Power Supply Filter

[planters]

I use a variety of lab power supplies to run drivers and various fans TECs and controllers. I notice that if I do not turn off the driver supply prior to turning off the modulation supply I will get a brief flash out of my diode lasers. I placed a couple of my BK supplies across a scope and both turning on and off the main power switch produced a short (us) spike up to 5V despite the output being set to 0.0 during the operation. This is not too good.

Any suggestions for an easy effective output filter that will tame this spike?

[-bart-]

Are you 100% sure it comes from the output ?

This could also be some kind of brief common mode spike, traveling over a complete other path.

See also http://youtu.be/BFLZm4LbzQU

I'm not saying there is no problem.
Best thing is to eleminate the source rather than blocking it somewhere halfway.

[planters]

Pretty sure. The spike is only present when connected across the outputs of the power supply and it has a configuration a lot like a noisy ring down from a shorted capacitor. The spike reverses polarity several times and by 100us has settled down to 0.0 V. The pattern is variable depending on how vigorously I flip the on/off switch. Finally, what leads me to look at this is the effect it has on the drivers. They seem to be responding to a brief voltage and producing a current flow to the diodes.

[-bart-]

Have you tried snubbers across the poles of the switch ?

[planters]

I did not think of these and I will try one and see how it helps. I also fond this:

http://www.pololu.com/docs/0J16/all

[mixedgas]

If you can wait for me to complete my mission I can scan you the US Naval Postgraduate School's soft start circuit for " submersible" diodes. It will will cure your woes to 5 amps or so. Simple enough to wire yourself. The parts count is low, and the parts are readily available. It has the added bonus of serving as a back up voltage regulator.
Steve

[planters]

I can wait. I assume that the 0-15 V variable supply I need this for will not be defeated by the voltage regulation function. Or, if it is then it would still be useful for my fixed V supplies.

[steve-o]

if I do not turn off the driver supply prior to turning off the modulation supply I will get a brief flash

This sounds to me like a power down sequence problem. In the ol' days, we used to always have to power down all logic inputs to any circuitry before turning off the main device power or it would damage the input gates.
Ok, I just re-read your post and it seems the opposite of this. Can you build a delay so that the correct power-down sequence is acheived?

Any suggestions for an easy effective output filter that will tame this spike?

I like zener diodes that are on the verge of conduction across the LD. They will also forward bias any opposite polarity spikes that may occur.