Hey guys. A buddy of mine asked me for help trying to repair or at least diagnose the cause of his xt65 dying. Its not actually dead, per se, as it still makes some noise. But the noise it makes is only a loud buzz, and the guitar signal cannot be heard.
This all started hen he played a gig and had the thing at max volume for an hour or two. He said smoke started coming out of it and its been worthless ever since. I hooked it up to a test speaker to measure the transformer voltages under load to see if the transformer itself was the problem. as soon as i turned it on, the cone on the speaker moved forward as far as it could, and began to smell like it was burning after only a few seconds. Obviously something is placing a high DC potential on the speaker.
So i have two questions
1. What could cause DC on the speaker like that?
2. Is it safe to test a transformer with just the primaries connected, and measure the voltages at the secondary side?
Yes of course you can connect a transformer to the mains with the secondaries not connected. But the power transformer is not your problem. It won;t place DC on the speaker leads.
You have the oldest problem in the solid state world, a blown output stage. Probably one of those TIP142 or TIP147 transistors is shorted. Check them with your meter. REplace the pair even if only one is bad. If they both seem OK, then continue.
Since it doesn't blow fuses, disconnect all speaker loads - the amp doesn't need them, and you don;t need to burn them up. Once the amp no longer makes DC, you can connect one.
First verify power supply. The metal tabs on the output transistors is a handy test point. Looks like 40v rails, so you got +40 on one and -40 on the other? More or less?
You mean test the Voltages on the transistors with no speaker attached? I want to make sure i understand correctly. And how exactly do i test the transistors? Remove them from the circuit and use plug it into the sockets on the multimeter? I'm used to dealing with tubes, so i don't know how to test most of this stuff.
i checked the voltages on the output transistors. I measured from the mounting tab, which should be at ground, to each leg. For the tip142 the voltage for every leg were between 78 and 83 volts, and on the 147, the two outer legs were both around 80 and the center leg was at 0. What does this tell me?
I also used the audible continuity test on my DMM. With one lead attached to the grounded mounting tab, there was continuity between all three legs of both transistors. As far as i can tell, they're both blown.
Now i'm a little confused. The schematic shows the two red leads from the transformer secondary going to the rectifier, but these are at 56 volts. The schematic is a combined drawing for the 65 and 120 watt versions of this amp. The third green wire is a center tap. Should i be using one red lead and the center tap going to the rectifier? Thats the only way i see to get the voltages low enough to have 40 volt rails and not 80. then should the third wire, the unused red one, be grounded where the center tap used to be?
My friend brought me this amp in a few pieces. He had tried to diagnose the problem himself, and he didn't pay attention to where the leads were originally.
Hi shinychrome, calm down, we'll help you.
Please post the schematic (or a direct link to it) so we are all talking the same, and start by building a series lamp bulb current limiter.
Look at my very poorly drawn example.
You should use a 40W to 60W lamp as current limiter, and *no* speaker connected.
Good luck.
I'd love to post the schematic but i had to get it directly from crate. I had to sign a waiver agreeing not to post it anywhere before they would give it to me. I can't find one online anywhere.
Ok, don't *post* it anywhere, but show it to a technician who is helping you repair it.
That falls well within the fair use of what they kindly gave you.
Just click on my profile and you'll see my email address.
http://www.ssguitar.com/index.php?action=profile;u=357
Good luck.
JMFahey
OK, first, the transistor tabs are NOT ground. That is why there is an insulating washer under each one and a plastic step washer on the mounting screw. The metal tab is not grounded. Use chassis as your ground point.
You didn't have to remove the power transistors to test them for shorts, but since they are out... Each one has three legs. That is three combinations of two legs. Are any pair of legs shorted together?
You will find that the center leg of the transistor is connected to the mounting tab. In each case that is the collector of the transistor.
When installed, the tab of one ought to have +40v on it, and the tab of the other ought to have -40 on it.
I am not sure what you mean the two red wires are at 56v. 56v from one to the other? That sounds right. But they are not 56v to ground. I bet you have about 28v to ground from each one. Look at the schematic, there is a center tap on that transformer and it is grounded.
If you rectify and filter 28VAC, you get just about 40VDC on the nose.
Even though you assumed wrong that the transistor tab was ground, you still got readings of about 80v betwen two of their legs
, so I bet the +/-40 are OK.
I didn't remove them, i tested for shorts while they were in the circuit. I guess that probable wasn't an accurate test. However, i did a continuity check from the metal tab to the heat sink. Both had continuity, despite still having the spacer under them. Could this have been from over heating when they failed?
On that note, was overheating probably what caused them to fail? There's a pretty weak heat sink on the transistors, so i want to try and make sure this amp is gig ready and won't just fry itself again.
The insulating piece is under each transistor? Pull the mounting screws. Are the tabs still shorted to the heatsink with the screws out? Each screw MUST have a plastic shoulder washer on it. Otherwise the screw shorts the transistor to the heatsink.
The shoulder washer - or T-washer if you prefer - os a plastic washer with a short tube projecting out one side. This bit of tube sheaths the screw shaft as it passes through the transistor hole. It keeps the screw cenetered in the hole and insulates it.
The heatsink is brounded to the chassis. The tabs of the power transistors are connected to the power rails. If the tabs are griounded to the heatsink, then you have a dead short across your power supplies. Fuses blow.
Even with the screws removed and the transistors pulled up so that they are not touching the heatsink, i still read continuity from the tabs to the sink. Could they be shorted through something else nearby?
I'd like to know what you call continuity.
Set the digital multimeter on the diode scale and please take 4 measurements.
To avoid being *too* wordy, when I sat Red or Black tip to some tab, the other one goes to chassis ground:
1) Red to TIP142 tab=?
2) Black to TIP142 tab=?
3) Red to TIP147 tab=?
4) Black to TIP147 tab=?
and now
5) Red to TIP142 tab, black to TIP147 tab=?
6) Red to TIP147 tab, black to TIP142 tab=?
please cut and paste this table, replacing ? with the number on your multimeter screen.
Wait a few seconds, because when charging big capacitors it may take that to stabilize, all will show momentary "shorts" or short beeps.
7) on each of the four power diodes: Black to strip, Red to the other end should show around 500 to 600, and infinite (1+blanks) or at least above 1000 on the other way; answer only if everything is as this or some of them shows an abnormal reading.
Good luck.
1) Red to TIP142 tab=1081
2) Black to TIP142 tab=655
3) Red to TIP147 tab=723
4) Black to TIP147 tab=kept counting until it read infinite
and now
5) Red to TIP142 tab, black to TIP147 tab=kept counting to infinity
6) Red to TIP147 tab, black to TIP142 tab=1223
7)well the rectifier diodes are not individual diodes, it is a rectifier chip. i checked the zeners just after the filter caps in case that would tell us anything.
one read close to 800 with black to strip, and infinity the other way, the other diode read 640 and 840 respectively on the same measurements.
i looked up the data sheet for the rectifier. Its rated for 6 amps. that sounds a little low to be running at full power for a few hours. Is there a way to test this rectifier as if it were individual diodes?
oh, and what exactly was i measuring there? I like to know what i'm actually doing.
OK shinychrome, we are doing fine.
What are you measuring?
Enzo asked you to check a couple things, I noticed you spoke about continuity, which had to be defined a little more clearly.
A forward biased diode and a piece of wire both show "continuity" , yet one is working properly and the other is a short.
Unfortunately diodes and transistors often turn into pieces of wire. :'(
The difference between both is that when some current (in this case around 10 mA, the exact value is not that important) goes through a piece of wire, the voltage across said wire is "0" or a very low value, say 20 to 40mV.
Across a (good) diode you will measure around 650 mV.
That's exactly what the diode test setting in your multimeter does: apply some current through the probes and measure the voltage across them.
Mind you: a forward biased diode and a dead one will *both* show "continuity", but the voltage across them will not be the same.
Now to what I asked you to measure:
1) Red to TIP142 tab=1081<-good
2) Black to TIP142 tab=655 <-good (you have a protection diode there)
3) Red to TIP147 tab=723 <-good (you have a protection diode there)
4) Black to TIP147 tab=kept counting until it read infinite <-good
and now
5) Red to TIP142 tab, black to TIP147 tab=kept counting to infinity <-good
6) Red to TIP147 tab, black to TIP142 tab=1223 <-good
So far your transistors *look* good, no glaring shorts there, which would have shown as "0" or very low measurements or as diodes where thereb should be none.
TIPs 14x are confusing because they have an internal backwards connected diode, which makes you think it's bad.
Now to the power bridge: it *is* a pack, but you have 4 diodes in it, look at the schematic: you have access to each diode which is connected across 2 pins.
Disconnect the power transformer secondary and measure both ways across each pair of consecutive pins (not in diagonal), across each "side of the square" if you wish: you should get around 550 one way and above 1000 the other way.
Build a lamp limiter for tomorrow's tests.
Good hunting.
well the bridge rectifier checks out ok. It might take a few days to get a current limiter rigged up. I'm in a dorm room so i need to scrounge some tools and materials. anything else i can try in the meantime?
Hi shinychrome.
I guess you are quite close to the end of the trail.
You can make a bulb limiter with a desktop (incandescent) lamp and an extension cord, some black insulating tape and not much more.
I guess that you have some tools available or you wouldn't be working on your amp.
Of course maybe you can't get that if you can't get out of school or whatever.
To work safely you need that limiter.
The option, a Variac, is more expensive and harder to get your hands on.
still working on getting a limiter together. Haven't really had a whole lot of time on my hands lately. What else are you going to have me test when i get it though?
BTW, from what i can figure out, i just need the light bulb in series with the hot wire in the extension cord correct? Which color is this going to be?
Don't know the standard in your country, but you should find it in 5minutes.
Yes, you need the lamp in series with the "hot" or "live" wire.
Alright, so i've got my current limiting desk lamp extension cord gadget. Now what?
bump. Still there JM?
I'm going to pull the board and inspect the filter caps. There is a significant amount (about 82v) of ac on both transistors. There obviously shouldn't be any.
Hi shinychrom, back here.
No speaker attached, all controls on "0", plug your amp and with black probe grounded to chassis, measure DC voltage on both power rails (you should have around + and - 40V respectively) and DC voltage on the speaker out hot pin.
I don't understand your previous posting of 82VAC on the transistors, there's nowhere a point in your amp that can show more than around 30VAC relative to ground.
Please tell me what scale was the multimeter on, and what were the probes touching.
I'll check that again when i put the board back in a little later. I've got it out right now to check some of the other parts of the power amp and filter section. The measurement i took earlier was from the case of each transistor to the earth ground connection. My multimeter was either on the 500v ac or 200v setting. Probably the 200v. The decimal point can be testy on this meter, so now that i think about it, its possible that it was 8.2 volts. But it was exactly the same on both transistors, and completely steady. I know there should be less than a volt of ac. The filter caps seem to be fine though. using the highest resistance setting on me meter they both show charging and discharging normally. There is some heat discoloration around the diodes in the filter circuit though, mainly around D28. It measures okay though as far as i can tell. I tested it the same as the transistors earlier. D28 read around 830 one way and infinity the other, D29 tested about 830 one way and 750 the other. There are a few other spots that look like they've been overheated. I'll figure out whats on the other side and test whatever i can.
Ok so i found a definite problem. the tracks connecting to Q19 are damaged, and one is completely broken. attempting repairs now.
Hi shinychrome
I don't like your D29 testing low both ways, just replace it.
Cheap available anywhere multimeters with only a couple AC voltage scales, typically 200 and 500 or 750 VAC do not really measure AC but use a diode in series with the DC measuring circuit, and approximately double its sensitivity because they expect to see one demicycle on, one off.
When you touch DC with that probe, the meter shows about double the DC voltage, so on a point that has , say, 40VDC you read over 80V "AC" which was never there to begin with. :duh
What is that scale there for?: to measure 110/220 V power lines, transformer secondaries, etc. ; *maybe* power output across a speaker, *never* to follow signal voltages across a preamp nor to measure ripple.
I see this error everywhere on this Forum, not really your fault but those greedy manufacturers (or merchants) :trouble
I will replace that diode. Q19 was the problem. I had to run short jumpers because the tracks were too badly damaged to repair. It looked like a bad solder job from the factory. But the amp appears to be working fine now. No DC at the speaker output jacks, so i connected a speaker and everything seems to be back to normal. I'll hook it up to its original speaker in a little while so i can get it a little louder. My bench speaker is really low powered.
So two questions: one, what could have caused this to blow? could it just have been a little heat from playing at full volume combined with the bad solder joint?
and two, is there a better way to repair the traces on the board? Its lifted off the board for a good half inch on all three legs.
1) GOOOOOOOODDD , it works :tu:
2) maybe there was a short, the thin underdesigned traces worked as fuses.
The "bad solder job" probably comes from overcurrent melting the solder while it was busy burning tracks.
3) follow the tracks backwards, cut and discard semi-molten, unglued parts and scratch the board to get rid of carbon deposits.
Get o a healthy piece of track, scratch carefully the solder mask paint from a 1/2" section and solder there a piece of wire.
When you are happy with the repair, fix the added wire with a drop of contact cement or similar to avoid it wiggling and tearing the track section
4) Play at will.
So what could have caused this meltdown in the first place? Just overheating on this transistor? It has a really dinky heat sink.
Rather than only the transistor overheating (which also happened), eventually it shorted and that track passed 10 times what it was designed for.
*That* burnt it.
There's something else, although they will *never* recognize it to you, many try to "repair" the amp with a higher value fuse, or even some cigarette paper or bypassing it.
It never blows, so some part of the amp must, to stop that overcurrent.
Either that, or your house burns.
I've made and sold over 10000 amplifiers (in 40 years) and obviously I backup them, both under guarantee and years afterwards.
When the original fuse blows, and there's no foul play, that's to say, the original type and value is used (as all amps state clearly written on user manuals and back panels), 90% of them only have output transistors blown, sometimes only one, the board looks virginal.
I use regular (fast) blow fuses and always supply an extra one or two taped to the power cable, so it's always there if needed.
Sometimes amps come with no fuse at all (they pull the incriminating evidence) but with burnt tracks, wire resistors no simply "cut" but when open show sizable balls of molten wire, grey cracked solder, resistors burnt beyond recognition, cracked transistors, the full horror show.
All those used way oversized fuses.
I'm starting to include hidden in-transformer picofuses just because of that.
So now you can imagine what might have happened there.
Was the short just something temporary? Because the transistor itself tested fine. It didn't seem to be any worse for the wear.
You got lucky.
Under designed tracks worked as fuses and saved you.
If the transistor tests well and works well ... leave it there.
Just write a "?" sign besides it to remember and if someday you are bored and have a spare, replace it, just to be 110% sure.
As is, I'd trust it 95% which is good enough in my book.
I realize this topic is ancient, but I have the same issue with an xt65r.
I picked it up cheap today, and It just hums...I checked the fuse, and it had a 5A instead of a 2A...So I am guessing too much current cooked something on the board.
I believe I will try Q19 first in hopes of getting lucky.
If that doesn't take care of it, I will perform all of these same steps as shinychrome0 was directed.
My fingers are crossed.....
Ok, I see no obvious marks on the back of the board as with shinychrome0's amp...and the tracks look undamaged.
Time to start from the beginning of this thread. Any info aiming me toward the most obvious failed parts would be appreciated.
I'm going to leave you to JM 'cos he already has the circuit for this amp, but in the meantime you can make up a limiting lamp as per;
http://www.ssguitar.com/index.php?topic=2093.0 (http://www.ssguitar.com/index.php?topic=2093.0)
...because you are going to need it.
Disconnect the speaker. Do any components seems to look overheated, browned off (if so, post photo)?
Replace the fuse with one of the correct rating, but do not try to power the amp up again at this stage.