SWR California Blonde troubleshooting tweeter/power amp repair

[J M Fahey]

Normally diaphragms are not glued but air leaks are avoided by cardboard or rubber gaskets.
Your picture shows a small crack in the plastic cover, you should seal/rebuild just that before mounting, and nothing else.

[DrGonz78]

It was all gooped up with some black sealer that was definitely holding it in a place as it took a little bit of prying to come free. I only removed what was there with the intention to copy how it was original constructed. To be honest I am pretty sure they put some of this black glue there to hold it in place better. Mainly to reduce the stress from pushing or pulling off the speaker terminals and to seal it up too. Watching vids on youtube I saw many types of tweeter rebuilds and this one seems rather cheaply constructed in comparison.

[J M Fahey]

Ok, if they used some kind of rubbery sealer, it must not have been something adhesive like liquid electrical tape but a waek ne, good for sealing but bad on adhesion, like the "liquid gasket" which is used in car engines.
It stands preasure, seals minor cracks, but is designed to be pulled away without much trouble and leaving no residue behind.

IF Celestion used anything stronger, either it was a Friday afternoon job where the proper one run out an there was no time to get more, or simply they consider this tweeter disposable.

[DrGonz78]

This seems a bit unusual... I am testing a sine wave at 1k 100mv into the input to look for cross-over distortion but I seen none. I have the bias set to the absolute coldest setting and the sine wave is just perfect. If I crank the gain knob up I can see the tops of the wave form clip, which I would expect. I guess I was expecting to see some cross-over distortion. Also, I think someone had been in the amp before and had set the bias really hot. I should have been more careful from the start of the repair now looking back with hindsight.

[Enzo]

Run your signal up, but not to clipping. Scope the output. Now cranko teh scope vertical up, we only need to see the part of the waveform crossing zero, we don't need to see peaks. This wil make any crossover notch larger on your scope scrteen.

[DrGonz78]

Run your signal up, but not to clipping. Scope the output. Now cranko teh scope vertical up, we only need to see the part of the waveform crossing zero, we don't need to see peaks. This wil make any crossover notch larger on your scope scrteen.

I tried this again and there is no crossover notch at all. Strange. While running I smelt some glue getting warm around R29 and R31 which get hot, but I don't remember there being any hot melt glue on those. Also, I have -15 and +14.75 rails which should be just fine. However, I am going to monitor something in that area as the hot met glue seems to be heating up in that area, possibly. Will check that out some more but still no crossover notch to adjust out.

[52 Bill]

It may be that the range of adjustment doesn't allow the amp to get cold enough to show any crossover distortion. Some of the SWR schematics show different or selected parts "due to transistor gain" or something like that.

[DrGonz78]

It may be that the range of adjustment doesn't allow the amp to get cold enough to show any crossover distortion. Some of the SWR schematics show different or selected parts "due to transistor gain" or something like that.

That was where my thoughts were heading. So you would need a 1.5k pot (only a guess) or something to get to that point to see crossover point. At this point it only worries me since I have no idea how far past the crossover point it is set now. How hot is the bias as it stands and is there any other way to tell? I did change out the bias Q4 transistor for a 2sc4153 instead of the original 2s3834 that was installed. It does say on the schematic 2sc3834 or 2sc4153 and I could not find 2sc3834 at digikey when I ordered parts.

Edit: Just to note: Come to think of it I did some late night checks last night and did observe some crossover notches. It was cooler in the shop and now today it substantially warmer, 80 degrees opposed to 60 or so.

[g1]

How hot is the bias as it stands and is there any other way to tell?

What is the voltage across the emitter resistors for the output devices, and what is their resistance value?

[DrGonz78]

What is the voltage across the emitter resistors for the output devices, and what is their resistance value?

R20 = 0.228mv and R19 = 0.008mv. This is on my fluke meter (reads precision mv values) and on the extech meter it will only read 0.2mv and 0.0mv respectively. The DC offset is -125mv but goes to about -158mv once it settles in.

Edit: Forgot to add R20 and R19 are 0.1ohms, they read about 0.25 on the meter.

[g1]

Strange. Sounds like they are both turned off to me. I'd expect to see crossover notch. Is this with or without a load?

[Enzo]

So tack a short across Q4 E to C, that is the coldest possible, see any notch?

Also in teh circuit, does adjusting the bias control do anything to idle current or anything else?

[DrGonz78]

Strange. Sounds like they are both turned off to me. I'd expect to see crossover notch. Is this with or without a load?

Yup it is strange. So far I have done all testing without a load. I think it is time to put the amp on a load, but was hoping to at least figure this out first.

So tack a short across Q4 E to C, that is the coldest possible, see any notch?

Also in teh circuit, does adjusting the bias control do anything to idle current or anything else?

I will try this out in just a bit Enzo good idea.

[DrGonz78]

Had to call Fender just to at least say I checked. They did have a bias procedure and it was in the same files for the california blonde. However, it is for sm400/sm900 type amps. I told him that I am pretty sure I saw this online already and he seemed to think it was still related to the california blonde amp, since the procedure was in the same files as the blonde's. The SWR2000 schematic has a power board that is like the Blonde's power amp on steroids. These amps talk about having a 2 ohm load attached with minimal input signal to set bias. The file he sent over was already one I had found over on talk bass and I am not so sure it really applies to this amp.

https://www.google.com/url?sa=t&rct=...398,bs.2,d.cGc

[DrGonz78]

So tack a short across Q4 E to C, that is the coldest possible, see any notch?

Also in teh circuit, does adjusting the bias control do anything to idle current or anything else?

So in a sense we are shorting the bases on Q5 and Q6 together. Those look like the easiest connection points to physically attach a short across. So whatever negative voltage is on Q5 base is now directly connected to Q6 without any of the bias circuit causing the voltage to become less negative (Am I understanding this part correct too?). Am I correct in thinking that the source of the negative bias voltage goes across R11 feeding into the bias circuit? Just trying to understand better what I think I already understand.