PiDP11 + NVME crashing due to flux residue

[Geoff Williams]

Hi Group,

I wanted to drop a note here on how I fixed up my PiDP11 in case it helps anyone else (or it breaks again!):

* Raspberry Pi 5 8GB
* Official Pi 5 power supply
* Ethernet connection
* Crucial NVME 2TB (CT2000P3SSD8)
* ORICO M.2 NVMe SSD Enclosure[1]
* 2 Ports USB 3.0 Male to Female Extension Cable for Front Panel Screw Mount Type (for easy USB access)
* Latest firmware on Raspberry Pi and USB enclosure

I've been running for over a year with this config, after a previous NVME USB drive crashed daily. It's been quite good but recently had started requiring a reboot every few days so I decided to upgrade to the Pimoroni NVME Base[2] as I thought this may be power/cable related and had had enough of fiddling with enclosures.

I thought this would be a 5 minute job. The base bolts right on to the PiDP-11 nylon spacers and worked right away until I reconnected the PiDP to the GPIO port. At this point I had to plugin HDMI display and keyboard, then I observed:
* Most boots failed completely (crash)
* Sometimes a boot would go very slowly rather then crash outright (5-20 mins)
* Disconnect GPIO, fixes all problems again

At this point I suspected power again and stumbled on a post about low voltage indications[3]. Following the advice from Oscar about reflow/flux residue I took a closer look at my board. The entire board was covered in flux residue and sticky to the touch,  especially around the GPIO connector. See before pics. I cleaned up the whole board with isopropyl alcohol and a toothbrush about 3 times until it was no longer sticky and after this the system booted perfectly with all devices attached.

After testing overnight and observing stability, I reassembled the case and found one last hurdle: when I put the back on, some LEDs are really bright and others are dim with the default light show and the knobs switched the lights randomly. Turns out this was just the USB extension sockets pressing on the assembly and the solution was to carefully coil these to avoid pressing. After this, the whole system works perfectly and has now been stable for 10+ days.

Anyway, hope this helps someone, its a great kit and a lot of fun. If you are getting weird NVME errors with this hat or your USBs, try inspecting for flux residue.

Cheers,
Geoff

[1] List of compatible enclosures - https://jamesachambers.com/best-ssd-storage-adapters-for-raspberry-pi-4-400/
[2] NVME Base product info - https://shop.pimoroni.com/products/nvme-base?variant=41219587178579

[3] Low voltage indications - https://groups.google.com/g/pidp-11/c/luSNcZd05Ik/m/8c5pShKzBgAJ

[oscarv]

Geoff,

>> Following the advice from Oscar about reflow/flux residue I took a closer look at my board. The entire board was covered in flux residue and sticky 

>> to the touch,  especially around the GPIO connector. See before pics. I cleaned up the whole board with isopropyl alcohol and a toothbrush about 3 

>> times until it was no longer sticky and after this the system booted perfectly with all devices attached.

I've learned that excess flux between the GPIO pins is indeed the #1 source of 'mysterious' problems. As in, of course a bad solder point somewhere on the board is the most-occuring problem. But the hard problems - they are often flux related. It's funny because I had always thought that the need to remove flux was a bit of an urban legend. Not so.

 

>> After testing overnight and observing stability, I reassembled the case and found one last hurdle: when I put the back on, some LEDs are really bright 

>> and others are dim with the default light show and the knobs switched the lights randomly. Turns out this was just the USB extension sockets pressing on the assembly and the solution was to carefully coil these to avoid pressing. After this, the whole >> system works perfectly and has now been stable for 10+ days.

This is REALLY important: 

More elegant solutions exist, of course: one is to tape a small rectangle of plastic onto the PiDP PCB itself. There's a footprint legend that shows where the USB cases will be when you insert the Pi into the PCB.

Kind regards,

Oscar.

[Bert Driehuis]

I just cut a piece of PET foil from a window box (https://www.ptpackaging.com/blog/what-is-window-box-packaging/), using an X-Acto knife to cut out a rectangle for the connector, and a perforator to create holes for the standoffs. It looks nice and does a great job of isolating.

Hmmm, now that I see it in print I should redo the sheet: the holes don't look that nice and the edges aren't straight. I see an excellent excuse to go buy some pastry in a window box!

With kind regards,

Bert

--
You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/7b926b86-daab-466a-8eef-5e6d19d16e6an%40googlegroups.com.

[Geoff Williams]

I do have the cardboard under the pi but I also have big USB extension leads attached and the thick wires were mechanically squashing the PI and either breaking the GPIO connection or causing enough pressure for my dodgy soldering to breach through the cardboard. Now the back is bolted on again I don't want to look(!)

You received this message because you are subscribed to a topic in the Google Groups "[PiDP-11]" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pidp-11/ue9SmMnIqlE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/CAN0x2A-T5dQWRy%3DOyDRYZ8-ZHNGcrVKsdx0RpZ6ZF-omXb_1mA%40mail.gmail.com.

[R Clark]

What I did was just put a gpio extender on the rpi.  That of course separated the board from the PCB.  Then just longer stand-offs. Even helps with cooling.

[terri-...@glaver.org]

On Wednesday, April 29, 2026 at 11:53:06 AM UTC-4 ge...@declarativesystems.com wrote:

I've been running for over a year with this config, after a previous NVME USB drive crashed daily. It's been quite good but recently had started requiring a reboot every few days so I decided to upgrade to the Pimoroni NVME Base[2] as I thought this may be power/cable related and had had enough of fiddling with enclosures.

I thought this would be a 5 minute job. The base bolts right on to the PiDP-11 nylon spacers and worked right away until I reconnected the PiDP to the GPIO port.

If you're one of the people who wants ro run your PiDP-11 with the back cover on AND you are using the DB-25 cutouts, you may need to flip the Pimoroni NVMe over (like an open book). This does block access to the Pi's SD slot, but if you're having an NVMe issue you'll need to open the case anyway...