Pi 3 with Cartridge SD

[Roller Toaster]

As I near the end of my build I have been running into issues with the SD card reader. I've been looking up and down the forums and haven't found a solution to my specific problem. Most often people say to twist the data wires, but I can't really do something like that since I am using a micro SD to full-size SD extension ribbon. I have cut the board part down so it fits in the case, and length wise it is perfect and doesn't wrap unnecessarily. I can plug the micro SD directly into the Pi 3 and it boots perfectly, but when running with the extension ribbon I get "Unable to mount root fs on unknown-block (179,2)" every time. I checked for shorts, and my multimeter did beep for an instant when I touch the VCC and GND, but I traced that back to the PRS board and it seems to just be a capacitor.

my wiring setupShow

As you can see I have done my best at twisting DAT0/DAT1 together, and DAT2/3 together. They are in a small space and the way the boards are situated would make it difficult to twist all 4 DAT lines together.
When I plug this whole setup into my computer it reads it as an SD card and I can access the files fine. Trying to boot gets me the kernel panic...

Any thoughts? I have not yet tried conductive tape although I worry about making things too bulky and inflexible as a result. Space is tight when doing a Pi 3 with cartridge.

[abrugsch]

ok, easy ones questions out of the way first, have you buzzed through the connections end to end with a meter for continuity? have you buzzed adjacent connections for shorts?
have you double checked that the pins go end-to-end where they should? with a meter?

[Roller Toaster]

ok, easy ones questions out of the way first, have you buzzed through the connections end to end with a meter for continuity? have you buzzed adjacent connections for shorts?
have you double checked that the pins go end-to-end where they should? with a meter?

I have checked continuity between all solder points, and also from the micro SD end of the ribbon to the solder points they should connect to on the card reader. Everything is all clear except for vcc and ground which has continuity for a fraction of a second and then nothing. I thought that was the source of the problem at first and traced it back to the PRS board itself and not my soldering, but I don't know if it's the issue.

[abrugsch]

that'll probably be the filtering caps but you shouldn't even see a fraction of a sec tbh...
(very small capacitors between Vcc and Gnd will act as a short circuit for a very small amount of time until they fill up and become effectively very high value resistors, essentially open circuit and then allow AC portions, i.e. noise, to filter off to ground (I assume the PRS board uses decoupling caps like this) but the size of the decoupling caps should be so small you'd never notice this with a meter...

can you take some pics of the PRS board in isolation?

[Roller Toaster]

can you take some pics of the PRS board in isolation?

Sure thing! It's already been glued in but if you need the other side I can tear it up for you.

SpoilerShow

Edit: I realize also that I widened the hole at the top because I have the cartridge with the different screw post. I was slow and careful not to damage any of the connections in that area I hope.

[YaYa]

You were careful but from the pictures, i see some kind of copper on the upper left corner of the hole you made... maybe the problem is there

[YaYa]

Also, the most upper pad of the C3 (capacitor ?) is cutted. Your hole is touching it.

[Roller Toaster]

Also, the most upper pad of the C3 (capacitor ?) is cutted. Your hole is touching it.

Do you think it's significant enough that it could be the reason for the problems? I can plug the setup to my computer and read the files just fine. I thought I had cut the hole carefully enough that it was only barely in the way. I cleaned it up with an x-acto knife and now the copper you saw is gone

[YaYa]

I cannot swear that this is the issue. I just saw two "possible" problems.
A computer may handle I/O from a SD card differently than a Rpi. I cannot guarantee nothing.

[YaYa]

The talked pad could be cracked inside and you don't see it. The connection maybe is enough for the computer (maybe with CRC over I/O) and insufficient with Rpi...
Maybe not... it's really hard to say...