Russian K40Y-9 started leaking and caused bias problem

1rebmem

Halffast
Subscriber
I noticed yesterday a 7591 power tube plate was glowing red.
I quickly connected my meter to the cathode 10 ohm resistor and saw .650v so I adjusted it down to .360v with the IBAM pot. But it ever so slowly continued to creep up. Swapped tubes around, same location was creeping up still. Verified the 10 ohm cathode resistor is good. Took some voltage readings and pondered over the schematic.
I realized a leaking coupling cap from the prior stage could put positive DC into the negative bias of the power tube control grid and cause the bias to go more positive. I replaced the coupling cap and problem solved.
I tested the removed cap on my Sencor checker. It went from original value of .068uF to .077uF and it had 9uV of leakage. I check all these Russian caps when I got them new, all were in tolerance. Installed them about 3 weeks ago.

Anyone had any issues with the Russian PIO caps?

Jef

DSC01165.JPG
 
This interests me, I have these in stock but have yet to use them.....Since reading about them being used many years ago, this is the first I have heard of one going bad.
 
You originally measured them at .068 before you put them in?
Yes...they all measured within the 10% tolerance with zero leakage.

I posted about this for two reasons:
- To see if anyone has had any issues with these caps. Maybe this a case of a rare failure?
- And if anyone sees their bias drifting, consider the coupling cap.

I personally have never had a new modern cap fail. On the job or in my garage. I have always been skeptical about using vintage NOS caps. But since I read so much about the great sound of these Russian PIO's I thought why not use them. I personally can't hear differences between caps. BTW...I was very careful installing them, to not bend the leads close to the body.
 
This is second time I hear about K40 caps leaking. I have them in 2 different units and so far so good.
I don't get them on eBay but prefer to get them from Jim McShane. He recommends them as best for your money and says that it will be sad day when supply runs out.
 
... just to say it one time again -- and I say it all the time -- there was no meddling. There was nothing to meddle with. There was no meddling with the cap-aign!"

Never heard of a PIO leaking unless they got knocked around enough to break the glass seal ...
 
Never had issues with those or other Military PIO caps. I also use a Sencor to match value and check for leakage at the rated voltage.
Knowing that heat is the enemy of any cap, I try to orient caps away from the obvious sources of heat such as power resistors.
Could proximity to an excessive source of heat been the cause?
 
Interesting that the K40s I got look different. Still an axial cap and silver colored, but shorter and fatter and having shields over the leads near the body. Also, looking online, there appear to be 2-3 types. Edit, mine are the ones that came with the Jim McShane kit.
 
I bought mine off ebay from a Russian dude.
First time I've ever bought them, I can't speak to the quality. $1.60 each.
I will do a follow up post if I have any other issues with them.
 
Interesting that the K40s I got look different. Still an axial cap and silver colored, but shorter and fatter and having shields over the leads near the body. Also, looking online, there appear to be 2-3 types. Edit, mine are the ones that came with the Jim McShane kit.

Here's what I have from Jim, also shorter and fatter:
K40Y.png
 
Interesting that the K40s I got look different. Still an axial cap and silver colored, but shorter and fatter and having shields over the leads near the body. Also, looking online, there appear to be 2-3 types. Edit, mine are the ones that came with the Jim McShane kit.

K40 is a generic cap type - different ratings and mfgs available that will make a difference in the can size, etc. The shields are real nice as a common problem is shorting the metal bodies to the chassis in tight spots.

The ones in my Latino custom build look way different, but I had time to kill and had some fun with them. Standard heat shrink on the leads, and the big stuff used for buried electrical cable connections (think water wells) with DIY paper labels.

pio-caps.jpg
 
K40y-9 is a specification code. That number decodes in to what type that is and how capacitor is made.
Size and shape will depend on capacitance and voltage ratings
 
This is the first report of a Soviet K40 series failure I’m aware of, and I have installed a ton of them over the years. Still have boxes and boxes left of late 80s production. The y-9 type are a nice and smooth pio for peanuts. Perhaps we are finally reaching the bottom of the barrel on them after all these years.

The coated ones are coated for moisture resistance in high humidity environments, not for electrical insulation value. If any are going to be near a short point, just heat shrink a 1kv clear dielectric sleeve around the case.
 
Those are 630 volt rated, thus fatter. I have the 400 volt.

Makes perfect sense, along with the other explanations that these vary in voltage and values.

One thing I recall reading is that changing these to 0.1uf from the original value (.068uf?) is the recommended new capacitance value. Can't remember the history/origin/rationale of this though, but this is likely documented in one of the many great posts on this forum for rehabbing/improving these fine units. Perhaps 0.077uf is close enough?
 
I don't believe you can just change capacitor value. Grid resistor value will need to be changed to maintain time constant for the stage.
 
Back
Top Bottom