DC power supply causes problems

[Aaron Maue]

I just finished a long round of troubleshooting on a prop I'm building for this year's show and would like some feedback on what I was doing wrong...

In short, I have a need to turn a 12VDC power supply on and off using a standard output channel on a CTB08 board. I chose a supply that I use elsewhere to power the lights in some lighted switches I use elsewhere in my display. Here's the one...

http://search.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go〈=en&site=us&keywords=285-1884-nd&x=0&y=0

The symptom I was seeing is that any time I turned on the channel that provides the 120VAC power to the power supply, all other channesl on the board would go full on, regardless what commands I was sending to the board. I finally discovered that when I used the output channel to turn on the 12V power supply, it would draw just over 150VAC.

Luckily, I didn't fry any hardware in the process. And I discovered that if I use a standard wall wart power supply instead of the other one, it works just fine.

What is it about the power supply at the link above that was causing my problem?

[WilliamS]

I could be wrong but the controllers are not designed for power supplys to be used on. Really anything like that at all that can produce a lot of electrical noise. They are used for lights and lights alone. You would need something with a isolator from the channels optical preferably to keep noise out of the system. I would bet this is the issue as soon as the power supply kicks on, as they tend to be dirty is causing a lot of "noise" in the controller which is firing up all your lights.

I would imagine someone will jump in here soon to straighen that out as well. I would think for what your doing as its turning on power supplys the DIO8RLY - 8-Channel Relay board would be your better use.

Also easier than that is just an outdoor timer to have them come on a few minutes before the show, and shut down a little afte.

[-klb-]

Switching power supplies usually are non linear loads, much like many CFL ballasts. They can really mess up the zero crossing detection for the controller, just by being on the same circuit. Just like CFL, some may be no issues at all, and others may be really bad.

[Aaron Maue]

Thanks, GoofyGuy. I wondered if that might be the problem. With the exception of this one output, I'm actually using the controller for light control. It's in a prop that needs just one 12VDC output, so I thought I'd take advantage of one of the leftover channels on the CTB to turn it on with the rest of the stuff. Oh well. I'll either stick with the wall wart solution (with more testing to see how much noise it's putting in the system) or I'll go with the outdoor timer idea.

If anyone else has further thoughts, I'd appreciate hearing them so I know what to avoid for next time. And thanks again, GoofyGuy.

[Aaron Maue]

Also, thanks to klb. Further confirmation that my idea itself was better than the results. A question, though... Should I expect similar results from a non-switching / linear supply?

[-klb-]

With a linear power supply, you will have inductance in the load. For a small supply, this is probably less of an issue than the switching supply. But do test.

Also, there are some old threads about a x10 filter that may prevent CFL lamps from causing issues with the controllers. Those filters may help either type of power supply as well.

[Aaron Maue]

Thanks for pointing out the CFL threads. For anyone else interested, I've posted one of the links below. They appear to be a little more spendy these days than they were when the thread was originally generated, but maybe worth a try.

http://forums.lightorama.com/view_topic.php?id=18850&forum_id=72

[Steven]

A simple (cheap) switching power supply consists of a full-wave rectifier and a large capacitor on the input side. It may be designed without any current-limiting circuitry, which will result in a high current when it is first plugged in as the capacitor charges from 0. Since your problem is happening only when the channel is first switched on, it could be caused by this initial high current, instead of electrical noise.

Try this: Immediately before the channel is turned 100% on, fade up from 0% to 100% over a period of 0.2 seconds. This may charge the capacitor more slowly on power-up.

Note: Fading the input to a linear (not switching) power supply is not a good idea, but it shouldn't hurt a switching power supply as they are designed to be used with a wide input voltage and frequency range.

I did read the specifications for the power supply you are using, which say: "Power Thermistor is built in to protect the circuit from high Inrush Current." However, this may not be working or sufficient, and it doesn't work until it has a chance to cool down.

[Gary Levelius]

Another suggestion might be to leave the power supply on all the time and just use the channel to switch a small 120v relay on and off to switch the 12 volt output of the power supply.

-Gary-