Power Switch for Electronics

January 4, 2007 8:02 AM

I have been assembling some simple test equipment that involves mating an Omron digital readout to an Omron 0-5VDC pressure transducer. I have been seeing erratic zero fluctuation that appears to coincide with switching the power on and off. My background is electrical, not electronic. I suspect that there could be a voltage spike occurring when the on/off switch is used. The switch is a good quality industrial double pole rocker switch with internal illumination. I want to eliminate possible switch caused spikes as a variable in trying to figure out if anything external is causing this wandering zero. I am familiar with switches that will control spiking in an electrical circuit but not with more sensitive electronics. What type of switch do I look for in my (Digikey) catalog?
posted by Raybun to technology (6 comments total)

Don't know the details of the transducer, but if it is piezo-electric, you could be seeing hysteresis. Perhaps a way to combat this is to heavily filter the power with a large choke (inductor) in series and bypass capacitors in parallel. If you think it is due to bouncing in the switch, try a knife-egde switch just to check.
posted by fatllama at 8:37 AM on January 4, 2007

I don't know anything about the Omron gear you are working with, but if you are using a mechanical switch to trigger a digital input, you are almost certainly experiencing "bounce" and need to debounce it.
posted by b1tr0t at 8:39 AM on January 4, 2007

Is using a capacitor not an option?
posted by veedubya at 8:40 AM on January 4, 2007

We do have older units that have an additional scope output. I will use one of them as a test unit to check for bounce. Thanks for jumpstarting me on this problem.
posted by Raybun at 9:23 AM on January 4, 2007

It's customary to not take any readings until the whole test circuit has settled down, so you should just ignore what you see til maybe 5 to 10 seconds after switching the power on.

I always allow time for equipment to settle down before taking readings. My oscilloscope takes a minute or two to fully warm up. Some oscillators can take a few minutes before the frequency can be precicely set.

There are ways to create a quiet, glitch-free power-up, but without knowing more specifics about your circuit, I can't make any further suggestions.

A more elaborate circuit is probably not warranted. Just don't take readings right away.
posted by Artful Codger at 10:11 AM on January 4, 2007

Seconding Mr. Arful Codger...

If you want to test, power the transducer separately. Then switch on/off the display unit using the switch you have.

If it's a bridge transducer (and why wouldn't it be?), when you apply power, excitation voltage is applied and it may be variations in this that you are seeing during the initial power up interval. Removing that from the equation by separately powering the the transducer should localize the problem to the transducer or the display, or at least demonstrate their dynamic.
posted by FauxScot at 10:54 AM on January 4, 2007