[Resolved] O2+ shutdown (power distribution??) correlated with use of GPS Expansion

William Scott · 15 Feb 2018, 18:36

Have been trying to reliably turn up four systems. Each lives in an Expansion Dock which holds a GPS Expansion. There is a Proto Expansion on which resides power conditioning for the Expansion Dock and a RS-232 - to - 3.3vdc UART conversion device (I/F to O2+ serial port #1).

Have isolated some issues to power distribution. As such I've changed the 5V rectifier to a beefy NTE960 which will accommodate 1 A minimum. This has significantly improved performance and made two issues vanish (won't bore you with those details).

Over night I've made the correlation between system shutdown and presence of the GPS Expansion. On multiple systems I've demonstrated: (a) without the GPS Expansion the system will boot and run indefinitely (multiple systems over 12 hours); and (b) in the presence of a GPS expansion each system would boot then operate successfully for perhaps one to several minutes before shutting down. The post boot operation duration seems to be random (probably isn't, but not enough data yet to say otherwise).

I'm looking for ideas on troubleshooting, unless of course someone knows of the silver bullet data point on GPS Expansion caused failures.

One thing I'll likely attempt is put a terminal on UART port 0 to see if the kernel announces some cause for shutdown.

What is actually happening in the shutdown? Here are relevant observations. Post shutdown there will still be 5vdc available on the two Expansion Dock ports expected. Also 3.3vdc is available on the Expansion Dock AND The O2+. Removing power, either external prime power or simply at the Expansion Dock power switch, clears the event and allows boot.

thanks!
Bill

ccs hello · 15 Feb 2018, 19:02

If it uses U-Box NEO-6 family device and is set for 3.0V operation, it consumes at least 111mA on that GPS component.
Suspect it's still the power issue. Omega2 is very, very picky/sensitive on that.

William Scott · 15 Feb 2018, 19:43

@ccs-hello Thanks for the tip! The U-Box NEO-6 is indeed the device in use.

Following is an excerpt from a power budget section of a test report for the system.

https://www.dropbox.com/s/dc6vo6567eytxjn/Screenshot 2018-02-15 11.33.14.png?dl=0

The 3.0V operation with a 111mA (minimum) load would put us in the upwards of 400mW dissipation range. As you can see from the measurements the unit is actually responsible for dissipating over 700mW. Ouch!!!

I'll rig up a couple of test cases (or one more complex). Will take a look at whether or not the O2+ reset line gets asserted as a function of the current load.

thanks!
Bill

Douglas Kryder · 15 Feb 2018, 21:14

@William-Scott hi Bill, i have nothing on the issue you face but wanted to know what software you are using to track/report on the system?

William Scott · 15 Feb 2018, 21:43

@Douglas-Kryder Not sure I understand the question. Could you clarify?

--Bill

Douglas Kryder · 15 Feb 2018, 22:12

@William-Scott sure, what software was used to generate the report in the image/screenshot you posted to dropbox?

William Scott · 15 Feb 2018, 22:15

@Douglas-Kryder Got it! That's just the Google Docs document application.

--Bill

William Scott · 15 Feb 2018, 22:56

This problem has been isolated to a power supply issue. Well....that's good and bad.

I've been experimenting with two units side by side. One unit is powered by a nice switcher at 12 vdc. I built this one myself years ago with a multi-output core purchased from DigiKey. I can't seem to get this unit to fail.

The second unit is powered by an old linear, a cheapo unit from Radio Shack. I can't get this unit to stay alive, sometimes even to post boot.

Therefore I'll mark this topic as resolved.

I may still have the problem. The two units referenced above are on the bench. I have two more units in my back yard. The power source for them is a Solsum 6.6F solar charge controller. One of the two backyard units runs for days. The other runs for minutes to hours, hours if I'm lucky. Now I wonder if the solar charge controller can provide a suitable load current for peak moments. Looks like I'll be in the game of rigging up an experiment with a controller tonight.

Now...can anyone refresh my memory on how to edit the damn subject line so I can "resolve" this topic?

thanks!
Bill

György Farkas · 16 Feb 2018, 05:14

@William-Scott said in O2+ shutdown (power distribution??) correlated with use of GPS Expansion:

Now...can anyone refresh my memory on how to edit the damn subject line so I can "resolve" this topic?

There were some forum motor 'improvements' during the last year:

simple posters can't upload files
no emoticons
no that sticked [Resolved] ... thread / folder
~~and it seems we can edit only the body of our post~~