Mesa 7i96 Output Wiring

PCW wrote:

Both are correct.

If you use the second scheme you should definitely fuse the 24V supply
since a short to ground (assuming 24V common is ground)
would likely damage the 7I96 output

Note that you should have flyback diodes across the relay coils

Am I correct thinking the diode positive (striped end) should go to 7i96 OUT+ (and can put the diode negative to OUT-)?

No, the diode should be across the relay coil

for example
Sourcing drive:
+24V --> OUT+ --> OUT- --> COIL+ --> COIL- --> GND

or
Sinking drive:
+24V --> COIL+ --> COIL- --> OUT+ --> OUT- --> GND

Diode always goes from COIL+ to COIL- (cathode=band to COIL +)
If you are unsure, test the relay+diode with a mechanical switch
replacing the 7I96 OUT+- pins

Your answer is what I was confused about. Is my drawing incorrect?
forum.linuxcnc.org/27-driver-boards/3845...iring?start=0#178037

PCW wrote:

No, the diode should be across the relay coil

for example
Sourcing drive:
+24V --> OUT+ --> OUT- --> COIL+ --> COIL- --> GND

or
Sinking drive:
+24V --> COIL+ --> COIL- --> OUT+ --> OUT- --> GND

Diode always goes from COIL+ to COIL- (cathode=band to COIL +)
If you are unsure, test the relay+diode with a mechanical switch
replacing the 7I96 OUT+- pins

As attached?

I'm a bit gun shy about hooking up my torch switch - I plan to use a multimeter to check voltage and current, but the polarity of the 7i96 output switches keeps breaking my brain. I read in another thread that an additional relay isn't necessary (at least for a hypertherm which uses a low switch voltage/current), but I'm tempted to use one nonetheless. I assume the alternative is to just loop through an output pair since the plasma is supplying the signal voltage for the switch (assuming it's a reasonable voltage/current and I send signal+ to OUT+ and signal- to OUT-).

The manual could be more friendly, i.e.

"Inductive loads must have a flyback diode. The output polarity must be observed (reversed outputs will be stuck-on)."

I can see what you mean now, but I interpreted that as flyback diode across the 7i96 outputs (which seemed to make sense since you could force current to only flow in one direction). I was also interpreting the statement about reversed outputs with respect to a diode across the outputs, which seemed to make sense since you would short the switch if you attached the diode in one direction. Big dumb.

It also doesn't say if the polarity sign convention is current flow, or just hook neg-neg, pos-pos. Hence the confusion. I empathise with AgentWD40, I need drawings to understand

AgentWD40 wrote:

Your answer is what I was confused about. Is my drawing incorrect?
forum.linuxcnc.org/27-driver-boards/3845...iring?start=0#178037

Yes, that looks correct if I understand those relay modules

Yes, those drawings are correct

OUT+ must be more positive than OUT-, otherwise the inherent diode
in the output MOSFET will conduct (so the switch will appear to be on if reversed)

This is not related to the flyback diode which is most often wired directly
across the relay coil.

5V/12V/24V Relay Module,4 Channel Optocoupler Relay Module Board High & Low Trigger (24V) www.amazon.ca/dp/B07QKLKPBJ/ref=cm_sw_r_cp_tai_rnFpFbZ9G2PSN

These relay boards are opto isolated on their inputs so I should not need a flyback diode correct?

PCW may correct me but Flyback diodes are only necessary where its an electromechanical relay (with a coil). Also I don't understand why you would bother with relay boards when the 7i96 has relays built into the outputs.