Enforcing power supply ordering

[pigrew]

I'm needing to use a Marki A-2050 amplifier, which uses +/- 5 Vdc. It states that the negative rail must be power on before or simultaneously with the positive rail.

Is there a (simple) circuit I can build that only allows the positive rail to be applied while the negative rail is powered?

I expect to use a bench supply, providing both +/-, but I'm worried that I'll accidentally turn a knob in the wrong way.

Otherwise, I'll probably buy a fixed bipolar power supply, and assume it turns on both positive and negative simultaneously.

Thanks

[JS]

You need to build a dedicated PS or just something to use once in a while in the lab?

You could use a relay, normally open on the possitive rail and coil powered with the negative rail. Don't assume both goes on at the same time, also, there are dedicated power managment devices to coordinate sequential rails, but that's more for an fpga or a complex design that needs several rails in a very strict order and delay intervals.

JS

[tooki]

A relay whose coil is energized by the negative rail, and whose NO contacts are in series with the positive rail. Don’t think there’s anything simpler than that.

[pigrew]

I want this to be a semi-perminant fixture which labmates can use to power the amp, but it won't be locked to the probe station, or anything.

I should ask Marki what the power supply tolerances are, but I don't know if they have any sort of spec written down. I'm assuming the negative supply is used to bias the gate of a depletion-mode FET (or does it use a HBT?).

I like the idea of a relay, but they are a bit slow. I just ordered a power supply supervisor circuit, and a charge pump to generate a negative supply. I think I'll whip something up using a 7905 regulator or the like on the negative rail to filler the supply. I'll use the supervisor on the negative rail to control a FET enabling the positive supply (200 mA isn't that much).

[tooki]

I like the idea of a relay, but they are a bit slow. I just ordered a power supply supervisor circuit, and a charge pump to generate a negative supply. I think I'll whip something up using a 7905 regulator or the like on the negative rail to filler the supply. I'll use the supervisor on the negative rail to control a FET enabling the positive supply (200 mA isn't that much).

That’s awfully complicated and error-prone, it seems to me. A relay is essentially foolproof. And how is it too slow??? It’s no slower than a power switch.

[Brumby]

I like the idea of a relay, but they are a bit slow.

How fast a response do you need?

[David Hess]

Besides power supply sequencing, also consider fault conditions.  If the negative rail must be present before the positive rail, then a short on the negative rail after power is applied violates that condition.

The easy solution (or start of a solution) is to use the output of the negative power supply as the reference for the positive power supply so they track instead of using independent regulators.

[GigaJoe]

p-channel mosfet for off/on ;  and opto as a signal , if no common ground , kinda simple things.