Uno R3 and USB Host Shield failure

Projects General Guidance
1

Hello all,

I’m having a curious problem with an UNO R3 and a USB Host Shield:

With the shield mounted on top of the Uno in the standard way, it works fine.

I’ve done this with many projects and many boards.

However, because of space limitations, I decided to wire the boards together to lower the height. I started with the host mounted normally, test my code, and all worked as expected.

Then following the documentation, I wired the active pins, but the USB signal didn’t get to the UNO. So I progressively added more connections, with continued failure. I wound up connecting every pin on the host to the Uno, including the 6 pin ICSP header, and it still doesn’t work, even after verifing continuity between every connection. I’m baffled.

Any suggestions? Thanks, Aaron Michael,

Retired EE with 20 years of Arduino experience.

2

Please check the wiring again and create schematics for the wiring. Something is clearly missing.

3

Bad connector or solder joint or overheated and broke connectivity at the board.

4

Thanks, but I checked the wiring and did continuity tests. Wiring is 1 to 1 and pin to pin between the boards. No, I didn’t drew a schematic. In this case it’s much easier to see the actual jumpers between the boards.

5

Thanks, but I checked the wiring and did continuity tests. Wiring is 1 to 1 and pin to pin between the boards. All good. I wish it were as easy as your suggestion, but thanks.

6

Did you solder the wires or use dupont jumpers.

7

I used jumpers and checked continuity on the boards’ solder vias.

8

Wiggle the jumpers randomly; it's not unheard of for there to be marginal connections at the ends.

9

Thanks, but been there, done that. I’ve tried all the usual tricks I’ve learned from my 65 years of designing and building electronics. Only one I haven’t tried is hitting it with a 10 inch crescent wrench.

10

Well ther's your problem, you need at least a 12 incher.

11

Do you have a schematic for that shield?

12

Hi, Yes I have the schematics, but thanks.

13

Solved! I have it working reliably now. The problem seems to be the unreliability of the header sockets on these Chinese boards. The contacts inside are 12 mils thick forks, very floppy with the prongs about 24 mils apart. Good quality jumpers have pins 25 mils square. More common jumpers have 25 mils round pins. They all mostly connect but the more connections, the more likely one or more doesn’t. And pins must connect at both ends so odds against go up fast. (For our non American friends, mils are thousands of an inch, 25mils = 0.635mm).

So with good jumpers, carefully placed, redundancy where possible, and a 6 pin harness I made with with a good 6 pin header and socket all soldered to connect the ICSP headers all works well. Now that I’ve found the problem, I’ll go on to solder all connections and likely delete the redundant connections. Thanks to all for the suggestions. Aaron

14

Looks like you nailed it.

15

You are welcome
Have a nice day!

16

But will he mark post 3 as the solution?

17

If you send me $10 I like post #3

18

Probably not.

19

Wouldn't be the first time an engineer eventually found a solution long after a high school kid on a summer job did. I love the attitude too, nothing has changed.

20

My problem is that since retirement I don’t have any high school kids working for me any more.