• State Verified Answer
  • Locked Locked
  • Replies 21 replies
  • Subscribers 62 subscribers
  • Views 18582 views
  • Users 0 members are here
Attachments (0)
Nordic Case Info
  • Case ID: 100561
Options
This post is older than 2 years and might not be relevant anymore
More Info: Consider searching for newer posts
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

How are you going about Production Programming for nRF51 parts?

Marc Nicholas

Howdy,

We're gearing up to program about 4,000 nRF51822s. They'll get the BLE version of DFU plus a stable version of our own code. For future production runs, we'll likely have our partner in Japan program the devices on the assembly line.

So, my question is: how are you handling your programming? Are you literally doing boards one-at-a-time with a SEGGER and a pogo-pins or similar?

Our PCBs are panelized as 5*5, so I've toyed with the idea of building an X-Y slide with stepper motors to fly a programming head across the boards. Buying 25 SEGGERs so that we can push 25 sets of pogo-pins down on to an entire panel is cost prohibitive.

Our world is also slightly more complex in that we create custom stickers that are unique to each programmed device, too :)

Thoughts? Current strategies? Horror stories? Please share! :)

-m

  • 0 in reply to John Kelley

    Our solution right now (still being built):

    • Some Sparkfun Koala protoboards to supply an FTDI for testing UART
    • Same board provides ~100mA 3.3vdc
    • Prototype area on board houses a TI CD74HC4067 multiplexer chip

    If I can get the SEGGER working with a Beaglebone Black, then I'll have that orchestrate the test firmware flash, testing of the device, success alert (Super Mario theme song), incrementing the multiplexer to the next device, and visual display of which devices have failed.

    If that doesn't work, the same protoboard will just become a single programming and testing station :)

    Will let you know how we get on (and post pics!).

    -m

  • 0

    a cute approach would be the thought of using a Techno-Isel (or like) CNC router frame with a nest fo the pvb's and "toolhead" thats composed of a Bed of Nails fixture. One could easily use the "spindle"/"toolchanger" controls to trigger a flash go..(click of the mouse button.). If your product has an LED, a light sequence on program,reset could give you a go/no signal.

    We have one that's used for light stuff and prototyping and certainy has the speed and accuracy of about.0004" repeatability, to hold a BoN. Our 10-year old machine has travel movement of 1200ips. Newer ones are probable faster. The movement is very reliable.

  • 0 in reply to roger hayes

    We actually considered exactly that as we have a ShapeOko not doing much right now.

    Accuracy/repeatability were a concern.

    It was at that point that we considered that pogo pins and PCBs are relatively cheap and we could use a mux/demux chip to switch between a single SEGGER.

    The $0.95 chips we bought multiplex 16 channels in both directions. So, for our panels, we'd still need two SEGGERs (which luckily we have) but that's a whole bunch better than 25 :D

    -m

  • 0

    Is anyone using something other than a Segger to program their nRF51822's in production?

  • 0 in reply to John Kelley

    Haven't tried myself. The nRF51 is a slightly peculiar SWD implementation, but I have no doubt that some of the work done with FTDI and SWDlib could be adapted (if it doesn't work out of the box).

    -m

Related