I have project that uses copper tape around my 3d printed enclosures to help protect it from EMI. In my current configuration i use CAT8 shielded twisted pair to make connections between the shielded modules. all of the cables shield drains are connected to the shielding of the enclosures they connect to.
The whole system is powered by an adjustable dc psu set at 14v. The shield ground starts at the master, the next modules gets its shield ground from through the shield drain and the same for my other 3 modules. basically all the shields are grounded through the cable drains and grounded at the master to the psu negative. the first module from the psu has its own ground terminal on the shield that connects back to the 12vpsu negative
I tried to draw a block diagram to help visualize my configuration.
I do NOT connect the ground to the shields anywhere except from the master module and cable drains. I don't connect and of the shield grounds including the master to any ground except the psu 12v gnd. The only module that has ground connected is master. and all the other modules get ground from the master through the cable drains.
I should link this to aner thread i started that sort of discussed this. i would just like to focus on this grounding issue now.
Sorry, the shield drain is the drain wire in the shielded cable. the only issue it seems in having is if i touch the arduino usb header. this causes a bootloop if i hold my finger on it. but if i use one hand to touch the the grounded shield and the other hand to touch the arduino usb header nothin happens. Im wondering if my grounding paths look okay and that its okay to ground these module shields through the cable drains like this
?
Just checking.
There's also double shielded where each pair has its own shield in addition to the main shield that encompasses all the pairs. In practice they should all be in close contact though.
Yeah...as far as I can tell. But given the phenomenon you describe apparently it's not OK. The explanation for that behavior is not in your diagram, though.
Btw, not sure if I got it right, but you do mean that none of the modules has its [circuit GND] electrically connected with the common [shield GND], correct? Or are [shield GND] and [circuit GND] connected in one or more places? If so, one, or more, and where?
And where is [safety GND] connected to all this, if at all?
What is the nature of the [shield GND] as it stems from the power supply; can you post a picture?
If [safety GND] and [shield GND] are not electrically connected anywhere, what is the measured potential between both?
Here may be the problem. If your copper tape has adhesive, one wrap will be insulated from the others, so in effect you have a big coil of wire wrapped around your enclosure. If you want to ensure that it really does shield, then solder copper wire across the turns of tape, about every 1 inch or so.
shield gnd and circuit gnd are not connected anywhere expect at the 12v psu negative terminal. all of the modules get circuit gnd from through a wire in the shielded cable.
no i mean the strands of wire behind it. i am using CAT 8 shielded cable with individually shielded twisted pairs
I don't have an earth ground. previously i *was getting problems when i connect earth gnd so i connected the shield to 12V power supply ground. I'm pretty sure my power supply has an ac transformer and its providing some galvanic isolation if that matters.
So i should measure the voltage between my chassis ground and the earth ground? AC or DC?
I check in various spots on the copper tape. it conduct very good thorough the adhesion believe it or not. it was between 0.0 - 0.2 ohm if this is okay
Well, mostly for shits and giggles, yeah. I'd check both ac and dc. Do make sure you set your meter to its highest range initially and then dial back if it's only a little.
Anyway, still no very clear clues as far as I can tell.
In practice a lot can depend on how physically tidy everything is assembled; that remains of course the achilles heel of your setup.
Should i attach the arduino usb header to the shield ground? should i attach the arduino to the shield gnd? it seems like my body builds up some capacitance and when i touch the arduino it drains through that.
Its crazy how many different things can affect the shielding.
Correct me if i'm wrong but if i connect the arduino logic gnd to the shield gnd this will create a ground loop?
Ill measure the voltages between things and report back with the results
It's connected to circuit ground. You're not,and since nothing is connected to earth ground, my best bet is that you mildly zap your circuit with a high voltage, low energy pulse whenever you touch it, which trips the uC.
As far as I'm concerned thr best solution is to connect your shielding to earth ground together with circuit ground in one place. You mentioned that doesn't work, so IMO that's what you need to figure out, why that doesn't work. You also mentioned that you suspect there's a transformer in your PSU that galvanically separates the low voltage side from the grid; well, I can assure you this is the case and it does in no way explain why your PSU would malfunction if you connect circuit GND to earth GND. In your place I'd try to idolate THAT issue and if turns out to be thr PSU, try a different unit. There might be something wrong with the one you have; a leaky cap or a heatsink contacting the case where it shouldn't or so.
Ill be sure to add testing the resistance between modules to the list off things to report back with. i did however test them individually locally and the layers conduct very good.
Did you use a standard multimeter for this? AFAIK those are not very suitable to measure very small resistances. At least mine aren't; theres no way I could tell the differemce between 10mR and 150mR with the ones I use the most.
I'm about to use a 20 dollar digital one. idk how accurate it is but it 0.0 when i touch that probes together so i'm trying to get as close to that as possible. I think this should be good enough for this purpose?
I just checked between every
module
shield
cable shield
I measured 0.00 ohm no matter where I probed.
I'm not measuring any volatge between the shields and earth, the USB and earth and between shield and arduino usb header.
I have been told yes and no about attaching earth ground to my shield.
Well, some improvement at least.
Does the arduino still crash if you ground yourself before tocuhing it? I.e. first touch the earth-grounded shield of your project and while holding on then touching the usb port.
