I've heard of a few modders who are connecting the pot ground to either pin 3 or pin 10 ground pins instead of the -Trim pin 9.
Was curious about this, as well as why 3 ground pins, so emailed the GE rep. This is the email...
Hi Cynthia
I received your Inquiry:
Comments: I'm working with Naos Raptor 20A - NSR020A0X43Z - and have a question regarding the 3 ground pins 3, 10, and 9. A continuity check shows that they are all 3 connected, but an ohms check shows that pin 9 (-Trim) has a 1.2 ohm difference between either pin 3 or pin 10, and pin 3 and pin 10 have 0.2 ohm between them. Question is why is this and why 3 ground pins and must Rtrim be connected across -Trim pin and +Trim pin, or can Rtrim be connected across pin 10 or pin 3 and +Trim. I'm wiring according to the datasheet and connecting Rtrim across -Trim and +Trim, but am curious as to the need for 3 ground pins. Thanks, Cynthia
Pin 3 and Pin 10 are true ground pins. Pin 9 Trim – does tie to Grd but thru another component. Tying Rtrim resistor between Rtrim + and Rtrim – is the best way to eliminate any noise issue that maybe picked up in the application from interfering with this reference point. Also there is always a need for Rtrim to be connected internally in case the external connections are broken they module would not go unstable.
I hope this answers your question
Regard’s
Michael Naglich
What I think the rep means from his last sentence is that if pot ground (or Rtrim if not using a pot) is not connected to -Trim pin 9 and is instead connected to either pin 3 or pin 10 and the pot fails (or the resistor fails if not using a pot), the converter will become unstable. The converter should default to its low output voltage (0.6v) if the pot fails (as long as the unused outer leg of the pot is tied to ground along with the wiper), so not sure if this is what the rep means by unstable that it won't default to low output voltage, or if it is something else that would make the converter unstable.
Now, I'll preface this with the fact that I am FAR from an electrical engineer. I've done a lot with audio electronics and computers, but I'm like you; I've learned most of what I have through self study and experimentation.
From what I can tell from the datasheet, since it's still essentially a sine wave output (Fig 31, 39, and 41), the module still acts as an amplifier in a way. Instability would be akin to feedback - that ringing you hear when someone is using a microphone too close to a speaker and starts a feedback loop. The module is designed for power supply usage (probably computers from the 3.3V and 5.5V availability). On the input, we don't have to worry about inductance, since it's a battery, and is essentially just a solid resistance. However, since we use coils for the output, which are, by definition, inductors, this
may lead to instability in the module. Thankfully, the wave is extremely small in relation to the voltages and currents we're using
and the coils are resistance wire and don't allow enough of an internal field to develop to be of much concern.
I would err on the side of caution and use pin 9, since that's what the chip is designed to use for trim. It's a lot more relevant if a design is using the "Tunable Loop" feature. The feature is used for tightening up tolerances for sensitive electronics (again computers and such). Such as for large swings like when a CPU throttles up. Our outputs are fixed between pulls, so it's just a fancy extra feature that the module is capable of. Again though, just to be on the safe side, I'd still use pin 9, instead of 3 or 10.