Programming nRF51822 via 6pin?

RE: Programming nRF51822 via 6pin?

bolteon — Fri, 11 May 2018 07:07:44 GMT

Update yet again...

So, I can program a fresh unit with the slightly older softdevice and app from Adafruit and see the connection LED blink and connect to the module... (s110_nRF51822_7.1.0_softdevice.hex & blefriend_s110_xxaa_0_5_0_150224_blefriend_release.hex)

...but, when I try the later pair of softdevice and app- the module never starts (seemingly) the app. On an older board, I can program this newer combo and see it start up. (s110_nRF51_8.0.0_softdevice.hex & blespifriend_s110_xxac_0_7_7_161213_blespifriend.hex)

Very close!

Any thoughts? The files come from Adafruits repository found here:

https://github.com/adafruit/Adafruit_BluefruitLE_Firmware

//edit

Just to clarify, this is both confirmed with nRFgo Studio and JLink commandline.

//edit 2

Tested on a much older Rev2 board- functions just like the latest units... works on older softdevice/app but not on the newer- must be something I'm missing.

Strangely, the high fuse set in the Moteino MEGA board settings to not erase the EEPROM chip upon upload via the AVR ISP MkII only works on the middle board (does not work, much like the BLE situation- on the oldest and newest boards).

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 10 May 2018 18:54:10 GMT

P0.09 & P0.11 are just gpio. Nothing special about them. The user may have wanted to repurpose the unused jtag pins for his own use.

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 10 May 2018 18:06:58 GMT

Update time again...

So we figured out the jtag pin issue- turns out the schematic was designed to be used with a raspberry pi (!?!?!) for programming than spending the money for the JLink even though they included a JLink when they shipped us everything. Almost 2 weeks lost to a 9-size font note in one of the pages of the schematics... hilarity.

Anyway, last night got the new cables set up (hacked the singular one we had to reroute the cable) and got both the MCU and BLE programmed properly.

After adjusting for a BoM change (swap from 10MHz to 16MHz oscillator)- got proper serial output coming from the latest boards and successfully have them serial'd so that everything is kosher- except...

My proceedure is as follows for bring up:

Prime EEPROM via this command:

"C:\Program Files (x86)\Arduino\hardware\tools\avr/bin/avrdude" 
-C "C:\Program Files (x86)\Arduino\hardware\tools\avr/etc/avrdude.conf" 
-v -patmega1284p -cstk500v2 -Pusb -e -Ulock:w:0x3F:m -Uefuse:w:0xFD:m 
-Uhfuse:w:0xDE:m -Ulfuse:w:0xDE:m 
&& wait 1 && 
"C:\Program Files (x86)\Arduino\hardware\tools\avr/bin/avrdude" 
-C "C:\Program Files(x86)\Arduino\hardware\tools\avr/etc/avrdude.conf" 
-v -patmega1284p -cstk500v2 -Pusb 
-Uflash:w:"C:\DualOptiboot_V5.0_atmega1284p_BlinkD15.hex":i -Ulock:w:0x0F:m

Then in Adruino burn the bootloader
Upload the configuration firmware via AVRISP MkII
Set the serial number
Upload the main firmware via AVRISP MkII
Upload softdevice to nrf51822 via JLink
Upload main program to nrf51822 via JLink

When I do all that, I got the board to get to the point of searching for it's BLE pairing (so, that means getting a cellular signal, locking onto GPS, and checking the cloud for firmware updates).

Once there, I can't see the device in a Bluetooth scanner (both from Adafruit, and Nordic) and the Connection LED is not flashing at any point.

I have to verify 2 things tonight- when I tried connecting to the nrf51822 after the above setup- it couldn't and reported back a reference voltage of 3.664V (which is obviously high). When I cleared everything back down to priming the EEPROM- I *could* connect to the nrf51822 and the ref voltage reported was 3.4V.

Need to verify if this is a design error or if there's something else at play- as I have (from the previous cable hack-together) managed to wipe clean a nrf51822 and reprogram the softdevice and main program and it went back to the Connection LED flashing upon issuing Reset (r) and Go (g). Second thing to check is if an older board reports back the same high voltage when running properly.

Any thoughts and comments would be appreciated, as well as any other ideas about what P0.09 & P0.11 (as Turbo J mentions, they may just be debug lines).

RE: Programming nRF51822 via 6pin?

Turbo J — Thu, 10 May 2018 07:43:36 GMT

My guess would be: Debug UART. But it could be anything thanks to the pin flexibility of the NRF chips.

RE: Programming nRF51822 via 6pin?

natprop — Sun, 06 May 2018 19:51:39 GMT

Jay T.,

what are you using the P0.09 and P0.11 for? MDBT40 lists them a GPIOs

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 20:13:07 GMT

Our hardware guy is on that right now, had a great chat with Jay as well and we'll send over schematics if we don't trace the error ourselves.

Hopefully, it's just a matter of the wrong pins having been routed to the wrong location and we can make a 1-off cable that will work. Hopefully.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 20:10:51 GMT

Ok, then can you verify the schematic of your board whether the pins are correct ?

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 20:07:44 GMT

I'm not comparing it like that... I'm using the layout of the MCU jtag to verify that my probing was accurate, and that our pin 1 & 5 on the BLE jtag are flipped in polarity.

//edit

...and to clarify (not sure if I mentioned this before) but I don't believe the BLE module was ever programmed via the jtag before- it either came pre-flashed or was placed onto a different hardware board, programmed, and then soldered to ours.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 20:07:28 GMT

We know that. Actually pin 3 is RESET and 6 is SWO. They are not used in nRF51. The problem was that when OP connected the JLink it showed ~1.25V. We are trying to debug the hardware why that happens.

RE: Programming nRF51822 via 6pin?

Jay Tyzzer — Thu, 03 May 2018 20:00:46 GMT

Vcc_nRF is the voltage from the Target board sent to the programmer as a reference to enable level shifters, etc. This is taken from the nRF51822 Beacon Reference Design...pins 3 and 6 are not needed to program the device.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 19:36:35 GMT

I suggest not to compare like that. It's like asking why my orange is not red like my apple. I already gave you the info about the ARM connector. You should check that against your schematic. Do you have the schematic of your board ? You had a board that worked before. How did you flash it then ?

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 18:03:28 GMT

The reads aren't off... this is the pinout for the AVR ISP MkII... which matches what's on the board for the pins to the MCU.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 15:26:58 GMT

It looks like you might have the multimeter pins in reverse. The black lead should be on a GND pad and the red lead to probe. What is the actual design voltage you are supposed to have. A DC converter could varies +- 0.05V from one board to the other.

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 15:18:59 GMT

I annotated polarity on that image, black - for negative and red + for positive.

There is a 3.59V (to be exact since 3.6 is the upper limit) current between any labeled pins.

My concern is that the MCU jtag matches the mirror image of this pinout list: http://www.tag-connect.com/Materials/ARM20-CTX.pdf

The BLE jtag isn’t even close.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 15:10:48 GMT

Did look at it. There are no indication of what on that pin. It is better if you can post a schematic and the layout. Since you measured 3.6V it means the pin out is correct right ?

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 15:00:29 GMT

Look at my image, first pin below the top hole is not +...

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 14:55:31 GMT

Not sure to understand what you mean. The orientation of that connector is follow. The single hole is the top and the dual hole is the bottom. Pin 1 is to the left bellow the single hole. So is the pin out correct ?

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 14:52:53 GMT

Well, you are running at the max voltage supported by the nRF5x. Any variation of the voltage could cause it to malfunction. It is best to reduce it to bellow 3.5V.

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 14:49:03 GMT

Shouldn’t the orientation of the pins match this PDF though? Including polarity?

http://www.tag-connect.com/Materials/ARM20-CTX.pdf

The MCU pad does, (pin 5, ground) but the BLE doesn’t.

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 14:44:38 GMT

3.6V

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Thu, 03 May 2018 14:43:20 GMT

The AVR ISP and ARM SWD are not the same. You can't compare them or put them together. Each processor family has different jtag protocol and pin out. You can't mixe them.

So what is the voltage you get on the pin 1 of the BLE jtag ?

RE: Programming nRF51822 via 6pin?

bolteon — Thu, 03 May 2018 06:45:19 GMT

Update time...

New JLINK showed up, same issues- so it has to be the board. Went digging with the multimeter and I’m guessing this is the issue:

The MCU ISP is correct- and programs without issue. Pin 1 on the right hand column, 5 at the bottom and 6 next to it. 3.6V between these (1&5, 5&6).

Now, it seems like something got really screwed up when rotating and routing the BLE ISP... 1&5 are in the other column, and their polarity is flipped.

Green lines are continuity- not sure if those make a difference.

Guess that explains the lack of functionality... I’m assuming here that regardless of programmer- the 6pin header pad is supposed to be wired the same (both AVR ISP MkII and JLINK).

If so, is there any salvaging this connection setup? Does the polarity make a difference to the JLINK? What about alternative programming solutions (open to whatever is easiest)?

Help!

RE: Programming nRF51822 via 6pin?

bolteon — Tue, 01 May 2018 21:23:50 GMT

Noooooowwwwwwww I see it.

The annotated the 20 pin properly...

1 | 2
3 | 4
5 | 6
7 | 8

...and so on and so on, but didn't bother to do the same for the 6 pin. I'll figure out orientation with the multimeter.

Back tonight with more, new JLink tomorrow. Thank you again.

RE: Programming nRF51822 via 6pin?

Nguyen Hoan Hoang — Tue, 01 May 2018 21:19:32 GMT

http://www.tag-connect.com/Materials/ARM20-CTX.pdf

RE: Programming nRF51822 via 6pin?

bolteon — Tue, 01 May 2018 21:15:24 GMT

...and the 20 pin? 1-10 left to right top row, and 11-20 bottom row (also left to right)?

Top row being the one with the notch above it?