[aeberbach]

If you follow the directions and use the servo then probably here will never be damaging DC going through your cans.

What if a component fails?

I was thinking of using an opamp (comparator) and relay per channel, one input being the AC coupled amp output and the other the DC coupled output. In normal operation the output of the comparator holds a relay closed. If the DC coupled output exceeds the AC coupled output then the relay would drop out and possibly light a warning LED.

 

[antomas]

Hi Aeberbach,

I'm trying to finish up my Gilmore and to place it at the center of my system; so having a watchdog to check that no component failed would be very nice. I suspect this would be appreciated by the builders of other DC-coupled circuits using expensive cans, too. Unfortunately I've not enough skills for helping with the design of the circuit. But I would be happy, in case, to see the results of your work.

If I've correctly undertood the logic of your circuit, and if it's true what I read in some posts at diyaudio.com (see the Doh posts in the Gilmore preamp thread), maybe the sound could be better if the relay would be normally inactive (no magnetic field from the relay); and the relay would last longer, too. Instead, keeping it normally active would help in case of a failure of the watchdog.

Ciao,
Massimo

 

[kevin gilmore]

While i have had absolutely no failures on any unit, i'm working
on a safety circuit.

Early krell amps had relays that shorted the input to the
amp while the amp was turning on or during fault conditions.
The relay was wired so as to unshort the input when the
coil was activated. Krell after a while found the continuously
energized relay to affect the sound, so they turned the
circuit around, so the relay not energized was the no-fault
condition.

 

[Possum]

Quote:

Krell after a while found the continuously energized relay to affect the sound, so they turned the circuit around, so the relay not energized was the no-fault condition.

Some people building preamps on diyaudio.com found the same negative effect and use this technique for their relay-based source selector circuits.

 

[antomas]

It's always a pleasure to read from the Master!

A naive question: why the Krell people didn't disconnect the output, where the signal is stronger and less prone to interference, instead of shorting the input? Another idea could be to disconnect the power supply of the amp with a latched relay and adding a reset button. I'm sure the Krell people considered these options, but... why they discarded them?

Thanks very much,
Massimo

 

[kevin gilmore]

quote

A naive question: why the Krell people didn't disconnect the output

Because series relay contacts (no matter how good
the relay) add some resistance and ruin the damping factor.