LORA - Serial interference. Workarounds?

T

Talkiet

Well-known member
HI all - no code here because I don't think it's the code. Please tell me if I'm possibly wrong about that.

I am using a Teensy 4.1 with a Ublox GPS on Serial1 and an external Ebyte LORA transceiver via a UART to RS485 converter that's on Serial5.

Serial1 is at 115200 and Serial 5 is at 19200.

The main loop keeps NeoGPS fed thusly

Code: 
  if (gps.available( gpsPort )) {
    while (gps.available( gpsPort )) {
      fix = gps.read();
    }
  }

and occasionally (as a result of an input being detected) the unit fires out a 28 byte packet to the RS485 port for the LORA to transmit.

This all works. The remote end gets the packet and parses it and all is good.. BUT...

The moment the LORA unit transmits, the GPS time stops updating. If I fully re-init Serial1 (via a keypress) then it recovers, and most damningly, if I physically move the LORA box and the GPS unit physically as far away from each other as their current leads allow (30cm), then the GPS time DOES NOT FAIL and carries on just fine. The moment I move them back close together and resend another LORA packet, the GPS time stops updating again.

So from that, I am inferring that this is not a silly code or serial buffer problem, and is in fact some serious interference from the LORA that disrupts either the GPS or the Serial1 badly enough that it loses sync(?) dies in another interesting way. I have also removed power from the LORA box (but left the RS485 converter in place) and this also results in no bad behaviour.

It's not feasible to physically separate the LORA box and GPS unit as they do need to live in the same enclosure so I am looking for workaround inspiration. At the moment my idea is to close the Serial1 for GPS before sending a LORA packet and then opening it again after - but I was wondering if there's something clever I haven't thought of yet...

20221207_104258 (Large).jpg

Any advice / inspiration gratefully considered!

Cheers - Neil G
 
Sigh. I knew I should have done 5 mins more testing... Just observing the PPS LED on the GPS unit shows that it's the GPS unit being badly affected by the LORA... The PPS stops for a few seconds when the LORA transmits. I will have to look into whether it is rebooting and why. (interference, power sag etc)

I hate hardware :-(

Cheers - N
 
I was wondering how you are powering the different units?
For example, do either of them connect up to the Teensy 3.3v pin? How much power does it require?
The 3.3V shows max of 250mA.

When I was playing with LoRa, I was using circuit with it's own 3.3v VR
 
I would put the GPS module and the purple board inside a metal box. That way it will reduce the interference from the LORA transmitter.
 
The LORA box is powered from a 9v buck converter and the Teensy is powered from a 5v buck converter. On this test rig they are both powered from a large 20v drill battery.

The GPS is powered from the 3.3v VCC from the Teensy.

I don't really think power is the issue as when I physically move the devices apart the problem goes away. when I move them back close together it dies again...

Currently bodging together a shielded box to put the GPS board inside to see if that changes the behaviour at all.

Cheers - N
 
You might also want to add a ferrite choke on the cables from the LORA transmitter to the purple board.
 
skpang said:
You might also want to add a ferrite choke on the cables from the LORA transmitter to the purple board.
Click to expand...

Interesting. This is easy as I have a few lying around... I know very little about magic airway electricitymagics - are you suggesting the interference may be feeding back (?) into the board from the LORA box? I imagined (with no basis except for it feeling right) that the interference would have been from the LORA transmission into the GPS receiver or chip...

Still, can't hurt to try!

Cheers - N
 
Fascinating...

In the interest of replicating the issue with slightly different hardware to eliminate a broken/flakey unit... I swapped in a different GPS module... I was using the larger Ublox M8N with integrated antenna (hmmm) and swapped in the smaller 7M instead (no integrated antenna). (I was using the external GPS antenna in both cases).

The problem doesn't occur with the 7M at all. I have held it up against the LORA aerial at the top, middle and bottom, I have held it against the LORA unit case (with a sheet of paper between) and I cannot make it fail with the smaller unit. I will try another sample of each but I wonder if the integrated GPS antenna on the M8N is contributing to the issue.

I have a few more GPS module types so I guess I'll swap them all in and look for common behaviours.

2GPS.jpg
Cheers - N
 
Diagnosis of power not the issue given the distance test seems valid.

Just for ref: Years back working with a GPS and a T_3.6 here I could power it from Teensy 3.3V.
While @mjs513 using a T_3.5 and AFAIK the same (Adafruit Ultimate) GPS it required use of external 3.3V IIRC.
There may be a power draw change between the M8N and 7M. So quick test with second 3.3V supply (2nd Teensy) and common GND might be easy to rule it out.
 
defragster said:
Diagnosis of power not the issue given the distance test seems valid.

Just for ref: Years back working with a GPS and a T_3.6 here I could power it from Teensy 3.3V.
While @mjs513 using a T_3.5 and AFAIK the same (Adafruit Ultimate) GPS it required use of external 3.3V IIRC.
There may be a power draw change between the M8N and 7M. So quick test with second 3.3V supply (2nd Teensy) and common GND might be easy to rule it out.
Click to expand...

Yep, well worth trying... I'll try that tonight when I take a deep breath and test all 6 GPS modules I have against it... I also aborted the shielded case for the module at this stage until I try all the modules I have. I'd love to find an issue apart from "interference" as that's the hardest issue to be sure I would have fixed.

Cheers - N
 
You must log in or register to reply here.
Back
Top