Overvoltage protection for negative voltage

"Daku"

My god you are thick.

... Phil

If you look at most crowbar circuits- they are basically 2-terminal circuits, maybe with a fuse in the supply rail.

So you'd design the crowbar for a -5V supply the same as for an equivalent +5V supply, except connect the + terminal to 0V, the - terminal to -5V through a fuse (if used).

Best regards, Spehro Pefhany

Turn that circuit upside down for (-) output.

Jamie

Voltage is relative. So your -ve and 0 could be thought of as 0 and +ve respectively(and your +ve would then be +ve - -ve). This means that your crowbar circuit is no different but, depending on how you look at it, maybe be "upside down".

Learn to think about how current flows in a circuit and through the components and it should be rather easy.

A simple example:

-ve = -5v, +ve = 7v

is equivalent to a circuit with

-ve = 0, ground = 5v, +ve = 12v

(of course we generally would relabel these so that 0V is ground and think of this as a dual positive supply)

I wouldn't even use the phrase "design the" -- you'd design one crowbar circuit for a given voltage, and hook one up to each supply in the right direction, then add fuses to each supply line before the crowbar. Job done.

I'm just picking nits about the _way_ Spehro said it though: the actual _action_ to take is the same regardless.

There should always be a fuse. You can't rely on a power supply that's gone overvoltage to have retained enough sanity to trip or fold back.

I'll nitpick right back at you, Tim. ;-) I said "design the" because it's quite common to have asymmetric supply capacities. You wouldn't need or want a crowbar capable of muffling a 30A +5V supply in order to keep a 0.1A -5V supply within reasonable bounds. So, I wanted to emphasize that the crowbar circuit and fusing should be designed to suit the supply (and the tolerance of the load to be protected, of course) and not just the nominal supply voltage.

BTW, perhaps I should have made explicit that I considered the fuse (if any) selection to be part of the crowbar design. In particular, the inevitable voltage drop across the fuse has to be considered.

If it's negative voltage, I suggest calling it undervoltage protection. Think about it...

All true. Very good points.

Don't think so.

"Undervoltage protection" has established meanings in various domains, and that isn't one of them.

Hi Daku, I just got a crowbar protection circuit working for my negative regula ted supply. I had looked at this thread, earlier, but didn't get a clear p icture from the comments, so far. So, I sat down with a lab supply and so me components and alligator clip leads and got a solution which I thought I 'd share for the sake of this thread. I started with a schematic that I found online (by Ravi Sumithraarachc hi) for a positive regulated supply. In my application, the negative regul ation is provided by a normally-positive regulator (7815 (TO-3 package)) th at has its (normal) output connected to Earth, and its ground pin is used a s the negative supply. Of course, there are negative regulators, but this implementation works, too. So, I put the fuse on the (normally positive) output on its way to Eart h and put the cathode of the SCR to the (normally Earthed) negative supply. The cathode of the 16 V Zener is still sensing the (electron-hole)input s ide of the fuse, but, in effect, I've fused the conventional return path, r ather than the conventional send path.

Please check out this drawing:

Maybe it will give you or someone else some ideas about how to impleme nt your own CPC...

servus, Andrew