light pop out of left speaker [resolved]

[Urbansandz]

I'm occasionally finding that when I first start up my Kaiju the left speaker will start out ok and then shortly after startup I'll hear a light pop and lose much of the volume out of the left speaker. I've switched the speaker cables in an effort to narrow down the issue and I did find that the issue would then occur in the right speaker, so I believe the speakers are OK. When this happens I'll switch the amp on/off a few times and eventually the sound out the left speaker will be normal and once it gets into the normal state it will stay OK until I shut it off for the day. This didn't happen last year and I switched to another class D amp during the summer months. It's only begun now that I'm starting to use the amp during the colder weather. I don't think it's the pre-amp because I used that all summer with my class D without issue. I've tried switching both sets of tubes on the Kaiju to the opposite side with no change.

[Paul Birkeland]

The method to track this down is to set up two sets of stacked books that you can rest the inverted Kaiju on.  Hook up some cheap speakers to the binding posts and fire it up, then poke around with a wooden chopstick until you are able to cause the problem, then this can narrow your focus on where you may have a bad solder joint or two.

[Paul Joppa]

I'm a little worried that this could be a connection failure of the  resistor that bleeds the time delay capacitor, or the resistor that drains the plate voltage to ground until the delay kicks in. I'll check in again in a couple hours.

[Urbansandz]

I'll just add that it's not consistent. I fired it up this morning without issue. Also, this unit doesn't move at all (it's so heavy), so I don't think it's a situation where wires are moving around even slightly.

[Jay]

heat can move things around.  Wires expand and contract, the top plate can expand or even flex...  It only takes a tiny fraction of a millimeter to cause a problem.  I have solved a similar problem on my eros using the chopstick test.  (Actually, I used a bamboo skewer)

Here's a link to my post https://forum.bottlehead.com/index.php?topic=14878.0

[Doc B.]

I would suggest trying what Paul Birkeland suggests. PB has been the lead tech at Bottlehead for many years and I can tell you that it is what he would do if he got your amp on his bench. It definitely sounds like you have an intermittent connection, and his approach is the most direct way to locate the problem spot.

It is also less likely but possible that there is an issue with a tube. The easy way to check that is to swap the tubes between channels, one pair at a time. See if the problem moves along with a tube.

[Paul Joppa]

On the C4S board, the 33K resistor (mounted just inside of the Y+/Y- points) discharges the 10000uF time delay capacitor after the amp is turned off. It takes 20 minutes to fully discharge, and until it is discharged the time delay is not active.

If the C4S turns on before the tubes warm up, then there is a risk of the 300B grid going positive relative to the filament and can cause a small arc in some 300Bs. For this reason, the Kaiju should not be turned on before the capacitor is fully discharged, i.e. for 20 minutes after it was turned off. If the 33K resistor is not connected, the capacitor leakage current will eventually discharge itself but the time interval is much longer, possibly hours. This is covered in the third paragraph on page 6 of the manual.

To test, monitor the voltage across the 10000uF capacitor (use clipleads) with the amp off, it should have zero volts. When the amp is turned on, it should increase slowly to 3 volts, taking 20-30 seconds. When turned off it should decrease more slowly, back to zero volts.

[Urbansandz]

I measured the left side 10kuf capacitor and it initially only increased to 2.38 volts, I then tried the right side which increased to just over 3v and then I tried the left side again and it also got to just over 3 volts. Both seemed to discharge at the same rate and not so slowly that it would take an hour.

I just returned from vacation, turned the on the amplifier and things worked fine for maybe 5 or 10 minutes before the low pop was heard and most of the volume left the left channel. If I turn the amp off and then on shortly after the volume will be normal for only a matter of seconds before fading to a very low volume. It's only when the amp has been off for awhile that I'll get the pop.

[Paul Joppa]

What tubes are you using? Stock?

When you said it "only increased to 2.38 volts," what exactly did you mean - i.e. how long did you monitor it?

[Urbansandz]

Yes, yesterday it peaked at 2.38v. Today when I gave it a try, it peaked at 2.367v and that was after a couple minutes. After it peaks it slowly falls and I've never waited to see how far it falls. I tried it twice yesterday, and the first time it peaked at 2.38v and then the second time it got up to 3.014v. Maybe that's why the amp only sometimes works normally, though since returning from vacation on Saturday I have not been able to get it into the normal state where the volume on the left channel stays good for as long as I keep it on. The tubes are stock. I did have Electro Harmonix 300Bs in there, but placed the stock PSVANEs back in today to see if there would be any difference and I still get the issue. The other tubes are stock as well.

[Paul Joppa]

Thanks for the data. I don't think the issue is with the soft-start circuit after all.  That's a relief because there was potential tube damage.

I suspect one of the C4S connections, but best to follow PB's advice, the chopstick test described in post #2.

[Urbansandz]

I poked around today and was unable to locate anything loose. I then plugged it in and again got the pop a few minutes in. Normally I turn off the amp after I get the pop for fear that something could be damaged, but today I let it continue and after the pop the volume slowly returned to the left channel in full. I should note that I have the DC filament supply. Does this additional information provide any other clues? A light pop and immediate reduction in volume in the left channel followed by a slow return to full volume in the left channel.

[Doc B.]

Take a close look at the 300B tube on the left side and see if the filament dims or stops glowing after you hear the pop and the sound fades.

[Paul Birkeland]

I would suggest that you remove the 5670 from the socket and see if the pop happens under these conditions.  This will help in kind of splitting the amp in half to figure out where your focus should be.

[Urbansandz]

I realized that the readings from the trim potentiometers on the circuit boards somehow got out of whack and were reading too high ~190v vs 175v. I adjusted the trim pots to bring them back down to 175 and I haven't experienced the issue since. All readings are normal now and it's sounding great.