Ampeg SVT 3 pro.

[J M Fahey]

I wonder because the meter does not even *have* enough digits to show 1 or 6 ohm values.
Meters show the most significant (important) digits, left to right, so :
20M ohms = 20 Million ohms = 20000000 , you need 8 digits to display that number in full and in that case, the rightmost digits means units (0 to 9) so *there* you might see a "1" or a "6" , what you claim you read.

BUT, a typical 4 digit multimeter , which can display from 0000 to 1999 (4 digits, remember?), will show only the 4 leftmost (in green), will not display the ones to the right, and definitely not the units marked in red, while trying to display 20M:
20000000
Even a 5 digit multimeter is not enough.

Most practical way to test semiconductors is to use the diode test function, which although part of the resistance scale does NOT display am "ohms" value (confusing, isn't it? ) but in fact passes an undisclosed value of current (a couple mA) through the junction we are measuring and displays the *voltage* dropped across it.
Around 0.6 to 0.7 for silicon semiconductors.

Note to other Techs: this is a somewhat simplified explanation, showing the lack of available digits, please donīt start a discussion about 20M being actually displayed as 19999999 because thatīs not the point now and might confuse the OP even further.

[g1]

i was always told when you measure a diode or a transistor ,set your meter to the highest ohms range you can. Is this not the correct way,or am i a victim of more mis information?

This would be the case if checking a semiconductor for leakage, otherwise use the diode test function. For resistance, you will get the most accurate reading when on the appropriate range.

[catstrat]

SVT.

Thanks everyone. You all have been giving me info that i just was not aware of. Tube amps have a higher voltage generally so i know how to check those. Even though i'm getting better with solid state, some of it is still a mystery to me. The only way to learn is jump in with both feet and get advice from reliable sources and try not to blow anything up. That's why i belong to this forum. I'm probably going to pull the o/p trans out of the circuit and see what i've got that way. By the way, does anyone on here have experiance with a Pioneer M90? Again thanks.

[52 Bill]

I'm probably going to pull the o/p trans out of the circuit and see what i've got that way.

Another thing to know about is that the outputs are static sensitive devices and need to be handled accordingly.

[catstrat]

SVT

HI 52 Bill. That's one thing i did learn while working at Compaq computers,while it was still Compaq. I don't know what it is now. I have a variac and it comes in handy sometimes if the amp has no relay. Generally i'll increase the variac voltage to where it's pulling maybe 2 amps from the wall and i wait and watch. The shorted transistor/s will start to heat up quickly. I'll remove the transistor and try it again. If the amperage draw is normal then i know i'm on the right track. I know it's un orthodox but it's worked for me in the past. I'm pretty sure about this, if you have a shorted o/p device then change both the drivers and the pre drivers on that side along with the o/p transistor. If you have one bad one i hear it's best to replace all on that side. I get customers that bring me stuff and they tell me, i want it fixed but, i don't want to spend more than 100 bucks. I try to be helpful but sometime i take beating on some of this stuff. Some of the ss guitar amps aren't even worth fixing but they want it fixed. That's enough of that rant. If any newbies on this forum are following this thread, any advice you get from the guys here, you can take to the bank. They've helped out of a pinch many times. Back to the SVT, i'll get back on it Monday and i'll report back. Thanks everyone.

[Jazz P Bass]

A really nice companion to the variac is a 'lamp limiter'.

BTW, a solid state amp should never pull close to even an amp at idle.

[catstrat]

SVT. Again.

Good Sunday everyone. I did some more test this morning. The positive voltage rail is at + 65 volts. The negative rail is at - 64 volts. The schematic shows 65 +/- volts on both rails. I changed a few t092 transistors on the OP board and at least now, it passes signal but the hum is still there but it will go away when you tap the board. The board itself is badly microphonic. You can lightly tap any part of the board on the tube side and it makes a loud thump. Also the bias is affected when you do this. I check the bias according to Ampegs recomendation which i believe is about .8 amps through the variac at idle. This is the way i've set all svt 3 bias in the past and it worked. No returns, no complaints. I'm still leaning toward a bad ground somewhere as the heart of the problem. I've checked for cracked solder joints and even reflowed the solder on ones that looked funny. Maybe i should touch up all solder points on this board. Are there any hidden ground points on this board i don't know about? I'm going to run a signal into the power amp in and see what happens. I think i've done that before but i'm doing it again but i'm writing it down this time. Maybe a bad bias trim pot? Thanks.

[J M Fahey]

Ok, letīs start actual troubleshooting instead of random parts replacing.
So far you reflowed solder and replaced "a few TO92 transistors" but you donīt tell us which or give a reason for it.
Yes, it might be a bad bias pot .... or any other 100 or so bad parts.
Only troubleshooting can tell.
Letīs start from the beginning ... do you have a scope?
If not, Iīll suggest a software one you can load in a notebook and a suitable safe probe/attenuator so you donīt blow the soundcard input.

[catstrat]

SVT.

Here is my decision. I'm not scared to tell someone that it's beyond my scope to fix it. I've done it before, especially when i have stuff waiting that i know i can fix. But for the record i replaced Q 6 and Q2, all the drivers and the predrivers, don't remember the board number for those. I sprayed the board with some cleaner and tested it with a signal into the input jack, and it worked. The hum was gone, it was pulling about the right wall amps according to Ampeg and it was working normally. But after 3 and a half minutes, the relay opened and the heatsink once again was very hot to touch. So i'm pretty convinced now it is a shorted OP transistor. Keep in mind, they don't want to spend more than 150 bucks on this. They brought two and the other is a Hartke HA 3500 that was blowing fuses. I fixed that one. I ran a signal into the Ampegs Power amp in jack and i got a good clean signal. I do have a scope and let me go and scope the output and i'll let you know. Before i did everything i did to it today it wasn't even passing a signal. Let me get back with you shortly. Thanks

[catstrat]

I just scoped the output with the scope on 1 volt and a 400 cps signal into the input, the op wave looked even and equal on both the positive and the negative half. Except the positive half was about .200 MV stronger that the negative half. As usual, after about 3 and a half minutes it shut down. The signal just stopped without any abnormal things like spikes or anything like that on the scope. Thanks.

[catstrat]

AS it stands right now according to the scope wave and everything else, the amps seems to be workng fine, except the shutting down. There is another thread started by some else about his amp running hot. I'll go and visit that post and read hrough it. Again thanks.

[MarkusBass]

But after 3 and a half minutes, the relay opened and the heatsink once again was very hot to touch. So i'm pretty convinced now it is a shorted OP transistor.

I doubt it. Shorted transistor causes open fuse almost immediately. It's rather incorrectly set bias. Have you check bias in the amp after it was fixed? For starter just measure voltage drop on ballast resistors (0R33). And turn on the amp for 10 seconds only (just to make the measurement). Otherwise you may get shorted output transistors.

Mark

[Jazz P Bass]

My thoughts are that 0.8 amps, mains draw at idle, seems a tad excessive.

That's 96 watts while the amp is doing nothing.

Where is that wattage going?

Into the heat sink.

I would try lowering the bias.

[catstrat]

SVT.

Hi Jazz P. .8 amps is a lot for an amp just setting there. But i have the bias adjustment procedure from Ampeg on how to set the bias. I'll have to look at that again and make sure i'm reading it correctly. I'll have to verify it's not a mis print. Every SVT 3 i've worked on i've set the bias to .8 amps or close to it.. No problems and no returns. I may try it the other way which is using the scope. That may be the better option in this case. Both methods come from Ampeg. Actually Mark, it was still overheating with the bias set to the lowest point possible. Every shorted output tranistor i've run into almost always is causing a fuse to blow. But again i've not enountered every failure mode out there. Yet. To answer a question from earlier , i changed the drivers, pre drivers and the bias servo transistor because i had them in stock and they don't cost a whole lot. I still don't understand why it's working perfectly normal, or seems to be, and then it just shuts down after 3 and a half minutes. I must have done something right. When i got it it wasn't passing signal at all.It does make you think it's a bias issue which is where i was headed in the first place. Most of the shorted o/p devices i've found generally cause the heat sink to overheat in the area around the shorted transistor. That's where my variac comes into play. But this, seems like the whole heatsink is getting hot equally on all surfaces. Anyway, i talked with the customer this morning and gave them an option, of paying the bench fee or giving me an OK to replace all the o/p trans. I told them that would affect the final cost. I just hate to replace all the o/p trans and the problem is still not fixed. It's up to them right now. Thanks

[catstrat]

SVT

Ok Jazz p. I just double checked the Ampeg data i have and it is .8 to .85 amps from the wall. The entire bias procedure is to verify it's pulling the correct amps , .8 to .85. Inject a 20 mv 1000cps signal into the "poweramp in" jack loaded with 4 ohms. Adjust the bias until there is barely a crossover notch visable on the waveform. Disconnect the generator and scope and verify it's still pulling between .8 and .85 amps from the wall and you're set. We are talking about an SVT 3 pro., not that that matters. A bad coupling cap maybe, or worse yet, a broken tracing. It has some zeners in the circuit. What do they do? Probably a voltage regulator of some kind. I've found leaking zeners in the past. Not knowing that much about solid state stuff is why i pester the heck out of you guys. A little recap, in the beginning it was a bad hum and it wasn't passing signal at all. The odd thing, you could hear the fan motor through the speakers. That's why i started to suspect a ground problem. Thanks. I'll wait and see wait the customer wants to do.