EMI causing Servo "twitch"

[pikapfw]

I'm having an EMI problem with my ioBridge / servo controller board (purchased from ioBridge).  Essentially, any time a reasonably high-current device (which is connected to the same circuit in my home) switches on or off, my servo motors "jerk".  The "jerking" of the servos is undesired (and the servos do not return to their default state), and I'm concerned about the safety of the ioBridge and any other connected equipment.

I've traced this down to "switching noise" / EMI.  The biggest culprit was a fluorescent light (workbench light) turned on or off by a standard power strip.  I confirmed this by performing the following:

1. I disconnected the power cord from the ioBridge (the "barrel jack")
2. I inserted a small machine screw into the barrel jack so that I could connect alligator wires to the positive and negative terminals of the ioBridge power supply output
3. I monitored the ioBridge power supply output with my oscilloscope (set the coupling to remove the 5V DC component and set a trigger at ~400mV)
4. Flipped the fluorescent light on or off
5. Oscilloscope triggered; extremely large spikes (+/- 20+ volts) observed!

Obviously this is not a good thing for a device expecting +5V DC.  Therefore, I've left my ioBridge unplugged for months now while I try to resolve this issue.  It is important to note that the wall wart for my ioBridge is the 4A model.  I carefully opened the case with a dremel and see that surprisingly, it's a SMPS (as opposed to the old-school jumbo transformer / rectifier style). 

Recently, I started experimenting with ways to reduce the noise generated by other devices on the same circuit.  I've added snubber (X2) caps across Hot and Neutral near noisy devices such as the fluorescent light switch I mentioned above.  That has significantly improved things, but the servos still seem to "jerk" at the slightest bit of noise on my AC power line. 

I'm looking for suggestions here on how to reduce the "twitching" of the servos, and reduce the (perceived) danger to damaging my ioBridge.  Has anyone else observed anything similar?  I searched the forum for EMI but didnt find anything really applicable. 

Thanks,
Pete

[nick]

I'm thinking zener diodes to clamp that surge and some bigger capacitors.

[jason]

I've witnessed the same servo twitch on my servos when I flip on a fluorescent desk lamp.  I don't think you need to worry about damaging the IO-204.  The module that runs my fishcam has been subjected to the abuse for 2 years without any damage.

The fluorescent light has a "starter" in it that produces a massive inductive kick.  The inductive kick not only get distributed by the AC wires in your home, but as an RF burst that gets picked up the signal wire of the servo cable.  

So how do you solve the servo twitch?  You can try using a digital servo instead of a standard servo.  I've tested a HiTec HS-5485HB digital servo with and without the servo smart board and it doesn't seem to be affected by the EMI problem.

A digital servo can be used just like a regular servo by plugging it into the servo smart board.  A digital servo can also be controlled directly through the I/O Channel.  Here is a post explaining how to do that: http://www.iobridge.net/forum/index.php/topic,328.0.html