What constitutes a leaky Coupling capacitor?

[Drewaudio]

Hello,

I've changed the coupling capacitors in a KT88 SE amp that I’ve been building. The KT88's are configured as triodes with 200k grid resistors.

I installed two new Sequa paper and aluminium foil coupling caps yesterday. The sound is glorious but the voltage on the KT88 grid for each channel is 0.2v and .007v. When I swapped the caps over the voltages were 0.7v and 0.1v.

I would really value some advice....is the grid voltage too high, shouldn't it be closer to 0v? Are the caps leaky?

Thank you
Andrew Parsons

[KaDe]

another reason may be grid-leakage (reverse current) from your KT88´s, swap them likewise

[DigitalJunkie]

What's the grid voltage at,if you lift one end of the coupling caps?
If it's lower,that could tell you that the caps are a tad bit leaky.

[Drewaudio]

Thanks for the ideas! Here's what I've tried:

1) I disconnected the coupling caps from the KT88 and measured a grid voltage of 14mv and 18mv on each channel respectively.

2) I replaced the coupling caps with 0.1uF orange drops and measured a grid voltage of about 19mv on each channel.

3) I then put the 0.047uF 600v Sequa coupling caps back in and measured a grid voltage of 128mv and 245mv on each channel respectively.

4) Next I changed the orientation of the Sequa caps in the circuit and measured a grid voltage of about 74mv and 291mv on each channel respectively.

For scenarios 3 and 4 above, I noticed that the initial leakage voltage was slightly higher within the first minute of operation before dropping by around 10% to the values recorded above.

The Sequa caps sound really good and the measured frequency response is great (-1dB 27Hz to 27KHz with only CFB). The question is whether 291mV of leakage is significant.

Kind Regards
Andrew Parsons

[Conrad Hoffman]

I've had some really obscure service problems on old test equipment due to leaky coupling caps, so I have little tolerance for it. Specifically, an increase in distortion in a waveform generator. I'd remove the caps and apply some significant portion of the rated voltage, then measure the leakage current with a meter. Any decent film cap will be nearly unmeasurable. They should be very low and at least match.

[Drewaudio]

Hi Conrad,

Further to my last post, I borrowed my Dad's 500v insulation tester. The orange drops measured better than 100M ohm. The better Sequa measured 100M ohms and the other 50M ohms.

Cheers
Andrew