Re: Remote Power Supply Connection

"Ian Iveson" <IanIveson.home@[EMAIL PROTECTED]
> wrote in message
news:DCgXj.44704$JE4.15748@[EMAIL PROTECTED]
> Iain Churches wrote
>>
>>> Also, transferring the filter caps to the amp would mean that both
plug
>>> and socket would be live at the time of parting. I have seen it said
>>> that
>>> a diode in the amp solves this problem, but I wouldn't entrust my life
>>> to
>>> that solution.
>>
>> You have to use a method (something like a sensing loop
>> with a relay) to shut of the DC from the connectors if the
>> cable is disconnected at either end.  If a round connector
>> is chosen, it must be one one with a deep screw collar or
>> bayonet latch, so there is no access to the pins while the
>> connector is still making contact. That is why an octal plug
>> socket cannot be used, although Patrick seems to have
>> found an innovative way to make that combination safe.
>
> Not safe enough to be legal here.

The argument agianst the use of the octal plug and socket
is the same as the argument against the use of the old-style
Bulgin plug and socket: it is possible with the aid of a tool
to touch the pins when the plug is partly removed but still
connected.   Does not Patrick's solution prevent this? I
have seen a similar idea where a black plastic capacitor
clip has been fitted around the octal chassis socket.

> Innovation and safety don't mix well, but to his credit I don't think
> Patrick was trying to be inventive. He's done the best he can do under
the
> cir***stances, probably. I wasn't keen on the detail that the HT dies
away
> within 10 secs IIRC. How long do electrocutions take? When is the most
> likely time a connection might be touched?

Agreed.  One cannot assume common sense, and correct methods
and routines should be carefully do***ented in black and white.
>

>> A power diode, as an additional safety measure works
>> well.
>
> Additional?

Yes, in addition to the sensing loop, which switches a relay
to disconnect the DC from the chassis connector if either
end of the umbilical is removed.

> Works well? How well?

I have a power amps with 450VDC on a separate PSU.
I have on several occasions unscrewed the connector from
the amp end (which has diodes for the B+)
and put the probes of a voltmeter between the relative pins
and ground.

The psu has a touch proof chassis connector, though I
suppose one could poke a sewing needle into it.  But once
again the DC is disconnected from the connector if the
umbilical is removed. There is no delay.  The DC is
off before the connector collar is fully unscrewed.

A lurker on this thread wrote to tell me that he has seen a
German built KT88 amp with a separate PSU and a DC
umbilical.  Both units have the DKE and  CE mark.
He has promised to try to find a link to the manufacturer
for me.

Iain