on one of the supply rails one line of investigation is to look at the op-amps powered by it, and in particular any that have exposure to the outside world via sockets ('tho it's a bit unusual to have a problem on one rail only, and worth looking for anything that is only powered by this rail).

It's unlikely that any input op-amp is going to be worried by voltages within the supply span, say +/-15V, but in the outside world anything can happen so it is possible e.g. that the output of another amp has been connected to an input (or output) and hit it with a higher voltage, or that some external mistake (such as not grounding the stomp supply) has resulted in a high mains leakage, or it got hit by static or even lightning.  You are dealing with tout le monde and about the only thing you can totally rule out is Aliens.  The older the amp the longer it has had to suffer some accident or techno-stupidity.

This is a really classic example of the forensic detective work techs often have to do, but unlike TV you don't always get an answer.

Guy brought in a s.s. PA which turned out to have a boofed output stage.  Fixed, returned.

Came back a few gigs later, same problem, but the other channel.  I told him I also wanted all the PA speaker leads.  Sure enough I should have done that the first time because one of them had an intermittent short inside a connector (twisted, ineffective cable clamp).

This is also an example of why you should keep records of jobs, write up a job report (and these days with piccies).  The client doesn't often read them but when it came back in I checked my job record to confirm that it was indeed the other channel.

You will only get a really certain cause like this roughly about 1/3rd of the time, 1/3rd you'll have a pretty good guess, and 1/3rd no real clue.  It's these latter jobs that you need to keep and eye on after they go back.

I'm currently dealing with a string of occasional FET failures in some buffers I built, and I haven't been able to get a handle on the cause yet. { :grr }
If you say theory and practice don't agree you haven't applied enough theory.

ilyaa

hm it sure is frustrating when the answer just does not want to appear!!

no evidence of a short or anything anywhere - and who knows WHAT was plugged into this thing.....

at any rate, im holding onto this amp for now and ill keep playing it to see if things remain stable with all outside circumstances controlled and safe -

another question has come up for me, though: looking around the amp in general i noticed that there are NO individual screen grid resistors - the screens are connected directly to R121 - a 350R 5W resistor that drops off the main B+ supply - they are sharing it. is that kosher? it doesnt seem to be a mod either - the PCB is printed with that configuration - (bear in mind the schematic i posted is for a 100W version of this amp - so mine only has 2 of the output tubes). i guess the main point is to put the screens just a tad below the plates, so maybe its nothing but just seems a bit strange to me....

finally, although the amp is loud and sounds fine, a power test with my scope reveals that its really only making 1W - ??? - and furthermore, when i try and look at the current that the power tubes are draining, they are each at only about 14 mA!! (this is with a homemade bias probe that adds a 1ohm resistor from cathode to ground) - what the hay?! i tried other known good tubes and same deal - tested the transformer, its fine - what else might be keepnig the power tubes from pulling normal current? the bias voltage is at -54VDC.....




Enzo

Reducing the voltage is the job of R121.  The job of individual screen resistors is to limit current through the screen of the tube.  There are plenty of amps that lack screen resistors, Fender 5E3 comes to mind.   You could add them if you like.

14ma of idle current?   About the same as a Peavey 5150.  So it is cooler than the  borderline of melting level most people seem to want.   To get it hotter, you have to reduce the bias voltage.

Roly

Screen limiting resistors are a bit of a "movable feast"; most amps have one per bottle, some only one shared, others none at all.  The object is to limit screen grid dissipation and stop the screens acting like anodes and hogging the current (just before melting down in the case of EL34's).  In the four-bottle Carvin each has a screen resistor because they conduct in pairs and two resistors are required to balance the screen current draws (else one bottle may be working harder than its partner).

A single common resistor for a push-pull pair is and economisation justified on the grounds that when one bottle is pulling the current the other one should be cut off and out of consideration.

Quote from: ilyaathe amp is loud and sounds fine, a power test with my scope reveals that its really only making 1W

:lmao:  Big Lesson No.1 - your ears are a lousy judge of power.  I've had something like this many times and it's always a bit of a surprise to be reminded just how loud so little power can sound.

The most common cause of that is loss of emission in the output valves, just worn out, but if a new pair doesn't fix matters then you need to investigate;

- is the screen supply okay (is/are the screen resistor(s) intact)?

- are they are being sufficiently driven, i.e. is the PI putting a peak voltage into the grids about the same as the bias voltage, are they being turned on hard?  (If you don't have enough drive voltage then you need to go looking back along the signal chain towards the input to find out why).

- if the voltages and drive seem to be in order with new bottles but you are still only getting pathetic power it's time to have a look at the output transformer for shorted turns.  This is thankfully not all that common but it certainly does happen from time to time when a valve amp has been driven without a load connected (which in turn may be because of faulty sockets, plugs, speaker lead, o/c speaker, carelessness, stupidity, etc).  Check this thread;
http://www.ssguitar.com/index.php?topic=3357.0

Really there isn't a lot else to go wrong in a valve output stage.

Such a low idle current combined with such a high bias voltage implies that this is a Class-B output stage idling close to cut-off.  Class-B provides higher power and greater efficiency, but at the expense of increased crossover distortion (which generally sound pretty nasty compared to Class-AB).

HTH
If you say theory and practice don't agree you haven't applied enough theory.

g1

  From the looks of the schematic, this should be around a 40W amp?
Yes 1 watt is loud, but probably not to be confused with 40W  :).
When I run into problems like this, (if there is not gross lack of B+ or very odd waveform on scope), I start looking at what I may have done wrong, or what could be wrong with my test setup.
  For starters I would throw another amp on the bench and verify the test rig.

ilyaa

QuoteFrom the looks of the schematic, this should be around a 40W amp?

oh i forgot to mention: its 6L6's in there, not EL34s like on the schematic....

ilyaa

QuoteFor starters I would throw another amp on the bench and verify the test rig

dont know what i was doing wrong but i came back and re-tested it and it seems fine!

QuoteFrom the looks of the schematic, this should be around a 40W amp?

but even though i was expecting about 50W, im getting about 40 on the scope - what about the schematic indicates the kind of power output i should be looking for, other than the kinds of tubes specified?

Roly

That's actually a harder question to answer than you might think.  The best guide is normally the various typical operating conditions found on the valve datasheet.

As a rule of thumb with valves of this class (anode dissipation) I'd assume that if it has cathode bias then it will be about a 40 watter, fixed bias then about a 60 watter, but if the anode supply and bias voltage are on the high side then it could be Class-B and maybe 75 watts.  It's not easy to tell just by looking at the circuit.
If you say theory and practice don't agree you haven't applied enough theory.

g1

  For a pair of modern 6L6 or EL34, running around 450V B+, around 40W at the onset of clipping is pretty normal.  Push it into clipping just a little and the power goes up.  I think most of the big manufacturers spec their power at 5% distortion.
Do your power calculations with a meter measuring RMS rather than using the scope numbers and you will see how the power jumps at clipping.