[HighAltitude]

I've pretty much given up on trying to get a reliable and consistent calibration of the on board magnetometer on my Revo board and purchased a NAZA GPS/mag.  It needs to have the DJI connector removed and a 4 pin micro JST substituted.
Rather than fry my new GPS by incorrect connections, can anybody see the attached photo well enough to advise me which wire to connect to which?
Thanks

[hwh]

Take a look at https://forum.librepilot.org/index.php?topic=2592.msg18084#msg18084

[HighAltitude]

I'd been following the conversation with Mark36, so I was somewhat familiar with the topic.  Since the wire colors on my NAZA are completely different from the pictures on that thread and their connections with the sockets on the DJI connector are hidden and inaccessible, I used some needles stuck into the wires and an ohmmeter to identify which wire was connected to which socket on the connector. 
After checking and rechecking 5 or 6 times to minimize the risk of reversed connections, I soldered the DJI wires to a connector that fits the main port on my Revo board and went through the vehicle setup wizard to enable the NAZA GPS.
Unfortunately, it's not working.  Apparently, the Revo can't find the NAZA, as system health reports no GPS.
Is there any other way to determine what's wrong?
Incidentally, when I plug my spare uBLOXNEO-M8N GPS into the Revo, it shows up on system health.

[hwh]

If it's wired correctly all you should have to do is go to the Configuration --> Hardware tab and change the gps protocol from UBX to DJI.

[HighAltitude]

hwh - Thanks for the suggestions.
I've rechecked the continuity between the various wires in the DJI cable and the sockets in its connector another 5 or 6 times, and I'm virtually 100% confident that I have them mapped correctly.  I've also rechecked, desoldered and resoldered the connections with the 4 socket adaptor/connector that fits the Main port on my Revo Board in accordance with the diagrams attached.  Even with reconfiguring and rebooting, my Revo board still can't connect with the DJI GPS module.
I've ordered another DJI GPS that's supposed to be here in 2 days, so I'll fabricate an adaptor rather than cutting and soldering the wires to test it. 
Wish me luck.

[HighAltitude]

OOPS
Here's the other diagram.

[TheOtherCliff]

A few comments:
- An ohm meter puts voltage and current on the probe wires.  The lower the ohm scale used, the higher the current.  The cheaper the ohm meter (especially analog ones) the higher the current.  I have never damaged a device by checking it with an ohm meter, but it is possible.
- Did you try configuring and trying the other port on the FC?  Both main and flexi can do DJI protocol.
- I have damaged a flexi/main port twice in my life.  Look at the FC port connector with a magnifying glass.  You are looking for one of the tiny pins to be bent over.  It can be repaired with a #11 xacto blade to get it started and a pair of tiny tiny needle nose pliers to "clamp" the pin back to straight again.
- You might consider cutting open the DJI connector from the old GPS to make sure what colors of wire come from which DJI pin to make sure you wired it correctly. 
- On the new DJI GPS, leave a little wire attached to the DJI connector to see what color wire connects to which pin. 

[HighAltitude]

Thanks for your comments Cliff.
Addressing them:
I used the ohmmeter to map continuity between the various colored wires in the DJI cable and the sockets on the connector, so none of the current from the ohmmeter should have reached the GPS module, as it would have had to travel retrograde up the wires to the module.
I've not tried the Flexi port, since the pins are nicely aligned in the Main port, the mini connector plugs in without resistance, and the Revo can recognize the MN8 GPS connected to the Main port when the GCS is configured for it. 
Predictably, DJI makes its components pretty bulletproof, so dissecting the connector will probably take some determined Dremel work, but I'll do it.
My new DJI unit arrived in today's mail.  Before I cut into the cable, I've set up a bench test with a spare Revo and connectors on a piece of Coroplast.  Before I power up the Revo (with a 4 cell NiCad) and maybe fry another DJI GPS, please take a look at the attached photo for any connection errors.  Hopefully, the photo will come through with sufficient detail to see the wire color and connections.
Thanks again.   

[TheOtherCliff]

I used the ohmmeter to map continuity between the various colored wires in the DJI cable and the sockets on the connector, so none of the current from the ohmmeter should have reached the GPS module, as it would have had to travel retrograde up the wires to the module.

When you had the ohm meter attached to two different wires (one at the pin you stuck in the wire and the other at the connector) you had a small current flowing through the GPS.  I wouldn't guess that you hurt it, but it's possible.

I've not tried the Flexi port, since the pins are nicely aligned in the Main port, the mini connector plugs in without resistance, and the Revo can recognize the MN8 GPS connected to the Main port when the GCS is configured for it. 

If Revo MN8 works there, you know the port and hardware protocol (serial) works, so the main port is fine.

Predictably, DJI makes its components pretty bulletproof, so dissecting the connector will probably take some determined Dremel work, but I'll do it.

All I am saying is cut enough of connector shell plastic at the cable end off to get to some wire nubs.  You then can ohm out the connector to find what color wire connects to what connector pin.  You can then use the DJI pinout to label the wire colors with the correct function and use those same color to function mapping on the GPS with wire attached. 

My new DJI unit arrived in today's mail.  Before I cut into the cable, I've set up a bench test with a spare Revo and connectors on a piece of Coroplast.  Before I power up the Revo (with a 4 cell NiCad) and maybe fry another DJI GPS, please take a look at the attached photo for any connection errors.  Hopefully, the photo will come through with sufficient detail to see the wire color and connections.

Cut the connector off.  Just leave a 1/4 inch (or more) piece of cable on the connector, strip it back to individual wires and ohm out the connector wire colors like above (probably the same as above). 

[TheOtherCliff]

I have to make sure you know...

For serial: TxD on one end goes to RxD on the other and vice versa.  That can throw you off too if you read TxD at one end on a picture and think the end it is attached to is TxD.

(On the other hand if you wire an external I2C mag (not a DJI/Naza) the I2C wires are SCL and SDA and you connect SCL to SCL and SDA to SDA since this is a "bus" and not a "point to point link".)

[HighAltitude]

Thanks again for the advice.
Since my last post, I did dissect the connector and verify the physical connections of the wires to the sockets on the connector.  I even managed to do it without any bloodletting from the Xacto knife slipping.  My continuity testing was correct. 
I resoldered the mini connector to the cut off DJI cable wires on the first module the same way I did before, insofar as I can remember.
I must have been holding my mouth right, or muttered the proper incantations, because BOTH the old and new DJI modules are now talking to my bench testing Revo board through BOTH the Flexi and Main ports, even with the Rube Goldberg testing setup.
Since I now seem to have two functional NAZA modules, I can install them on two of my two quads and hope it warms up enough outside to calibrate the sensors and try some test flights.

[TheOtherCliff]