Sho Bud MAJOR DC offset

[Prattacaster]

I'm currently working on a Sho-Bud single channel amp. I had to replace the output transistors along with everything that has a red square around it in this schemo: show_bud.jpg picture by Prattacaster - Photobucket
I've quadruble check all the components orientation and I still get a whopping 3.2vdc across the speaker while using a 25w bulb limiter. Rail voltages are about 32vDC. I am wondering if a leaky cap in the preamp would cause this or maybe someone knows a better place to look.

I replaced the components with the closest match I could get except for the output transistors I used MJ15003's and have an extra pair in parallel with the replacements

Things worth mentioning: The top NPN driver gets fairly hot - not hot enough to burn but close. The PNP driver stays cool. I also added a zobel network. I did not replace the PNP matched differential pair at first but i was getting wacky voltage readings so I replaced them, when I did, the voltage readings were drastically improved and so was the DC offset, but the offset it still way to high. I also had to replace the trimmer pot to get a functioning bias circuit, before I replaced the pot I could not bias any cooler than 60mA

With no load I can bias so that i get NO DC offset, however with i plug in an 8 ohm load then it jumps back to ~3vDC.

[Prattacaster]

This was said about the Citation 12 which is pretty close to the Shobud.

I decided to fire up the completed unit on a variac instead of merely plugging it into the
wall' and was dismayed to discover that both channels exhibited extremely large DC
offset when I applied AC power. After some weeks of rechecking my work and findingno errors, I finally decided to connect it up and plug it into the wall anyway*. When I did,
the woofers on my speakers plunged fully forward as if attempting to escape the
amplifier, but after a few seconds they returned to normal and music appeared. Thus I
discovered one of the Citation's few faults: an enormous subsonic turn-on thump.

[Enzo]

OK, so the amp is not burning up or blowing fuses or anything? Then get rid of the bulb limiter and see how it works.

[Prattacaster]

Gotchya

[olddawg]

If you still have a dc offset, is the dif-amp a dual package with 5 leads or two discrete transistors? If they are discrete, they probably need to be a beta matched pair. In the old days they would have a dot of red or white paint on them. If your multimeter has an hfe function you can match the diffamps yourself, but you will need a bag of 6 or so transistors. Also you can set the crossover notch with a very low 1khz tone and a scope.

[Prattacaster]

I could not go thru with full wall voltage testing, I'm not brave enough yet. I popped in a 40w bulb and the DC offset increase to 5vDC with an 8ohm load. I took some voltage measurements without a load. Here they are: show_bud-1.jpg picture by Prattacaster - Photobucket
All the posistive and negative voltages are correct as far as what they read on my DMM. Its seems wacky that the bias voltage divider would be 13.7vdc and yet on the Base of the NPN driver I'd have -3.3vdc. Also the Differential pair emitter voltage seems off.
I replaced the bias diodes. This did nothing to change the outcome.

If they are discrete, they probably need to be a beta matched pair

They are discrete not matched. This shouldn't cause such a huge offset tho, right?

[olddawg]

They are discrete not matched. This shouldn't cause such a huge offset tho, right?

Sure...a mismatched pair can cause an offset to rail. Can cause full cone inversion and smoke voice coils. I had a friend jumper the speaker protection relays on an amp with a diffamp input. I said, "Im not suprized you don't have a bunch of cooked speakers." He opened his garage and had 6 blown cabs. Check your emitter resistors. Look for an open 500k. You can tickle the diff pair with some freeze and it will increase/ decrease the offset.

[Prattacaster]

Sure...a mismatched pair can cause an offset to rail.

Thanks. I did have the original matched pair in before and was still getting about the same offset.

Check your emitter resistors.

Are you referring to the Output emitters or the diff. pair emitter? I've replaced the output emitter resistors with 7 watters, i havent check but running with a 25w bulb I doubt I blew one of those, I check for ya anyway.

You can tickle the diff pair with some freeze and it will increase/ decrease the offset.

I have to say I've never tickled a pair before.

[Enzo]

If the diffy pair was way off, I would think it would show up without a load as well. How would the diffy pair know if there was a load or not?

[Prattacaster]

If the diffy pair was way off, I would think it would show up without a load as well. How would the diffy pair know if there was a load or not?

Ok, so that fact that the DC offest goes from 3vdc with load to -4.5vdc w/o load would mean its not the diff. pair?

With a load I cannot bias under 82mV which means I've got 248mA at idle. If I do not have a load a can bias all the way to shut off.

[Enzo]

With no load I can bias so that i get NO DC offset, however with i plug in an 8 ohm load then it jumps back to ~3vDC.

Ok, so that fact that the DC offest goes from 3vdc with load to -4.5vdc w/o load would mean its not the diff. pair?

So which is it? You either have no offset unloaded or you do.

With a load I cannot bias under 82mV which means I've got 248mA at idle. If I do not have a load a can bias all the way to shut off.

And I am not sure what this means. 82mv of what, where?


If outputs are getting hot, then they are conducting when they shouldn;t be.

Working with no load. there are two bias diodes in series. WHat voltage is dropping across each one? The bias control appears to be 100 ohms. With power off, does it indeed adjust from about zero up to about 100 ohms? Set it to zero ohms for the work ahead.

There are two 3.3k resistors, are they OK and close to value? There are three 100 ohm resistors, are they OK and close to value?

WHich way do you turn the bias control for lowest current? The coldest setting? Set it there. Now, without a load, is there DC offset at the speaker leads? If there is now, we need to cure that before we worry about bias.

[Prattacaster]

And I am not sure what this means. 82mv of what, where?

82mV across the .33 ohm emitter and collector(negative rail) power resistors.

If outputs are getting hot, then they are conducting when they shouldn;t be

The outputs aren't getting hot but the driver NPN is very warm, not quite hot enough to burn maybe 120degrees F. The PNP driver stays cool.

There are two 3.3k resistors, are they OK and close to value? There are three 100 ohm resistors, are they OK and close to value?

Yes they are all very close to value, I've checked everything at least three times.

With no load I can bias so that i get NO DC offset, however with i plug in an 8 ohm load then it jumps back to ~3vDC.
Ok, so that fact that the DC offest goes from 3vdc with load to -4.5vdc w/o load would mean its not the diff. pair?
So which is it? You either have no offset unloaded or you do.

Like I said, with no load a can turn the trimmer clockwise and get the DC offset to come within the acceptable range, but at that bias point I would kill the outputs. With an 8ohm load the lowest I can bias is 82mV across .33ohms or 248mA and I have +3vdc offset.

WHich way do you turn the bias control for lowest current? The coldest setting? Set it there. Now, without a load, is there DC offset at the speaker leads? If there is now, we need to cure that before we worry about bias.

Counter-clockwise, and there IS a negative DC offset of -4.4VDC. I agree that I cannot go on until I get rid of the offset. Thanks for all your help so far.

[J M Fahey]

a) disconnect load until further instructed.
2) the right diff transistor voltages (Please label all transistors and other parts for clarity (Q1 .. 2 R1 .. 2 ... etc.), repost the relabelled schematic) are impossible, so something is very bad.
You show the same + 0.63V both at its base and emitter.
Either it's shorted base-emitter or a rogue solder drop is bridging two pads it shouldn't.
Also itīs connected to -4.5V, yet it shows +0.63V.
Either its shorted as I mentioned before or the path from the speaker rail and its -4.5V is broken.
3) the image is somewhat small for these old tired eyes, can't you post a somewhat larger one?
Good luck.

NOTE if everybody posted accurate voltage readings as you and the Rogue amp guy are doing, repairs in this Forum would be *much* faster.
jm2c

[Prattacaster]

Also itīs connected to -4.5V, yet it shows +0.63V.

I realized last night where that discrepancy originates from, the -4.3vdc on the output is in reference to ground where as the .63v on the diff. pair is in reference to the output. So..... I did take accurate readings but I mismatched my reference point, which is good for nothing. I will repost tonight with correct voltages in reference to ground.

I found something else that might offer some help, when testing the voltage between speaker positive and ground without a load I get -4.5vdc until I turn the amp off... at switch off it quickly swings to about +5vdc and drains down to near 0.0vdc.

I remember when I have a load plugged in my output collector voltage and the positive rail is only 6.6vdc which tells me I am pulling way too much current; indeed I am because I have .082mV across the .33ohm output resistors.

I mentioned that the NPN driver gets hot, so I measured the voltage drop over its 100ohm emitter and it is only pulling 6mA. Thats ok right?

[Enzo]

You skipped over the bias diodes. VOltage dropped across each?

Seriously, if you minimize the resistance of the bias control and set the amp to coldest, you should then be able to sort out the offset. Changing the bias setting and watching the output DC volatge change isn;t telling us much, other than that something is conducting. The amp SHOULD be able to sit there stable with zero DC on the out. In fact, you should be able to short across those diodes - effectively shorting the bases of the 40409 and 40410 together, and the amp should work. There would be crossover distortion, but the amp would center at zero if it was otherwise OK. SUch conditions might make it easier to spot what is wrong causing the offset.

having said that, proper power supply is critical. A functioning amp should center between the power rails. If they are both the same voltage, then that would mean zero on the output. But if you have -32v and +6v, then center becomes something like -19v.

Now my guts talk to me. They could be TOTALLY WRONG, but what the hell. You mention the +32 getting hauled down to +6. And pose that that means you are drawing too much current from it. MAybe. But is that too much current from a healthy supply? Or too much current from a wounded supply to handle? You mention that you get 6v at the collector and + rail. Are you assuming the rail is +6 because it is connected to that collector? Did you measure it back at the V+ filter cap and fuse? A resistive connection could be the source of your voltage drop. If the main supply is dragged down to +6 at the filter cap that is one thing, but if that rail stays up at the filter but falls to 6v at the transistor, that is quite another.

I am pulling way too much current; indeed I am because I have .082mV across the .33ohm output resistors

OK nitpicking, I assume you mean .082V OR 82mv. But do the math. .082/.33 = 248ma. Now that may be a lot more current than the idle adjustment specified, but I don't think 1/4A is enough to load down that power supply if it is working. With those V supply voltages and an 8 ohm load, that power supply needs to source SEVERAL AMPS, not a few milliamps. SO I am led to think we have either a power supply problem or a connection problem.

(Due to an issue with my browser, I will break this into another post, sorry)