[MCC]

I've unfortunately ran into more problems with the vintage McIntosh receiver I rebuilt some time ago. I noticed earlier today that when I turned up the volume (for a low volume source), the cones on the speakers were traveling significantly farther than they ever should. They moved in a seemingly random pattern and it looked to be caused by a varying DC offset. I immediately switched off the power and checked inside- my multimeter confirms that the can caps are fine, and everything else looks fine visually. My output tubes are also biased within .005v of each other.

I substituted in some cheap Chinese tubes for the telefunkens, powered it up, and took this video with no inputs attached:

Oscilliscope reading video

What could cause issue, and where should I look?

Thanks.

 

[MCC]

I worked on it again today- it looks like the issue is on both channels. I ran across this description of a similar problem on another site:

"The Williamson amplifier could exhibit a constant low-frequency oscillation resulting in a "breathing" action of woofer cones moving in and out, triggered by wow or rumble in the reproduction of long-play vinyl discs."

Breathing seems like a good word to use, as that's exactly what it looks like. Except my issue is not constant. I also noticed that the oscillation is not present until the volume is turned up beyond about 9 o'clock. If I turn on loudness at a volume that does not have the issue, it begins to oscillate again. All with no source connected.

Here's the service manual/schematic for anyone who wishes to help:

Manual

Thanks.

Edit: I'll post a video of the oscilloscope reading showing both channels after my classes.

Edit 2: Here it is: 2 channel oscilloscope

 

[SiBurning]

The schematic's kind of hard to read. Not to mention, I'm clueless about tubes, so this is probably just useless rambling, but at least read the last paragraph.

Do you mean C112, 113? Those look to be the ones in the feedback path. Looks like variable R10 and.. is that R149, R150? also in the feedback/bias path, as well as R101 (?) next to the caps.

Okay, so that's the bottom left of page 7, which biases the output tubes. Where that connects to, I can't see at all. Whatever's to the left of that is part of the bias circuit.

What I'm trying to get to is how to check and adjust the bias, since that quote you mention talks about how they got rid of this problem using 20dB of feedback, and that it'll oscillate at 35 or 0. I think you have to adjust the bias before you can tune the feedback right. Also, you checked the caps with a meter, but if their value drifted, you need to adjust the bias to compensate and maybe the caps drifted too far in which case you'd have to replace them. I'd also want to check any coupling caps leading in from the previous stage for leakage. I have no idea if or how tubes drift with age, but with a lot of old solid state circuits, the simple fix is to replace the transistors and, if necessary, replace the (input and emitter bypass) caps with something new, and, again, if necessary, maybe 10X larger.

Anyway, the best place to check for vintage gear is AudioKarma I've only visited the Vintage Solid State forum, but they also have a Tubes forum.

 

[MCC]

I've played around with the bias as suggested and found that the effect is significantly subdued at .7v (as specced). One of my EHX 7591a output tubes has gradually gotten weaker and weaker, to the point that I had to decrease the bias for all the other tubes to ~.67v.

The tube in question is only a little over 1/2 year old, but it seems to be par for the course considering the quality I've seen from these things. Two have already shorted and begun to glow red in the past 1-1/2 years the receiver has been under operation since I rebuilt it. Thankfully I was there on both occurrences to switch off the power in time. I don't leave it alone for long anymore.

Anyway, back to the point. I didn't realize how much of an effect lowering the bias would have- I thought I might get clipping at higher volumes, but never thought of oscillation esp. since it seemed to be fine until now. Unfortunately, I had to bump the final input voltage (after the inrush current limiters I added) to 120v from 117v to get the bias back to .7v. I found that internal voltages were most in-spec @ 117, so it looks like I'll need to buy another pair of tubes.

Sorry about the schematic SiBurning, and thanks for your help. I don't think I would have realized the issue otherwise.

 

[Nemo de Monet]

Quote:

Originally Posted by MCC /img/forum/go_quote.gif One of my EHX 7591a output tubes has gradually gotten weaker and weaker, to the point that I had to decrease the bias for all the other tubes to ~.67v. The tube in question is only a little over 1/2 year old, but it seems to be par for the course considering the quality I've seen from these things. Two have already shorted and begun to glow red in the past 1-1/2 years the receiver has been under operation since I rebuilt it. Thankfully I was there on both occurrences to switch off the power in time. I don't leave it alone for long anymore.

From what I've heard, the Electro-Harmonix tubes run really, ridiculously hot, more or less by design. Have you thought about adding a quiet little fan to provide some extra cooling? Might help with more than just tube life...

 

[MCC]

Actually, I have a 2-section vertical fan on the side of the receiver already. It's set up to blow air directly on the power section and separately over the output tubes. I would put a fan behind it, but my "A/V rack" (an $18 particleboard TV stand from Wal-mart) doesn't have any room in back.

The power section would overheat if I moved the fan to the other side and closer to the tubes, so unfortunately it's stuck there.

Edit: I'm having trouble keeping the bias on the bad tube- I'm buying a new pair, at $45.

 

[SiBurning]

Don't take line-level voltages too seriously on old schematics. The schematic for my Kenwood 7300 has 50V all over the place, which isn't even possibly given a 33v transformer. Realistically, the output is and always was more like 43-46V. Call it +/-10% and don't sweat the small stuff seems to be the philosophy on these manuals. The other higer voltages (e.g. -48.5, 14, 48.8) come off that, so I don't take them too seriously, either, although some are clearly a junction away from the line, which is easy to adjust for. Bias voltages (callouts on some schematics) tend to be more accurate. For example, on your amp around V14 there's 450 and -22. The 450 is a line voltage, so it's a very rough guideline. The -22 is probably also only approximate, but most likely doesn't vary more than 5%. You need to exercise some judgement. Since you're biasing this, you understand all this already. I'm just trying to point out that this approximate stuff is very common in schematics on older amps. Don't ask me if it's still common practise--I just don't know.

If you suspect the amp is out of tune and might be killing the tubes prematurely, you should check out the rest of the bias circuit and the feedback loop. Pay particular attention to the input caps to that circuit--just test for dc leakage, or scope it if you suspect frequency cutoff. Also, lift and measure the impedance of the can caps--C112, 113. I'd lift and measure all those resistors as well.

Heck, do all of that regardless.

Was kind of hoping some tube dude would step in but meanwhile if I can help ::

Rereading your last post... It looks like the power section needs an overhaul before you buy yet another set of tubes after this one. I hope this thing sounds worth the trouble. I'm close to giving up on my Kenwood, since it's really banged up physically, so I might just repackage the phono preamp and toss the rest. It has high DC offset at the speakers, and my new amp blows it away, so it's not worth any more work. Besides, there's plenty of people out there who'd love to have a pair of those unobtainium output chips.

 

[SiBurning]

Have you looked into tube rolling? There has to be some other more reliable tubes you can try.

 

[MCC]

Thanks again SiBurning. I guess I probably have been taking the voltages too seriously- they're all close enough to be a non-issue most likely.

I overhauled the power section when I found this receiver a couple summers ago- it's my modifications that actually cause the overheating unfortunately. There isn't much empty space in the power section. I replaced the tiny original silicon diodes with heatsinked TO-220 ones, put in larger wirewound resistors in place of the original ones, etc. There are air holes for convection cooling but not enough to dissipate all the extra heat.

As far as the tubes go, I'm probably stuck. The only other current manufacturer (JJ/Tesla) apparently makes even lower quality tubes, and used (not NOS!) quads are at least $150. They're rare- the story is that someone came to America some time ago and bought every single 7591 he could find before returning to Asia.

Other tube types can apparently be substituted with some rework but I don't want to risk it. That seems a bit more risky than the other part replacements I've done and this receiver is indeed worth keeping. It totally blows away anything else I've had a chance to hear, including an early 90's Pioneer Elite Reference series.