Intermittent device reset in modified ble_app_beacon example

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 15 Mar 2019 16:53:01 GMT

Sigurd,

OK, I think covers everything. This ticket can be closed as far as I am concerned. It has been a pleasure working with on. Thanks for all the time and energy.

-- Brian

RE: Intermittent device reset in modified ble_app_beacon example

Sigurd — Fri, 15 Mar 2019 16:20:38 GMT

Hi,

I'm happy to hear that the issue is resolved.

[quote user="bstromquist"]official documentation that describes how to connect an external board to the DK for programming when that board runs under its own power.[/quote]

There is general production programming white paper here. For the DK, the only doc for debug out/ programming can be found here. I see that we should add some more documention for this. The offical doc from Segger can be found here.

[quote user="bstromquist"]You said that the GNDDETECT line should be used instead of GND. I can't get that to work.[/quote]

I tested it here now, and connecting GND on the sensor to GND the DK is correct. Previous statement with the GND DETECT is not correct.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 15 Mar 2019 12:35:43 GMT

Update:

The sensors have been running for a week without issue so I think we can call the damaged battery theory is correct.

I am concerned that you have not provided a pointer to official documentation that describes how to connect an external board to the DK for programming when that board runs under its own power.

You said that the GNDDETECT line should be used instead of GND. I can't get that to work.

If we can resolve that then this ticket can be closed as far as I am concerned.

Would it be better to open a new ticket for the GNDDETECT issue?

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 12 Mar 2019 16:04:59 GMT

First, I tried GND DETECT but then it wouldn't program. When I changed to GND as above then it worked.

Is this documented somewhere?

RE: Intermittent device reset in modified ble_app_beacon example

Sigurd — Tue, 12 Mar 2019 13:58:07 GMT

Very good find with the reverse current.

[quote user="bstromquist"]Can you confirm?[/quote]

Yes. If you are already powering the device, e.g. with a battery, you should not connect VDD from the DK to the sensor-device. And in that case, I don't think you need/should connect GND from the DK to the sensor-device either, but connect GND from sensor-device to GND DETECT instead.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Mon, 11 Mar 2019 12:52:08 GMT

Sigurd,

All this focus on battery current got me thinking along these lines: Something is damaging the battery. There is no way so many cells from multiple brands of battery could be exhibiting this effect on their own. This lead me to measure the battery voltage while the sensor was connected to the programmer (the DK).

I found there was a negative current to the battery up to 1mA in some cases. This has to be the explanation. In further reading about coin cells I found they are very sensitive to reverse currents and, obviously are not meant to be recharged.

To test the theory I programed 12 sensors and quickly put in fresh PKCell batteries after disconnecting the programmer. Since te sensors has enough capacitance to run for several seconds without power, the processor kept running and the programmed time stayed correct.

So far, after two and a half days, there have been no resets and battery voltages are holding up better than before. So far so good.

Way back when I started this project (over a year ago) I searched the Nordic documentation for how to set up the DK to program an external board. I could not find any official instructions for this. but I did find this diagram in a forum post:

So I tried it and it worked. I tried searching the infocenter again just now. I can't find anything on how to do this or even any explanation of VTG, etc.

After reading a bunch of different posts, I think this is what I need inorder to program the sensor with the battery in place:

Can you confirm?

Please document this and make it easy to find.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 08 Mar 2019 14:00:32 GMT

For reference, here is what it looks like with the 680uF cap removed. I had to use a smaller shunt resistor (0.333 ohm) to capture this so the peak current is 14./2000/0.333 = 22mA. The average is 662uA.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 08 Mar 2019 13:48:16 GMT

Here is th epeak current with the Sky chip active during Tx: ~6mA

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 08 Mar 2019 05:19:10 GMT

Here is the current for the sensor as it has been tested all along - with the SKY chip disabled.

The peak current is 2.02 mA which means the 680uF cap is covering a majority of the current needed for Tx. The battery appears to operating well within the PKCell pulse current max of 10mA.

Next I will turn on the Sky chip and post that.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 08 Mar 2019 05:02:24 GMT

Here's the circuit:

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 08 Mar 2019 04:57:21 GMT

Here is what I have set up to monitor current. It was adapted from this project.

devzone.nordicsemi.com/.../image1-_2800_8_2900_.pdf

I verified it measures within +-3% between 10mA and 10uA by replacing the sensor with various resistive loads.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Wed, 06 Mar 2019 14:28:02 GMT

I am getting set up to more accurately measure the current. Do keep in mind during all these tests the SKY chip has been disabled.

I am finding out about the DCDC components too late i'm afraid.

RE: Intermittent device reset in modified ble_app_beacon example

Sigurd — Wed, 06 Mar 2019 13:46:43 GMT

Hi,

[quote user="bstromquist"]The current ranges between 180 and 220 uA when operating at 2.5v. [/quote]

Could you measure the peak current ? (You will see the peak during active radio tx/rx)

I see that at least the PKCell is rated at "Pulse current: 10mA (maximum)", I suspect that your peak current is much higher...

Looking at the SKY66110PA, it's rated at operating supply current at 11 mA, so with that component alone, you seem to be over the max current limit.

I don't see any DCDC components on your design, so the Radio TX(0dBm) will at be at ~ 11.6 mA. If you had DCDC, you could have reduced that by ~50%.

So while the average current is within the limits, the peak current might be too high for these batteries.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Wed, 06 Mar 2019 13:13:37 GMT

Panasonic CR2477 test failure

It took five days but now 3 out of 15 of the Panasonic CR2477 batteries have shown the voltage step up and subsequent reset. The battery voltage slowly ramped down to somewhere under 2.4v and now is at 2.6v.

Any ideas?

How can I keep the processor from reseting when this happen?

RE: Intermittent device reset in modified ble_app_beacon example

brian — Mon, 04 Mar 2019 15:51:26 GMT

Panasonic CR2477 test

This test has been running for over 3 days now without a reset. I have been measuring battery voltages twice a day and did notice one odd event: one battery jumped from 2.7v to 2.9v. Similar to what I saw with the PKCell batteries but not as large or as common.

So I'm still a bit concerned as these tests are all statistically small sample sizes.

What options are available ti mitigate the reset behavior of the NRF52832 when the supply voltage rises quickly by > 0.3v?

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 01 Mar 2019 19:04:14 GMT

So, we have eliminated all possible causes except the battery and the battery connection. So I've been doing battery swap tests the last few days as you suggested. The results are interesting.

First, I lightly sanded and then cleaned both the battery surfaces and the board's battery contacts and then cleaned all with alcohol. Testing resulted in 10 of 15 boards reset within 12 hours.

Then I rotated the batteries amongst the boards and retested for 12 hours.

This was repeated 3 times. Here are the results. X indicates failure. The red lines follow a battery through the tests.

The failures tend to follow the battery. Here are the same results by battery:

11 of 15 results tracked with the battery movement. I realize there not a lot of statistical significance in this data to claim confidence but I'm going to assume the battery is the issue and not the contacts.

During each those tests I put a scope on one of the boards' Vdd line. I set the scope up for a single-shot trigger with a rising-edge trigger and the trigger threshold set 0.1v above the measured Vdd.

On the last test run, the board I chose had a measured Vdd of 2.08v. So the trigger threshold was set to 2.18v. 8 hours later the trigger had fired and Vdd was now measuring 2.60!!! The battery voltage had INCREASED 0.42v at some point during the test AND this board had RESET.

This kind of jump would cause a reset as you stated earlier in this thread.

What would make a coin cell suddenly jump up in voltage 0.4v? The load is dynamic but the load pattern isn't changing.

I would have never expected this from a battery. Googling around found no similar experiences.

So let's talk coin cells. I'm using PKCELL CR2477. PKCell is the cheapest CR2477 out there so it is worthy of suspicion.

I am now testing Panasonic CR2477 and will also test Sony brand when they arrive next week.

Sigurd, Any experience, comments questions you can share on these latest results would be appreciated.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Fri, 01 Mar 2019 14:09:19 GMT

The current ranges between 180 and 220 uA when operating at 2.5v. At 680uF right now, C7 is able to keep the device running without a battery for 10 seconds.

Thanks for the link.

RE: Intermittent device reset in modified ble_app_beacon example

Sigurd — Thu, 28 Feb 2019 15:48:54 GMT

Hi,

[quote user="bstromquist"]has Nordic ever seen battery connection issues like this? [/quote]

Yes, we have seen similar issues with coin cell batteries, when the current consumption is too high, exceeding the rated drain current for the battery, and causing too large voltage drop that then causes a brown-out reset. We have a whitepaper on this subject here.

Could you measure the current consumption on your board?

Based on this result, you might need to increase the value of C7, and/or reduce the overall current consumption.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Wed, 27 Feb 2019 12:42:11 GMT

So it seems that we have ruled out everything except the battery and it's connections. Assuming the connection more likely to be the problem, has Nordic ever seen battery connection issues like this?

These boards use this metal clip and rely on a large circular solder pad for the connection to the negative side of the battery.

The contacts appear clean and the mechanical fit is snug but maybe...?.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Wed, 27 Feb 2019 11:39:44 GMT

The test ran without error for 22 hours.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 26 Feb 2019 14:22:49 GMT

Step increase: I have measured the step response and it is steeper that 1mV/mS so if for some strange reason the battery or battery connection creates a step of that size it could cause a reset.

I've set my scope on one board to single shot trigger just above the battery voltage. Maybe I'll get lucky and capture one...

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 26 Feb 2019 14:15:22 GMT

Started a battery swap test with 3 boards.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 26 Feb 2019 14:11:39 GMT

After twelve hours I reduced the PS voltage to 2.0.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 26 Feb 2019 12:48:28 GMT

No resets after 10 hours.

RE: Intermittent device reset in modified ble_app_beacon example

brian — Tue, 26 Feb 2019 03:17:43 GMT

I started a 10 sensor test from a 2.5v power supply. 1 hour - no resets.