Should I replace this cap? [solved]

[Paul Birkeland]

There's a resistor between 26 and 31.  You have high voltage on one side, and 0V on the other.

Either you have a short after the resistor, and you are drawing all the current through that resistor (this is 99% unlikely since the voltage at 26 is so high).  The other possibility is that this resistor is loose or has a funky solder joint that is preventing it from passing voltage to the rest of the circuit (and your plate choke for that matter).

-PB

[Neuronal]

great! I'll start by reheating those joints and will let you know how it goes!

[Neuronal]

Definitely onto something.

When I measure resistance between 6-11 I get the expected 680 ohms. Between 26-31 my ohmmeter varies, like I am charging a capacitor - it slowly ramps from the kiloOhm to megaOhm range. If I come back later the resistance is back in the kiloOhm range. I measured capacitances across all of the caps on both power supplies and they seemed comparable, and I couldn't identify any obvious weak joints or shorts. Any ideas for where to head next?

Also, a related question: this all came about after I replaced the shottky diodes and the 1 uF parafeed caps that run from A2-20 and B2-40 with Auricaps. The serial numbers on the Auricaps and their dimensions are consistent with them being 2uF/600V caps (what I ordered), but when I measure their capacitance in the context of the circuit (installed) they measure about 20 uF. All of the other caps in the power supply measure what they are rated, so I am not misreading the meter (at least this time). Does the parafeed caps reading 20 instead of 2 make sense?

As always, thanks in advance for your collective advice - I can feel myself getting closer to fixing this thing!

[Paul Birkeland]

Definitely onto something.

When I measure resistance between 6-11 I get the expected 680 ohms. Between 26-31 my ohmmeter varies, like I am charging a capacitor

Is the 680 Ohm resistor present between 26 and 31?  If so, can you measure it?

[Neuronal]

Yup, the 680 Ohm resistor is present (and seems perfectly fine from the outside), but when I measure across it I get this very strange behavior where is looks like a capacitor (where the resistance slowly ramps from kOhms to mOhms). I could even measure a capacitance on my meter across this junction (about 30 uF). I get this result whether I measure the top or the bottom tabs, which makes sense (680 Ohms on the working side, the changing resistances on the broken side). I reheated the joints at 26 and 31 and added a bit of solder, which did not change this property. Comparing the capacitors on both sides of the power supply give essentially the same capacitance readings, and when I use the ohm-meter to measure resistance of each capacitor, all of the caps charge and discharge as one might expect, and so from my naive perspective the caps seem fine. 

Is the right move to replace the resistor? The fact that the 26-31 junction looks like a capacitor makes me think I must have a short somewhere else, but I don't understand (although I wish I did!) the underlying circuit.

[Doc B.]

If the other 680 is reading fine that 680 is probably open. One thing that is important is to measure with the probe tips touching the actual leads of the resistor, not the terminals it is soldered to. That way the measurement can't be affected by a bad connection to the terminal. If that doesn't bring any new info to light, the most sure way to verify the resistor's value is to disconnect one end from the circuit so it can't be affected by caps charging or other components' resistance.

[Neuronal]

OK, I will remove one end of that 680, remeasure, and report back!

[Neuronal]

So I removed that 680 Ohm resistor between 26 and 31, and sure enough, it was open!

When I now measure resistance across terminals 26 and 31 (in the absence of the resistor) it charges like a cap, just like before when the broken resistor was in place, so I assume the capacitance is some sort of circuit property that reveals itself when the resistor is blown or removed. I will call the Queen today for a replacement. If there is anything you think I should check before replacing the resistor, please let me know. Will report back when I install the new 680. Thanks so much Paul and Doc - you guy are amazing!

[Doc B.]

Yes if the relatively low value 680 ohm resistor is functional it will effectively shunt any charge buildup across a large value cap that is in parallel in the circuit, and thus you won't see much charging behavior in the ohm meter. Once that resistor opens up, the cap will exhibit that typical "rising ohms" characteristic when the meter battery tries to charge it.

[Neuronal]

I just wanted to thank everyone again for their help with this problem - I replaced the resistor, and the diagnosis was correct! Voltages all dead on, and really great music from both channels. Thanks so much for the help - you guys really are the best and make this hobby a ton of fun.