Mesa Board with high tolerance for high frequency with 24v Step-gen?

PCW wrote:

Here's the thing: For driver chips to be damaged with a plasma starting transient
(in the order of a few usec at most) there must be in the range of a 20V transient
between the 7I96 ground and the  servo drive ground in the same metal cabinet
This_IS_ a grounding or shielding issue. 
 
 

So I suppose what you are saying is there's now way the transients from an arc start/stop are killing the chips? And we are absolutely certain that it's just "grounding and sheilding"

In that case it makes it extremely easy to just say "here's how you fix this" and be done with it, you will never kill another one after this. So lets just fill in the blanks...
1) buy some shielded Belden cable off Amazon and tie the drain wire to <you insert grounding point>
2) verify mains ground sufficient by doing <you insert here>
3) (suggested before) ditch the RS422 step connections and go single ended and obviously couple Servo drive and 7i96 IO supplies as otherwise it wasn't necessary.
4) anything else......what else? no caps resistors or anything like that

I'm not an expert at.....anything really. But I would be inclined to let my imagination run wild and say something like maybe the plasma cutter and servo cabinet are on the same AC wiring and maybe the plasma cutter is causing some voltage drop that as soon as the arc extinguishes maybe the DC bus caps in the servo drives recover too quickly and surge the Servo drives IO lines......That might be a stupid suggestion but this is how my imagination works.

 

tommylight wrote:

Blazini i did read that, but in no way does it compare it to striking a plasma arc in open air from a high frequency start, it just states that even when done correctly, it is adding to the existing problem, not removing it.
 

Uh, that's a wrong assumption on all counts....

I don't mean that to be a jerk but that's a far reach

Ignoring HF start, the main issue with plasma and EMI is at torch start.
This causes a basically instantaneous current step in the table/work
(and everything connected to it) This can cause voltage transients in the ground
because all ground paths have relatively high impedances at the up to GHz
frequencies involved (arcs starts can generate picosecond risetimes).
One main solution is to keep all related control electronics
in a metal case _and_ to make sure all included electronic grounds are bonded
to the case with very short connections and that all external wiring is
shielded/filtered in such a way that it does not conduct EMI into the box

Is the 7I96S Frame ground connected to the case?

Are the servo drives frame grounds bonded to the case?

Is the 7I96Ss 5Vsupply grounded (at least for AC) to the case?

 Is the 7I96S Frame ground connected to the case?

Are the servo drives frame grounds bonded to the case?

Is the 7I96Ss 5Vsupply grounded (at least for AC) to the case?

Well the first thing I wanted to get some kind of way to confirm is how do we know that ground is actually ground? Let's speculate there's some wires going into a double gang box with a 3 prong dryer plug  next to the machine, and everything plugs into it. "Grounding" is useless ground is really ground. Short of telling the guy to drive a copper ground rod into the cement next to the machne.....I mean there's a wire going to the bottom prong of that outlet, but maybe we get our multi-meter out and do <insert here>


After that, he has the servo drives in a separate steel box from his Mesa card IIRC so......

Machine frame
1) make sure machine frame is grounded through AC entrance cable
2) Make sure plasma cutter power supply unit is properly grounded through AC entrance cable

Enclosure A (with the 7i96)
1) tie Earth ground from entrance cable to enclosure ground
2) tie 7i96 "GND PNT" screw (the one above the ethernet jack?) to chassis ground
3) tie AC input to 7i96 IO (5v?) supply to chassis GND
4) tie DC common output of 5v IO supply to enclosure chassis gnd (yes or no?)

Enclosure B (servo drives)
1) tie Earth ground from entrance cable to enclosure ground
2) Make sure all Servo drive frame ground points are tied directly to enclosure chassis ground
3) Make sure servo motor cable ground is tied to servo drive's frame ground (however that is done on that drive, probably the same point as the chassis, maybe it's done in the connector)
4) servo drive IO DC supply is expected to be handled inside the IO supply

Common Between Enclosures
1) Use differential step outputs (yes, no?) with SHIELDED cable and tie drain wire to (what?)
2) what do we do about tying 7i96 5v supply to servo drive IO supply or do we not bother because of differential stepgens?

So when we know exactly what to do with these things and when we know these things are done we can fairly confidently say you no longer have a grounding or shielding issue.