I have a sim racing wheel stand, metal, attached to the stand is the Fanatec DD1 and a couple of small cheap tablets all USB connected to the PC.
I'm building a shifter setup, aluminium cased, I need two (it's complicated) so I have them connected with a 4 pin jack cable, with a Micro in one of them, again connected USB. This works perfectly when on the test bed, but when I bolt the shifters to the frame I get really spurious results, the Micro starts reporting all sorts of output when there is no human input. Take it off the frame and it works perfectly again.
I've also found if I turn off the DD1 wheel base it all works perfectly. So do I need to earth the arduino to the aluminium frame of the shifter box? And more worryingly what current or signal is the DD1 putting into the frame? Is the DD1 earthed to the frame through the bolts, or is there current leaking somehow?
Although I'm quite handy with a soldering iron and have put together some crazily complicated stuff my basic knowledge is somewhat lacking!
Can you show a schematic or drawing of what is connected to the Micro boards ?
Are the pullup or pulldown resistors strong enough ?
Couldn't you find something better then a 4 pin jack cable ? Are the contacts of the plug and the sockets at least gold plated ? Is it a shielded cable ?
I assume that the Micro board is not connected to the aluminium case ?
The Micro board should be connected to the computer and not to something else.
So it is just the electrical noise of the case that influences the Micro ? That is weird.
A grounding problem is for example when the current in a ground cable has multiple ways to return to the power supply. The current might decide to choose a longer route via thin wires instead of a short thick wire.
Are the cables hold together with tyraps ? The cables with small signals should not be close to power cables.
A drawing wont show much, there are 4 microswitches, 2 in each box. The cable just connects 4 wires to the switches in box 2, but the problem is exactly the same without the cable connected so that's not the issue. But I agree it's not ideal, just what I had on hand.
So, Metal frame.
DD1 bolted to frame.
2x shifter boxes bolted to frame.
Shifter 1 switches wired to pins 10 and 14. (Micro is in Shifter 1)
Shifter 2 switches wired to pins 15 and 16.
All pins are PULLUP.
GND is taken from the Micro and shared from a chocolate block.
The Micro and usb cable are shielded from the case with a simple 3d printed cage.
There's definitely no loose or wayward wires, there are no connections touching the box at all.
What I find most confusing is the problem goes away if I turn off the DD1. I tested yesterday with the shifters on a table, not connected to the frame at all, worked perfectly. But if I put one hand on a shifter and one hand on the frame (with the DD1 on) I got streams of random incorrect data.
I've tried shielding the boxes from the frame, but of course there are other paths of me touching the frame so that makes little difference. I can't think of a way to shield the DD1, I can print a panel to put under the unit, but the bolts are still needed.
I previously built a button box with all sorts of joysticks/buttons/toggles and dials on it, this worked perfectly, but it was in a plastic enclosure box.
A drawing would help a lot ! Schematics, photos, links to used parts, that would help.
Just the weak internal INPUT_PULLUP ? That is not good enough for any final project. Start by adding 1k (220Ω to 4k7 or so) pullup resistors. If that is not enough, then perhaps optocouplers can be added.
It is hard to understand for me. I really don't know how all the wires are going and how it is connected.
You have not convinced me that the cable is not the problem The whole project should be able to deal with some noise.
When everything is connected, but not powered, can you use a multimeter to check for leaks between different parts of the project ? A test for capacitive leaks is even better.
If you power it, how much voltage does the Micro board get ?
Everything matters when dealing with noise. I can probably mention 100 things that can influence it.
If you, for example, have a cheap clone "Micro", then it might not have the right amount of decoupling. The Micro board could be very noise and pick up noise very easily.
A cheap power supply could feed the whole project with an enormous amount of noise.
... and so on, up to 100
If you can find the cause and fix it, maybe you should isolate the aluminium box from the frame as well. Just to be on the safe side.
The MIcro is genuine and getting full voltage with a premium cable, I don't use the cheapo things, already had too many bad experiences with those.
As mentioned I've tried to isolate the shifter boxes, but as I have to touch the wheel on the DD1 my body becomes a short.
Again, as for a drawing, there are 4 micro switches on pins 10, 14, 15, 16, sharing GND from the GND pin. That's it.
Honestly, forget the cable, if I take the cable away and only work with shifter 1 it's the same, there are no differences in the data with or without the cable.
I've just tried something daft and this is boggling me even more. If I connect a wire to the outside of one of the boxes and leave it dangling the problem goes away. What on earth does that mean?!
On the subject of schematics, I stress again, the whole setup works perfectly until I bolt it to the rig. I don't have a noise problem in my setup, I have a noise problem coming from the wheel controller.
As a temporary solution I've isolated the shifter with shims and nylon bolts but this isn't ideal.
Could someone please include a link for the components you suggest I add to the input lines? I'm really tight on space, I can't add anything where the micro sits, I'll have to put them in line on the input wires and shrink wrap them, not sure if this is optimal, but I don't have much choice.
The whole project should be able to deal with some noise.