choosing a bleed resistor

[john swift]

looking at PCB art work
http://electronics-diy.com/soft-star...wer-supply.php
it looks like its a project take from the Dutch: Elektuur magazine

John

PS

found it
the delay is from the July August 1997 issue of Elektor



the B250C1500 is a 1.5A 500V bridge rectifier
I would of thought a 1A bridge rectifier would be OK
or four 1N4007 diodes

[the great waldo]

Hi John.
Thanks very much, saved me some time this evening hunting for the article.
Cheers
Andrew

[Doddy]

Or use a NTC thermistor.

BTW, that design - the resistor wattage values are wrong - if, as claimed, it's intended to limit the current to 5A then the resistors should be rated at 250W (P=I^2 x R). I'd also caution that C1 is specified at the limits of its operation - failure of that would result in the explosive failure of C2 C2 and likely destruction of R1 and RE1.

Treat with care.

Oh, you dissipate about 6W in that capacitor and relay.

And A-Camera, electric shock isn't the only hazard of a charged capacitor. The biggest hazard at this potential are burns and physical injury resulting from a spontaneous reaction to a sudden discharge. You mention batteries - of course you don't use a bleed resistor on batteries - that's a divisive argument. But the first instruction in any vehicle maintenance manual is to remove the -V lead before any work. Why?

[Neale]

Why build a second power supply just to run the relay? How about driving a 24V relay from the existing power supply output via a suitable zener? With a zener chosen to drop enough volts to give nominal relay voltage when PSU is at working voltage, the relay will operate a little earlier than that, but by then the initial surge will be over. Even simpler, use a series resistor to give nominal relay voltage at PSU nominal output. I'm not convinced that surge protection is needed at all (my system works fine without but I chose my input MCB carefully so nothing trips, and that's the only reason why you need to bother) but if it is, make it as simple as possible. I've never used the NTC approach, partly because it looks as if you have to choose carefully to match PSU current draw, which is pretty variable driving a CNC machine.

[cropwell]

I would like to know what the design input voltage is to this soft tart circuit in Elektuur. We are assuming 50v as that is what we need, but it has 40v caps and a 24v relay.

[Doddy]

Nope, it's 220-240 on the primary of the toroidal, something that should be borne in mind for our US readers.

It uses a potential divider across the C1 capacitor and the relay coil. It works as-is, but is fraught with problems when people start substituting components.

[cropwell]

Sorry Doddy, I was trying to delete my post as I realised it was mains input. Just had a senior moment !

[the great waldo]

Nope, it's 220-240 on the primary of the toroidal, something that should be borne in mind for our US readers.

It uses a potential divider across the C1 capacitor and the relay coil. It works as-is, but is fraught with problems when people start substituting components.

Hi Doddy
I'm using a 500va 2x18volt toroidial transformer with 3 or 4 4700uf 100 v caps. 35 amp rectifier in my power supply, would the elektor circuit be ok with that power supply ? or should I hunt around for another soft start circuit ? Thanks in advance.
cheers

Andrew

[Doddy]

Why build a second power supply just to run the relay?

Feed from the secondary?, that's just too easy!

[cropwell]

Why build a second power supply just to run the relay? How about driving a 24V relay from the existing power supply output via a suitable zener? With a zener chosen to drop enough volts to give nominal relay voltage when PSU is at working voltage, the relay will operate a little earlier than that, but by then the initial surge will be over. Even simpler, use a series resistor to give nominal relay voltage at PSU nominal output.

Interesting idea ! Why not use a filament light bulb as the inrush protection resistance. If the bulb fails nothing will happen and if it is OK it will go out when you are at operating voltage. Just an idea that 'flashed' through my addled brain.