"Brian Gaff" wrote in message
m
I got the impression he wanted to short the ht out to
make it safe though. Another thing to ask is is the load
inductive? When I could see Ive welded many a rely like
this!
I guess for the crobar effect one could use a
semiconductor, and maybe one the other way around so to
speak to do the soft rise in voltage. Its a highish
current, but not impossible.

Legacy crowbar - fire a thyratron.

Phil Allison wrote:
"Ian Bell"

Safety. One of the functions of the relay is to disconnect the HT if the
PSU connector is unplugged. If I just switch the ac and bleed the HT there
will be a significant HT voltage on the PSU connector for long enough for
someone to get a shock from it. That's one reason why I prefer to switch
the dc HT.


** Bell must be the biggest idiot on the face of the planet !!!

What sort of UTTER MORON uses a connector with exposed and ** touchable
pins** for the output of a high voltage PSU ??

Has the pommy ASS never noticed how AC supply leads ALWAYS use
connectors with female pins on the end that goes to the equipment ????????

Unbelievable.


Of course I use female pins. Unfortunately, female pins are not by
themselves sufficient to stop a stupid enough person from connecting
themselves to them. Disconnection is the only really safe option.

Cheers

Ian

Jim Lesurf wrote:
In article , Ian Bell
wrote:
David Looser wrote:
"Ian Bell" wrote
Yes, but unless it consumes a significant fraction of the load then
its decay time will be rather long.

True, but it's normal. How many bits of valve kit have relay connected
discharge resistors? - almost none. Is there any particular reason
why you need a very short discharge time?


The PSU is remote and the relay has an interlock to turn off the HT if
the PSU HT output lead is disconnected. I need to either disconnect the
HT or bleed it very quickly to avoid a possible shock hazard.

For the HT source side I assume you use female connectors, so physically
preventing accidental contact with fingers, etc.

Of courcse.

For the HT sink side you can use a silicon diode following a shunt resistor
to ground. Once the plug is pulled this isolates the HT from the prongs of
the male plug.


I don't quite follow that. Could you explain a little further.

Note this means the cable is *not* reversable as it has a male at one end
and a female at the other to ensure that whichever end is opened you can
apply the above correctly. The shunt resistor before the diode will also
help remove the charge from any cable capacitance if the disconnection is
at the source end.


Yes, that is the plan - female connector on PSU, male connector to
female connector in the cable and male at the equipment end. Still not
clear how the diode and shunt R work.

Afraid I don't recall you saying where the smoothing/reservoir caps are
located. So I'm assuming both/either ends of the link.


Mostly at the PSU end but there will be some in the equipment too (
order of magnitude less capacitance).

Also use a bleed resistor on the caps.

You could use something like a crowbar which trips if a loop via two
conductors of the cable is broken. However though neat in some ways it can
fall into the trap of being over-complex so designing in extra failure
modes. Hence if you are very safety conscious I'd still do the above.


That's effectively what the relay is supposed to do:

1. A NO contact disconnects the HT when the relay drops out
2. A NC contact rapidly discharges the resevoir caps..
My experience is that almost any type of conventional relay will end up
causing problems if the unit is used for many years. Avoid if at all
possible. Solid state relays would be much better if you can find one that
suits your purpose.

I think I just grokked the series diode shunt R at the equipment end.

Thanks

Ian

Slainte,

Jim

"Ian Bell" wrote

Of course I use female pins. Unfortunately, female pins are not by
themselves sufficient to stop a stupid enough person from connecting
themselves to them. Disconnection is the only really safe option.

Wouldn't a better option being to simply house the thing somewhere where a
stupid enough person can't get anywhere near it?

David.

David Looser wrote:
"Ian Bell" wrote
Of course I use female pins. Unfortunately, female pins are not by
themselves sufficient to stop a stupid enough person from connecting
themselves to them. Disconnection is the only really safe option.

Wouldn't a better option being to simply house the thing somewhere where a
stupid enough person can't get anywhere near it?

David.



Unfortunately, it is surprising how enterprising a health and safety,
sorry I meant stupid, person can be.

Cheers

ian

"Jim Lesurf" wrote

My experience is that almost any type of conventional relay will end up
causing problems if the unit is used for many years. Avoid if at all
possible. Solid state relays would be much better if you can find one that
suits your purpose.


I don't share your jaundiced view of relays. Whilst they are not suitable
for every switching requirement, there are many applications for which they
are the most logical choice. I suspect your negative view is based on the
effects of cost-cutting by equipment manufacturers. SSRs are a lot more
expensive than ordinary relays and they have their own limitations. They
only make economic sense when very many operate-release cycles are expected
over the equipment life.

I might cite the two relays in the control box of my 30-year old central
heating boiler. The boiler is well past it's sell-by date and due for
replacement in 2010 (esp if I can get Gordon Brown to chip in £400 :-), but
the relays have never been any trouble.

David.

On Fri, 25 Dec 2009 09:45:32 +0000, Jim Lesurf wrote:

snip

My experience is that almost any type of conventional relay will end up
causing problems if the unit is used for many years. Avoid if at all
possible. Solid state relays would be much better if you can find one
that suits your purpose.


For goodness sake don't tell the thousands of small generating stations
that have been using relays for generator control continuously for the
last 40 years or so... They might get suspicious if they find out that
their technology is so unreliable. ;-)

Providing relays are specified correctly (and not run at too high a
temperature) they can be very reliable indeed - often better than solid
state switching as they aren't affected by moderate spikes on either the
coil or contact circuits. When that is coupled with the very simple plug-
n-play maintenance (which is often all the systems get) it makes them
ideal for use in remote generation plant.

The main problem is wiring. They are very labour intensive when a lot are
needed, not to mention the space needed for relay racks. I've yet to find
a 4-pole SSR with changeover contacts... :-)

--
Mick (Working in a M$-free zone!)
Web: http://www.nascom.info
Filtering everything posted from googlegroups to kill spam.

"Ian Bell"
Phil Allison wrote:
"Ian Bell"

Safety. One of the functions of the relay is to disconnect the HT if the
PSU connector is unplugged. If I just switch the ac and bleed the HT
there will be a significant HT voltage on the PSU connector for long
enough for someone to get a shock from it. That's one reason why I
prefer to switch the dc HT.


** Bell must be the biggest idiot on the face of the planet !!!

What sort of UTTER MORON uses a connector with exposed and **
touchable pins** for the output of a high voltage PSU ??

Has the pommy ASS never noticed how AC supply leads ALWAYS use
connectors with female pins on the end that goes to the equipment
????????

Unbelievable.


Of course I use female pins. Unfortunately, female pins are not by
themselves sufficient to stop a stupid enough person from connecting
themselves to them.


** Works for several * BILLION * mains power leads in use round the world.

YOU ****ING AUTISTIC MORON !!!


Disconnection is the only really safe option.


** Execution is the only safe option for congenital anencephalics like YOU.




..... Phil

In article , David Looser
wrote:
"Jim Lesurf" wrote

My experience is that almost any type of conventional relay will end
up causing problems if the unit is used for many years. Avoid if at
all possible. Solid state relays would be much better if you can find
one that suits your purpose.


I don't share your jaundiced view of relays. Whilst they are not
suitable for every switching requirement, there are many applications
for which they are the most logical choice. I suspect your negative
view is based on the effects of cost-cutting by equipment
manufacturers.

Based also on kit I designed myself and where I chose relays. The problems
may take decades, but in my experience eventually turn up. If you think
about it, physical contacts all eventually wear or corrode, springs weaken,
etc.

SSRs are a lot more expensive than ordinary relays and
they have their own limitations. They only make economic sense when
very many operate-release cycles are expected over the equipment life.

I might cite the two relays in the control box of my 30-year old central
heating boiler. The boiler is well past it's sell-by date and due for
replacement in 2010 (esp if I can get Gordon Brown to chip in £400 :-),
but the relays have never been any trouble.

However the problem Ian is converned with is a safety issue, and for all I
know might affect users in some decades time from now.

And FWIW I had to change the controls for our central heating some time
ago.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

In article , Ian Bell
wrote:
Jim Lesurf wrote:
In article , Ian Bell
wrote:


For the HT sink side you can use a silicon diode following a shunt
resistor to ground. Once the plug is pulled this isolates the HT from
the prongs of the male plug.


I don't quite follow that. Could you explain a little further.

Think of the connector on the unit at the receiving end of the cable.

First put a shunt resistor between the pins. The put a series diode in the
internal lead from the connector into the unit so that it is in between the
connector and any internal capacitor. Put this to way round so it will
conduct in normal use.

Now pull out the cable from the connector.

The power supply has been removed but there is still a charge in the
capacitance of the unit. However to reach the connector it would have to
pass back though the diode. But for this current diection the diode is
reverse biassed, and its other end is connected to zero by the shunt
resistor. So the diode isolates the internal capacitance from the
connector's 'live' pin as the resistor connects it to zero.

If the cable is still connected to the unit, but disconnected at the other
(PSU) end, then the shunt resistor also bleeds away the charge in the cable
capacitance.

This clearly means using a diode with a high enough reverse bias limit, and
accepting a diode drop in the supplied voltage in normal operation. But it
is a cheap and effective way to have a 'one way valve' in the system.


I think I just grokked the series diode shunt R at the equipment end.

OK, the above should confirm that. :-)

A loop sensing relay should work. And physical relays do have a low
insertion resistance, and good isolation, when working well. But I do have
a worry about reliability over time and with any physical abuse. Solid
state methods like diodes, crowbars, etc are better in my view.

Based partly on signal relays. But also with working in the past with 2k5V
PSUs for klystrons and higher voltages for BWOs, etc. ...and with having
shocks when other farted up the system without telling me. That was how I
found I *could* do a back flip. 8-]

No harm in having a loop relay as *well* of course. :-) But I'd then
wonder about things like leakage in the cable if using the system for
safety critial purposes in the long term where abuse and degradation might
occur.

BTW One 8kV 100mA PSU I used to use had a habit of tripping at a given time
of day. Turned out to be sunlight getting into a crack in the case and
shining on a photodiode in an isolator that sensed a connection problem.
:-)

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html