SWR California Blonde troubleshooting tweeter/power amp repair

[DrGonz78]

The amp passed signal originally on first check but it sounded like a nasty sharp cutoff fuzz distortion.

When I opened the amp it was clear that R17 was fried which I replaced. I then tested the speaker but admittedly forgot to test the tweeter. Anyhow I was testing voltages and realized some of the legs on the 2sc3264's were cracked right at the solder joints, not to mention a cracked leg on Q4 too. This seems like a common issue with SWR amps and I was able to reconnect the legs of those transistors. At that point all the transistors were testing good, just 15mv on the output and it all seemed ready to check under load. At that point with the amp connected to the speaker output I smoked R17 and apparently this time shorted Q6 base to collector. At this point I got suspicious of the tweeter and pulled it out for testing. The tweeter is 16 ohms but dc resistance reads OL and capacitance reads 8.9pf. I thought perhaps this tweeter was a piezo and that it is a capacitor type. I remember that from another amp I worked on but even then DC resistance measured 500k on that piezo. So should this type of tweeter read roughly 16 ohms with my meter? I am almost certain it should but just need someone to confirm my thinking. I don't deal with tweeter's very much evidently...

I attached a picture of the tweeter and here is a place that sells replacement diaphragms.
Celestion CR153, CR122, RTT50 HF-50 Diaphragm T5501/R 16 ohm - Speaker Exchange

[Enzo]

Piezos will not measure 16 ohms, they measure like a cap. Also, if you find replacement diaphragms, that means it is a dynamic tweeter, piezos have no diaphragms.

R17 is 100 ohm? Only one path for failure current that burnt it up, and that is through Q5, which is then suspect. That makes output transistor suspect as well.

[DrGonz78]

Thanks Enzo that clears up the tweeter question. Funny thing is that Q4-Q8 all tested good before connecting it to a load. Strange that Q5 still tests good even now using diode check mode on my meter. So Q5 is probably leaking? I am ordering replacement transistors Q4-Q8 to start fresh and a new tweeter diaphragm. So strange how so many of these SWR output transistor legs just break off right the solder joints.

[J M Fahey]

Yes, that´s a dynamic (voice coil + magnet drive) tweeter.
Piezos measure like .1uF to .22uF capacitoors usually, quite large caps which load an amp output, some don´t like that at all so they are often used with a series resistor, between 10 and 100 ohms, just to "separate" that annoying capacitor from amp out.
The very low 8pF you measured is cable capacitance.
Yours has a full fledged CLC crossover, normal in a non piezo tweeter.
Personally I´d add a 12V, 12W or 15W car parking or courtesy lamp (with matching lamp holder) in series with the tweeter, to protect it.
Worst case solder it in place.

[52 Bill]

I don't remember that those came with Celestion tweeters, but I don't seem to remember a whole lot of things now-a-days.

I normally would not solder in the output transistors until the board was mounted and the transistors were also mounted. That way there will not be any tension on the leads. Even if the outputs don't need to be changed, you can remelt the solder joints to relieve some of the stress at the joint.

[J M Fahey]

Those power amp PCBs are partially held by the power transistor leads, combine that to typical MI amp moving around and vibration, doubly so in a Bass amp, and you´ll find higher than normal stress cracking.

[DrGonz78]

Thanks for the help guys! I like the idea of inserting a lamp in series to protect the tweeter. This will be my first time installing a speaker diaphragm. No doubt I might have some more questions once the part is delivered.

I normally would not solder in the output transistors until the board was mounted and the transistors were also mounted. That way there will not be any tension on the leads. Even if the outputs don't need to be changed, you can remelt the solder joints to relieve some of the stress at the joint.

That is exactly how I was thinking to install the new output transistors. Mount them flush up to the heat sink then slip the power amp up into position and solder once fully mounted with screws. I did resolder the connections to restore continuity on about 4 contacts of the output transistors that are in there now. I hit all the joints to make sure. It was strange taking voltage checks and to not see voltage where expected on the legs of the transistor. Then some ohms readings made me look real careful to find the cracked off legs. I later read a post of yours, Bill, that mentioned this type of problem on some SWR amps.

[Enzo]

Diapragms. Don't remove the old until ready to go, it will keep dirt from getting in the gap. FOld a triangle of masking tape sticky side out. Fold like a flag. Use the corner points, slide them into the gap, and run around the gap, to clean out any dirt or soot. Do this as many times as it takes for the tape to come out clean.

Some diaphragms only assemble one way, others can be put in either way, so make notes or take pictures as you take the old apart to make sure the phase relation is maintained. IE plus and minus terminals to the right places.

[DrGonz78]

Diapragms. Don't remove the old until ready to go, it will keep dirt from getting in the gap. FOld a triangle of masking tape sticky side out. Fold like a flag. Use the corner points, slide them into the gap, and run around the gap, to clean out any dirt or soot. Do this as many times as it takes for the tape to come out clean.

Some diaphragms only assemble one way, others can be put in either way, so make notes or take pictures as you take the old apart to make sure the phase relation is maintained. IE plus and minus terminals to the right places.

Thanks for that added info Enzo! I did not take it apart yet and definitely am waiting to remove the diaphragm once I have the replacement ready to go. Oh yeah I have OCD when it comes to something where I get only once chance to get it right. I will no doubt be online here asking more questions before this thing is over. Thanks again to everyone here for the help!

[Dave H]

D4 should limit the voltage across R17 to below the smoke threshold so if R17 is smoking D4 must be open circuit.

[DrGonz78]

D4 should limit the voltage across R17 to below the smoke threshold so if R17 is smoking D4 must be open circuit.

I replaced D4 along with the charred R17 when I first received the amp. Funny thing is that I tested the removed D4 component and it was just fine. I thought the very same thought when I first looked at the burnt R17 too and was very surprised that D4 was still intact.

[DrGonz78]

So I am in the process of replacing transistors and will start testing the amp out soon. Quick question... I want to start out by setting the bias to the coldest setting. By measuring voltage across the ballast resistors would that tell me how hot or cold the bias is set? Lowest millivolts across ballast resistors will indicate bias set to coldest setting is what I am thinking. Also, there is no bias adjustment procedure to follow on this amp. So I guess it is then set by monitoring a signal through a scope and then setting it based on the crossover notch point.

Also, later in the week I will tackle the tweeter replacement. Anyone have a good recommendation for what type of adhesive to use to glue it into place?

[Enzo]

Think about how the bias works. The bias is a voltage space between the bases of Q5 and Q6. The wider that space, the more the outputs turn on and conduct current. Q4 controls this. If we short E to C on Q4, we pull those bases close to each other, the coldest setting. IN a working circuit, the closest to that short would be turning the transistor on as hard as we can. To do that we want to drag the base towards the collector, or in other words away from the emitter. If I short base to emitter, it turns Q4 off, which would me the most voltage between the driver bases, not what we want.

So look at the bias adjust VR1. If we zero out the VR1, it leaves 470 ohms between base and emitter - as close to zero as we can get. SO THAT would be the hottest setting. Schematic says that is Clockwise. So set the thing all the way counter clockwise, now the resistance between emitter and base is max, the most Q4 ca conduct. STart there.

I have no idea how many milliamps you will find, but I find the coldest setting based on circuit function, rather than a meter.

[DrGonz78]

So look at the bias adjust VR1. If we zero out the VR1, it leaves 470 ohms between base and emitter - as close to zero as we can get. SO THAT would be the hottest setting. Schematic says that is Clockwise. So set the thing all the way counter clockwise, now the resistance between emitter and base is max, the most Q4 ca conduct. STart there.

Okay great I see the light! I measured resistance at Q4 base to emitter with pot turned all the way counter clock-wise it reads 1470 ohms, coldest. Then turned all the way clock-wise it reads 470ohms, hottest. Sometimes I get rusty in my own thinking but Enzo you always get my gears spinning again, thanks!!

[DrGonz78]

So I got the tweeter diaphragm installed into the driver enclosure. I had an idea to use brush on electrical liquid tape to seal up the space where the terminal tab meets the plastic shell of the driver. I am pretty sure this stuff should work fine in this application. When dry this stuff gets pretty hard in place. Not as hard as epoxy but I think it should be good enough. Anyone have a reason not to use the liquid electrical tape in this manner? Anything better to use? Here is a pic to get a better idea of the the little area that I am talking about.