John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#10,342
I would like to turn our attention to output relays again. A colleague of mine has just made an extremely interesting measurement, on a power amp, that usually has almost no distortion. He has measured a huge 3rd harm. now, and spent whole day finding the problem. It was output relay. Look:
 

Attachments

  • relay.png
    relay.png
    89.9 KB · Views: 551
#10,343
PMA said:
I would like to turn our attention to output relays again. A colleague of mine has just made an extremely interesting measurement, on a power amp, that usually has almost no distortion. He has measured a huge 3rd harm. now, and spent whole day finding the problem. It was output relay. Look:
Click to expand...

It is a common problem - I've observed it many times. When an output relay contacts deteriorate, huge distortion may occur, mostly 3rd order. It is usually happening after the relay was subjected to a fault condition in the amplifier or if the relay driving circuit does not engage the relay properly, for example it the coil current rises slowly and the relay switches at the lowest possible voltage.

Alex
 
#10,344
PMA said:
I would like to turn our attention to output relays again. A colleague of mine has just made an extremely interesting measurement, on a power amp, that usually has almost no distortion. He has measured a huge 3rd harm. now, and spent whole day finding the problem. It was output relay. Look:
Click to expand...

A well known issue. In the forum contect, it was addressed at least by Bob Cordell here:

http://www.diyaudio.com/forums/showthread.php?postid=1469876#post1469876

and by yours truly here:

http://www.synaesthesia.ca/Auxiliary-circuitry.html

"Care should be taken when selecting the output relays. A high current model with silver contacts is definitely recommended. We experimented with a number of relays and definitely some cheap models are introducing an inacceptable level of distortions due to improper contacts materials and/or mechanics."
 
#10,356
monopoles

jneutron said:
Hey, symmetry demands those puppies..

I'll let ya know how the monpole detector works out. I think they're installing in in the next couple of weeks. (I'm not joking)..

We made a styrofoam box 8 feet tall, 3 feet by 3 feet, radiused the corners, and covered inside and out with mu metal..then the physicist came to me and said""can you degauss this thing??? And no, we have no money...
[snip]
Cheers, John
Click to expand...


Please John, keep me informed.
BTW, you can buy 'those puppies' here, only $1.25. Moreover, they sell more niceties like 'Zero-Resistance Wire -various gauges available'.

Cheers,
Edmond.
 
#10,357
john curl said:
Relays CAN sound bad, but if they are working properly, they can be a practical approach to several difficult problems.
Click to expand...

I would suggest the following approach that in my opinion should sovle many problems. Although it is known, it is still good and carefull enginering practice.

First use the RZ relay which is a good component.
Second, protect your contacts optimally in the following way ( see Ott's book for a complete treatment)

Avoid glow discharge by keeping surge voltage under 300V
Avoid arc discharge by keeping Voltage rise time under 1V per microseconde.

Glow is avoided by clamping on BOTH sides of the contact with clamping diodes to rail. You have inductance on both sides so be carefull, it costs nothing.

Arc is avoided by using a capacitor shunting the contact but NO serie resistor.
The goal is to avoid on opening a surge voltage rise time higher than 1V/microsec. This is equivalent to C ( in microfarads) higher than I load max ( in Amps).
A resistor in serie is often used to attenuate the high inrush current on closing. The capacitor has the full voltage then.
But then on opening there will be some voltage induced by this resistor and some initial arcing which is to be totally avoided in audio. The obvious way is NOT to use this resistor and switch the contact on before applying any signal and avoid offset.

Well known but not to often seen.

JPV
 
Status
Not open for further replies.
Top Bottom