Failed first test, no LEDs

[Daniel Crawford]

Hi all,

I finally got around to putting the kit together, and yesterday hit the point of first test. So the pi's set up, plug in, aaaand no LEDs.

Spend a little while playing around with everything, then remember the high power requirements. Noting that I had a few other things plugged into my powerboard, I moved to a different, empty power point, aaaand LEDs light up.

Unfortunately, about half don't light up, and the rest are very dim. So set it aside and come back today.

Today, I have a poke around, look for poor solder joints, short circuits etc. Plug it in (to an empty socket) aaaand no LEDs. I've tried a few power supplies, and power points, and nothing works.

Any ideas on what's going on?

- Pi 4b

- official power supply

Cheers

[Oscar Vermeulen]

Daniel,

Can you mail me a picture of how the front panel looks (like the ones on the web site?).?That about half the LEDs light up sounds OK in fact, if most of them are blinky or 'animated'.

Lift up the LAMP TEST switch to test them all. And you can use the python3 test script in /opt/pidp11/bin for more testing. Make sure to leave the PDP-11 simulator first: CTRL-E, exit.

If all is not how it should be, some first pointers for the bug hunt:

- did you put some cardboard or plastic (not just tape) between the Pi's metal USB cases and the PiDP board? Symptom: a few VERY bright LEDs in the middle of the panel.

- most of the time, big trouble (not just a dead LED or two) is because of solder problems on the GPIO connector. And most of the time, the following fixes it:

  -- reflow the solder on all GPIO pins, make sure there's no shorts. But some shorts are not easily visible, so reflow them all.

  -- afterwards, clean between the GPIO solder points with alcohol (vodka is fine too, I am told, and even a moist cotton tip with water or a fine whiskey is OK).

     Sometimes the solder flux, the goo you might see around the solder points causes a current leak. I once thought that was nonsense as I've never cleaned a soldered-up PCB myself, but it really is true.

The high power requirements are just for the Pi 4 itself. The PiDP board adds no meaningful extra current draw! So if your Pi is fine on its own but not when plugged in, there really is a problem.

Let me know if the above steps fixed your issue. If not, we need to delve deeper. Time for a bug hunt!

Kind regards,

Oscar.

[Daniel Crawford]

Hi, thanks for the suggestions!

- I'm trying to get a photo, but the LEDs are quite dull and work intermittently (I did briefly see them light up just now testing)

- I'm using the official power supply, although looking closely it's actually rated at 5.1V not 5.35V. I'm not sure that's significant. Actually, I can't find any official 5.35V power supplies around here, although I could probably track down a non-branded one (which might be iffy)

I'll try those steps over the next couple of days and report back

Cheers

[Daniel Crawford]

Solved!

I missed the diode next to the IC. Added that and all the LEDs light up, at full brightness. All appear at the same brightness so looks like everything is put together correctly

Shorting the test switch results in a test pattern, and running the test script appears to be ok.

Note that I did touch up a couple of pins on the IC socket and also the GPIO header, but I suspect that was unnecessary.

For future readers:

- the official 5.1V power supply is sufficient

- I did insulate the Pi, but actually in my case there is sufficient clearance between the pins and the Pi, ymmv

- check all the diodes before first test :)

Thanks!