STEPS grounding issue

[peterpan188]

Hey guys,

I finished my STEPS board for a little while now, and I have just case it up. I am using a standard Hammond alumimum case. As per tangent suggests in his article, I grounded the case, with a wire running from the ground of the IEC inlet ground to its bolt. I did sand off the alumimum anodization, then secure the wire in.

BUT now, I have been getting little shock from touching the case. when I put one hand on the top and the front (its a custom front plate, made with plexiglas), nothing happens, neither when I touch the back. It's always the time that I don't expect it happening that I get shock

. Well, nonethless, I would much rather get rid of this problem, and make it a functional and safe unit.

Any suggestions?

Thanks in advance,
Peter

 

[tangent]

Are you certain that it's not just static discharge? That case is now an excellent path to ground, much better than a doorknob.

If that's not it, then you may have an intermittent short of the DC voltage to the case. Up to 30V is certainly harmless, and some will tell you that the threshold is more like 40V. I've been zapped by 48V telephone ringing voltage many times and I'm still *erk* *ak* *pffttt* OOOOKAAAY.

So ennyway, as I was saying, I'd measure DC volts from one of the case screws to a known-solid earth ground. (Be careful: many older buildings have missing or faulty earth grounds. Use the cold water pipes under a sink if you're not certain your home has good grounding.) If you get a solid reading near 0V, it's grounding properly. (It won't be exactly 0.) If the reading fluctuates a lot, the connection isn't good, or it may actually be floating. If you get a substantial voltage -- especially one that looks like the unregulated or regulated DC voltages in your PSU -- you have a short somewhere inside the case. Probably the most likely ways to have a short are bad standoff choice or the heatsink touching the case.

 

[peterpan188]

Thanks, tangent.

I live in a dorm room in a college, so it's not very easy to find a real ground. Nonetheless, after what you mention about the standoff, it reminds me that I drilled a hole on the board to mount the it with alumimum screw and standoff. It seems to me that the hole that I made is part of the ground plate. Would that be the cause?

I will check the case voltage later with a DMM.

Thanks,
Peter

 

[tangent]

Quote:

Originally Posted by peterpan188 I live in a dorm room in a college

If the dorm is less than about 30 years old, it's a pretty good guess that the wall ground connector really does go to ground.

You can get an indication of how good your ground is by measuring voltage from neutral to ground. If it's more than a few volts difference, it's questionable.

Quote:

I drilled a hole on the board

I gotta ask: why? There were four holes there already, made for the purpose.

Quote:

the hole that I made is part of the ground plate. Would that be the cause?

It shouldn't be a problem, but the unregulated DC lines go through that area. You might have gotten into one of those.

EDIT: Come to think of it, you really shouldn't be tying DC-side ground to AC-side ground. You need to rethink your mounting scheme.