Problem - how to discharge a capacitor ONLY after the voltage has dropped to around 6VDC.

Window comparator, or simply bias a transistor off at 6v+. Since it looks simple, I suspect something's not made clear.

NT

An SCR shunt that is triggered only by a negative dV/dt on the 15VDC, would do it. That'd take a transformer or inverting amplifier to generate a positive gate current pulse when the voltage drops. It'd retire when the voltage hits a half volt, for a few milliseconds. A modest series limit resistor would keep the currents out of wire-melting range.

If you have access to the input power then it is very easy. Could even be as simple as a relay powered from the AC mains with it's NC contacts and series resistor across the big filter caps.

If all you have access to is the big filter caps then either of the other suggestions you have so far could be explored. (Level sensitive switch or falling dv/dt triggered switch.)

piglet

Wouldn't it be easier to replace the power supply with one that meets your needs?

Or a relay that disconnects the bleeder resistor when the OUTPUT is above 4V. Resistor would only have to dissipate enough energy to discharge the caps, as it wouldn't be connected under normal circumstances.

Trying to sense input dv/dt is fraught with issues when the power line glitches.

2 transistors are much cheaper.

NT

Hey, we need complete detail about that supply, so we can be sure not to accidentally use it.

--
 Thanks, 
    - Win

Cool, I like to learn things. Please publish your circuit.

That would be any supply with an UVLO, a feature becoming more and more common for switcher controllers and even linear regulators.

I did:

Left tr turns on at the desired supply rail voltage, robbing right tr of bias.

I suppose you cold take it a step further and add a bias divider to the right tr so it stops conducting below another threshold, if there's really a reason to.

NT

What regulator is that? It's kinda weird.

A switcher has a negative input impedance, so maybe the problem is that your supply can't furnish enough current to get it over the hump; it gets trapped. When it's playing dead, what's the input current?

We use regs with UVLO, or we add an undervoltage enable circuit, like a MAX809 thing. A safe supply recovers from brownouts as well as from obvious off/on cycles.

Your proposed discharge circuit might just reduce the probability of hanging up.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Two caps, a resistor, a diode and a JFET. The caps and the diode make a charge pump that runs off of the switch pin and pumps the JFET gate below ground. When the regulator stops switching, the JFET gate drifts to 0V, the JFET turns on (cuz it's enhancement mode), and sucks current.

Pretty sure you could do the same thing with a MOSFET, but with the resistor going to the supply voltage.

Or find a different regulator.

--
www.wescottdesign.com

Right! who made it? part number?

George H.

8 to 15 VDC +--------------+ Vin -->>--+-----+---------------|Vin GND Vout|----+--->>-- Vout | | Q1 +--------------+ | Cin \| | | | |----+-----, | | GND L| | | GND | / C1 | | | Cin/100 | | R1 | R2 | 220R GND 10K | | | | '------------+----------------------'

This might be workable and is pretty minimal. Adjust the resistor and capacitor values to suit your circuit. It will require some amount of time powered up to charge C1 before it can reliably pull down Cin. To prevent this parallel R2 with a diode.

--

Rick C

You could use the standard configuration of a '431 in the base of a PNP swi tch to R to GND, with the '431 ADJ input on a voltage divider to your regul ated output. As long as the regulated output has dropped out, the PNP drain s the input capacitor. This should work down around 3V. The TL431 is good t o 36V, so no problem there.

Interesting, however what if the power was restored before the voltage was low enough to turn of the SCR? Might be smoke time...

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
John's  Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

Fussy, fussy.

It is for pinball games and other vintage devices that use the 78H05 and are looking for a replacement. The part I want to use is a PSU7 from:

The problem is the large filter cap on the primary side of the unregulated +12 source. This unregulated 12VDC is used elsewhere in the game.

Any more questions?

Thanks!

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
John's  Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

On a sunny day (Tue, 9 May 2017 10:18:36 -0700) it happened John Robertson wrote in :

How about these?

I use those, and no problems so far. You can set voltage and current limit.

Well, simplest, but inelegant, is a relay that drops out at ~5 V. The N.C. contacts would put a resistor across the capacitor. This will drain it to

0 when the relay drops out.

The SCR someone suggested could get triggered on a momentary voltage dip and then stay on, causing excessive heating. So, I don't like that one.

A depletion-mode FET could be rigged to turn off when the regulator is running. Those are oddball parts, though.

So, two transistors could be rigged to do the same. A small power FET with resistor pull-up to the capacitor, and another transistor that would ground the FET's gate when the regulator was running. The pull-up resistor could be high-value, so draw minute current when the regulator is on.

Jon

Link has expired...

John :-#)#