0

Before I ask, I did see other similar posts, but it's not easy for me to apply them (I'm not an experienced EE..).

My question in brief is, how to get common ground, to avoid potential gremlins. (Or, if it's ensured by board design than it's all good .. )

So I have a devboard. It has 3 options to be powered, one through USB, other is from 3.3V regulated, from batteries, and also 5V from external power supply. I need to get away from the first option - through the default USB 'debug' plug, but, I do still need it connected - for VirtualCOM.

I'm also in need to connect external JTAG to it, through dedicated JTAG connector. VirtualCOM and JTAG would be connected to same PC. Mmm, so as I understand, I still have 2 different/uncommon grounds here : my external 3.3V (or 5V) and my PC (VirtualCOM, JTAG).

I do follow devboard guide on setting jumpers right, to supply board externally. But, not much about ensuring common ground ... ( I guess the basics are obvious, but .. not in my case).

I'm worried for gremlins in my setup, which will cause me weird issues, and potentially cause me hours debugging (for me, that's SW ..).

The board guide/description is here : SimpleLink Launchpad kit user guide

I follow sec 2.5.4 Power Connections, 3.3V input (for now, but 5V also an option..). I also do as sec 2.6.2 Battery Power - again more jumper settings.

The design archive with schematics is here : archive with design files - my board is CC3220SF_Sch

Unfortunately, don't know how 'common ground' works, so for that purpose schematics not for me ;( ...

v01d
  • 57
  • 5

1 Answers1

0

JTAG has ground pins, which most likely internally are connected to USB grounding of your PC. USB has shield and ground, that should be connected to the same ground as JTAG. Your PC is grounded through the power supply to Earth. The external supply most likely is not floating either, meaning, that it's connected to the same ground as PC. This will create a ground loop from your supply through USB-PC-Ground and you might see extra current consumption.

Generally digital electronics like that don't have too much trouble with such setups (I have such 2 setups next to me - JTAG/USB on prototype board, which is powered by an external PSU). It becomes an issue with sensitive analog measurement circuits, but that's more board layout issue IMO. Also weird things start to happen, if you try to power your board from USB and external supply at the same time. That's when things start to go pop.

stiebrs
  • 789
  • 4
  • 9
  • hey, thanks for answer. My power supply is a mini cheapo eBay little board, which supplies either 3.3V, or 5V. It is powered through a USB cable, which I plug into just special socket on my mains (you know, they sell now days those extensions where you can put your cellphone to charge say). The weird thing I observe - tbh not totally sure its weird , duh - but, say I turn off my supply (my external 3.3V), then, if the JTAG still on/debugging , then I will observe that the board has some power/voltage on it - so it seems my JTAG providing it..I see leds very dimmed/barely on. ... – v01d Oct 30 '19 at 10:08
  • .... Is that normal? As soon as I stop the debug session, JTAG seems to cut off that power, and all leds become off - thats what I would expect when I turn off all power. Should this be happening? – v01d Oct 30 '19 at 10:08
  • I'm looking at that Cortex connector I'm using : [10-pin connector](https://electronics.stackexchange.com/questions/302444/new-debug-connectors-for-arm), and it has some 'magic' to me too: like , 2 grounds/GND, and also GND Detect; and per that link looks like they usually tied to board ground.. It also has that VCC, which (per link) does "VCC line is just to tell the programmer the IO levels of debug interface." But , should I be seeing voltage flowing from JTAG to board , with board supply is off ....? – v01d Oct 30 '19 at 10:14
  • Yeah, JTAG can leak some current to the target board, it depends on particular set of devices. My JLink doesn't seem to do that, but my old Olimex debugger used to do it. If target MCU pins in shutdown state have weak pullups, current can flow from JTAG (DIO/CK/RST and other) pins through MCU internal pullups into your power bus, which lights up some leds or might even power some ICs – stiebrs Oct 30 '19 at 13:39
  • thanks for your help. Me too, using JLink, but cheap EDU one (don't know if they necc. worse). – v01d Oct 30 '19 at 16:53
  • do you know how I can prevent the current leak? some custom connector mod ... – v01d Oct 31 '19 at 21:10
  • It shouldn't affect your normal operation (without the debugger). You could try to limit the current with series resistors on JTAG pins, but that could affect your signal integrity. I haven't had any issues with those small currents, so haven't investigated any remedies myself – stiebrs Nov 01 '19 at 13:21
  • Thanks stiebrs agian. I will then ignore if it doesnt' cause problems, but I do have some 'weird' JLink issues with it, and I'm not sure why. – v01d Nov 02 '19 at 03:08