John Stumbles wrote:
On Thu, 07 Jan 2010 21:07:00 +0000, Ian Bell wrote:
I am using a smoothing technique first described by Scroggie

Not in Foundations of Wireless, IIRC. Second Thoughts, perhaps? I never
really got my brain round that one as well as I did the former (which was
a pressie to me when I was a spotty 'erbert from an uncle who worked in the
beeb).


I think I have in in a Wireless World article form sometime during or
soon after WWII.

That was pretty much the plan - the NC contact of the relay connects to
a bleed resistor equivalent to say 20% of the load, the NO contact is
the output and the common goes to the supply.

Well even with that arrangement any arc drawn as the contacts interrupt
the 'normal' load current would then have somewhere to go - into the bleed
resistance - so shouldn't be as bad for them as trying to break the
current to nothing. Why would the bleed resistor be sized to draw only 20%
of the load current? Why not all of it? (assuming the mains input to the
PSU can be guaranteed to be off at this point and so the bleed R isn't
being asked to dissipate the 70W load continuously).

John Stumbles wrote:
On Thu, 07 Jan 2010 21:07:00 +0000, Ian Bell wrote:
I am using a smoothing technique first described by Scroggie

Not in Foundations of Wireless, IIRC. Second Thoughts, perhaps? I never
really got my brain round that one as well as I did the former (which was
a pressie to me when I was a spotty 'erbert from an uncle who worked in the
beeb).


I think I have it in a Wireless World from during or shortly after WWII.


That was pretty much the plan - the NC contact of the relay connects to
a bleed resistor equivalent to say 20% of the load, the NO contact is
the output and the common goes to the supply.

Well even with that arrangement any arc drawn as the contacts interrupt
the 'normal' load current would then have somewhere to go - into the bleed
resistance - so shouldn't be as bad for them as trying to break the
current to nothing. Why would the bleed resistor be sized to draw only 20%
of the load current? Why not all of it? (assuming the mains input to the
PSU can be guaranteed to be off at this point and so the bleed R isn't
being asked to dissipate the 70W load continuously).



Whoops, sent the last reply without finishing.

Because , if I use an interlock to detect if someone removes the dc
cable then the bleed resistor could be asked to dissipate 70W continuously.


Cheers

Ian

"Ian Bell" wrote

Because , if I use an interlock to detect if someone removes the dc cable
then the bleed resistor could be asked to dissipate 70W continuously.


Though if the relay also disconnects the a.c. it won't.

I recommend adding a snubber - say 100nF + 100R. Actually you'll want two,
one from each of the NO and NC contacts.

David.

David Looser wrote:
"Ian Bell" wrote
Because , if I use an interlock to detect if someone removes the dc cable
then the bleed resistor could be asked to dissipate 70W continuously.


Though if the relay also disconnects the a.c. it won't.

Indeed, but that requires a separate supply for the relay - which may
not be a problem anyway. Nice idea, thanks.

Cheers

Ian

I recommend adding a snubber - say 100nF + 100R. Actually you'll want two,
one from each of the NO and NC contacts.

David.

On Fri, 08 Jan 2010 10:43:05 +0000, Ian Bell wrote:

Because , if I use an interlock to detect if someone removes the dc
cable then the bleed resistor could be asked to dissipate 70W continuously.

Ah, gotcha.

I know, since you're into thermionics how about using one of those
thermionic LEDs as bleed resistor? Cheap, readily available (give or take
recent EU rulings), nice glass enclosure, bright visible indication to
the user that the PSU isn't happy with what they've done ... ;-)


--
John Stumbles

Many hands make light work. Too many cooks spoil the broth.

John Stumbles wrote:
On Fri, 08 Jan 2010 10:43:05 +0000, Ian Bell wrote:

Because , if I use an interlock to detect if someone removes the dc
cable then the bleed resistor could be asked to dissipate 70W continuously.

Ah, gotcha.

I know, since you're into thermionics how about using one of those
thermionic LEDs as bleed resistor? Cheap, readily available (give or take
recent EU rulings), nice glass enclosure, bright visible indication to
the user that the PSU isn't happy with what they've done ... ;-)




LOL, that's a very good idea! I think I have some very nice candle
shaped ones somewhere.

Cheers

Ian

In article , Ian Bell
wrote:
John Stumbles wrote:
On Fri, 08 Jan 2010 10:43:05 +0000, Ian Bell wrote:

Because , if I use an interlock to detect if someone removes the dc
cable then the bleed resistor could be asked to dissipate 70W
continuously.

Ah, gotcha.

I know, since you're into thermionics how about using one of those
thermionic LEDs as bleed resistor? Cheap, readily available (give or
take recent EU rulings), nice glass enclosure, bright visible
indication to the user that the PSU isn't happy with what they've done
... ;-)


LOL, that's a very good idea! I think I have some very nice candle
shaped ones somewhere.

An additional advantage of using an old fashioned 'light bulb' is that its
resistance reduces as the applied voltage is reduced. So it may actually
have a lower value in normal use to when it is removing the charge as the
cap voltage falls.

Given what you've said previously about the long time required for the
charge to be removed using a bleed resistor that may actually help.

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

On 09 Jan, wrote:


So it may actually have a lower value in normal use to when it is
removing the charge as the cap voltage falls.

Oops. Should be "higher value".

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