Hey everyone, I'm working on this system that may or may not be suffering from ground loop problems. I'm VERY new to this kind of work, so new that I've yet to garner any experience identifying these sorts of common issues. Here's a block diagram of our system:



All the Tx/Rx lines between the Red Pataya and the Transceiver are connected via coax cable. Everything "works" (the Red Pataya can send/receive signals to/from the transceiver) until we turn o the 15V power supply, in which case it call crashes (data stops sending/receiving). Everything is powered from the same power bar coming from a single AC outlet.

I've been doing quite a bit of research into ground looping. I see that it's common when connecting instruments via shielded coax cables. However, in this case, I don't see how it's possible to separate the 5V and 15V grounds, given they're tied together within the Transceiver circuitry. Furthermore, would optoisolators on the Tx/Rx lines (between the Red Pataya and Transceiver) really do anything, given their grounds are tied together via the Transceiver board anyways?

Posting on here in case there's something obvious to more experienced eyes. Any pointers would be greatly appreciated!

 

Also, maybe someone could save me potentially blowing something up: what would happen if I just disconnected the ground connection from the 15V power supply? (The one on the supply side - not the AC side!). My thinking is the 15V signal would just go into the 5V ground via the Transceiver. But my thinking may also be way off base...

 

hello zurn. dont disconnect - from the power supply. you suspect ground loop. since we're not sure. i would try to check with a multimeter or similar to check why red pataya is disconnecting. maybe noise in the inputs? i am not familiar with this setup, is this sort of a sdr transmitter receiver? you could add ferrite beads or a simple choke, but i would first check out why the pataya is disconnecting.

 

Hey Joao... ya I've spent quite a bit of time trying to uncover why it's disconnecting. At first I thought it was an impedance mismatch, but looking at the schematic of the Transceiver it doesn't really add up (both inputs are op-amps with very high impedance), and plus, it works without the 15V line on.

And yep, you're right abotu the Red Pataya. It's just a little single board computer that can be used for, among other things, an SDR tx/rx.

Well, I'll keep at it. I was mostly just hoping to poll the community of any obvious signs of ground looping in my above block diagram. It's a tricky bugger!

 

I'm no expert, but a block diagram doesn't really show how things are connected in practice.
Someone who knows a lot more than about it than me, always stressed the need for "star" grounding.
I found out recently what can go wrong if you get ground wrong.
I had a GPS module running remotely with its own DC adaptor, connected by CAT5e to a PC running a Pico oscilloscope.
As soon as I connected the ground terminal at the PC end, it crashed the PC.
Won't do that again - next time, it gets powered from the PC USB.

 

You might gain some insight by trying to power up in a different order or connecting 15V with the red pitaya disconnected from the transceiver. Then check potential between red pitaya ground and transceiver ground. Then connect if OK.

 

Hey all, I actually zeroed in on the problem: the external clock (will eventually be used to run three Red Patayas in tandem, though I've only been connecting one of them at this point). I set everything up with an older Red Pataya configured to run on an internal clock and everything works, no shutoff when the 15V line gets switched on.

I'm going to zero in on the clock circuitry and see what's up. As far as I know I'm doing the "star" grounding method (I believe it's the same as "mecca ground", do correct me if I've misunderstood), but something tells me if there could easily be a ground loop it in the clock signal to Red Pataya path.

 

Problem solved - it was a busted transducer, not even depicted in the above block diagram. Everything works as it should now.

A wild goose chase yes, but hugely educational