Dumb Electrical Question (not Imp specific, but related to my Imp project)

Being a software guy, I am very rusty on my Electrical Engineering skills…

My IMP is currently powered via USB. The 5v input voltage also powers a RS323<->RS485 Adapter. That adapter is connected to my pool controller’s RS485 line (which is used for the remote controls for the pool, I am using it to scrape info about the pool, and eventually control the pool). I currently only have the two data lines connected. I am actually seeing a voltage difference between ground and any metal part of the IMP (the SD card shield thing) at around 2V!

I actually get a little bit of a shock when I touch the imp, when this is connected (The IMP does not do this, of course, when the RS485 line is not connected. I also get a shock pulling out the USB cable from the wall, when the RS485 is connected to the imp). Reading more about RS485, it is suggested you also use the ground from the device that you are connected to…Does that mean, that I just wire ground from my pool controller to ground of the IMP and it will take care of the voltage difference (I am having some other problems, that transmission is noisier then it should be, which is likely due to the voltage difference making the signals harder to detect, I assume). I am guessing ground/common is called such, so they are “common”, but I don’t want to fry my imp. Also, I am assuming there is 0 voltage difference between the input grnd line and the output (after voltage regulation, next to the 3.3v pin) ground, right?

Hi
while RS485 uses only 2 wires to carry data, those two lines must be referenced to some common ground, within the limits of the common mode tolerance of the used RS485 driver IC.

you can not just connect the data lines from different things together and hope that all goes always well. If all devices in the system share same ground potential, then it should work if not you have problems.

You should power the imp from the same power supply that power the rs485 IC on the pool controller, or at least have grounds connected.

Thanks. That was my next try. The pool controller is outside (inside an electrical box) and the imp is inside (it is easier for me to play around that way). The “controller” line is 4 lines, 12v, RS485L1, RS485L2, Gnd. I have the RS485 lines and ground hooked up to an ethernet cable, and brought inside. I could run the power in and power the imp (which is sounds like is the best route), or I could put the imp (it is small enough), in the outside box, but then I have wifi in a metal box, will probably block the signal enough to not be worth while.

I just tried tying the ground from the pool controller into the ground of the IMP and still had a voltage difference on the metal on the IMP. When I put the RS485 ground to a metal post sitting next to me, the IMP no longer had that voltage difference… Ok, I am going to hurt myself here:-)

I am going to power the IMP from the pool controller, I guess I use the battery terminals (unregulated) to power the IMP…

Thanks

(Disclaimer: be careful when playing with AC powered things!)

If you use a USB charger type thing to power the April, this should be an isolated supply; that is, you can move its ground potential around and all will be fine.

If that’s the case, then you ought to tie imp GND (which is the metal shield on the socket as well as the GND pins) to the pool controller GND. If the pool controller is in a metal box, I’d wager that the metal box is - or should be - tied to earth too, for safety.

What are you measuring the different in potential between… earth and the imp GND when connected to RS485? If you put the meter to AC and do this measurement you may see ~0.5x your line voltage - ie 55v in the US and 115v in Europe. This is common (and expected) with isolated supplies as they will tend to float halfway between neutral and live; I always used to get shocks from my Sony laptop (metal case, with 2-wire power supply) when I had one hand on the laptop and the other hand on something earthed.

This is very very low current leakage through the Y-cap on the power supply. The Y cap bridges primary and secondary to reduce EMI.

Can you draw a diagram of the system and post it? (showing supplies, earths, etc)