SNR drops after a short period, does not recover

running Reachview 2.5.3 on this module.

I am having a strange issue where the signal strength shown in Reachview drops by [edit] 5dB after a period. When cold booting the signal drop occurs after ~40 seconds. If the antenna is unplugged and reconnected, the initial signal appears good, but then after around 10 seconds drops right down again.

The drop is enough to reduce 10 green sats to 2 to 4, and the signal seems to gradually deteriorate after then, so after a few minutes only 1 or 2 sats are in the green. I noticed the problem when trying to get good signal on one of my rover units, but it is occurring with the module I have configured as base as well. I have tried a couple of stock tallysman antennae (on copper plated cd ground planes which have worked well in the past), and with a maxtena m1227hct. They all exhibit the same behaviour. I have tried with power via USB and from the Pixhawk 5V pin.

Raw log attached - the antenna was positioned with good view of sky, on a sunny afternoon. The initial dropout occurs around 43 seconds into the log. I then disconnected and reconnected the antenna twice, so the logs should show the signal drop to zero, then recover with plenty of green sats, then around 10 seconds later, the signal strength can be seen dropping back to low levels.

The practical effect, of course, is to make achieving a solid float/fix solution rather unlikely.

I am uploading a (very boring, slightly out of focus) video of the Reachview display, showing the initial good signal following boot, the initial dropout at 42 seconds and then shortly afterwards the effect of unplugging/replugging the antenna:

Any pointers on what could be going on? I had been blaming my setup on the drone in question, but the issue seems independent of drone or other local electronics which could be causing interference.

I should add that I have previously had good results wit all my reach units. It has been a while since using the system for work, so I’m not sure exactly when the problem started occurring.

raw_201707181352.zip (495.4 KB)

edited - it’s more like a 5dB signal drop

First thing first. Update Reach to 2.7.0

Is this first time happening? Only at this location?
Done any other modification to it?
How about running power from a batterybank with min 2amp outout?
Are reach supporting other device with power?

Multiple locations and versions of Reachview. It’s hard to be certain how long the issue has been going on. I have tested this on 2 of the 3 Reach I own, and will have a look at the other when I can. I’ll update to 2.7.0 and test again tomorrow - I have been holding off firmware updates for a while having been bitten by the Galileo/missing events issue. Currently only updating if I have time to run a proper test of the new firmware before using it…

In this case I was powering the reach from a lipo to usb adaptor, as I always power my base station in the field. The physical setup of the unit I was testing today was unchanged from my usual setup, which was working when I last had cause to use it. I have had the same issue powering from the DF13 Pixhawk, which has also worked fine in the past. And nope, nothing else powered from the Reach.

Cheers!

Funky… Same issue on more then one device?

Yep, seems to be. From different batches as well, one is from the second batch I think, the other was about 6 months newer. I’d like to revert back to some earlier firmware and see if that has any effect, not sure if the old firmware is available anywhere? The galileo ublox update might prevent reversion beyond the point that was released possibly, otherwise I would just flash back to the original Reach image…anyone know if that would cause any issues?

The timing of the signal drop out seems reproducible across multiple units as well, which is really weird. I thought at first it must be some sort of power issue, or something to do with some other hardware starting up or something, but as far as I can tell, removing as much other hardware as possible from the equation hasn’t made a difference, and neither has running through a few different ways of powering the module.

Any thoughts on possible causes/solutions very welcome!

What is the data correction link?

I’m investigating a similar problem. In my case, it seems that on my TCP connection, I had several servers accessing the base data. The Client Rover, if I understand correctly, on some setups will respond to the first server to respond if the servers both have the same address. The screen signal levels look very much like the video.

In my case, I could not get a good FIX, but when I found the other server and disabled it, I immediately saw the SNRs stabilize and went from all FLOAT to good FIX.

I am still investigating how to prevent this.
This , I think would only occur if the correction is TCP Server-Client.

Your problem may be entirely different, but thought I would suggest looking at correction link.

Would hook up those bad units and run them at the same time to see if the issue is synced.
Never heard of multiple units with the same issue. Interesting problem btw :sweat_smile:

No connection, this is running as a PPK setup so I don’t bother with a correction link.

Got it… Thanks…

Add a little more here. After looking at .UBX file in PPK, it does look like an SNR problem, possibly related to start up mode.

I ran a test on my Base Reach which I am in the process of setting up. I can control the environment a little easier. The signal SNRs are a little lower than open sky as the antenna is under a rooftop skylight dome.

I was investigating whether using USB power from a USB hub might be introducing some interference. What I found was that the USB hub actually produced better SNR results than a good battery pack. About 12 SNRs in the green versus around 6 with the battery. I also observed a change at around 30-60 seconds after boot up. This suggests a hardware-software interaction, but is only a possible clue.

For info…

very interesting, thanks for ths info. I’ve not had any more time to investigate but will try to be a bit more systematic. Stuck in urban canyon for a few days while we try to trim to negotiate a very tight tunnel in our boat…I did have a fiddle this morning, the easiest thing to recreate seems to be the plugging/unplugging of the antenna, where you (or at least, I) can see may sats in the green immediately after plugging the cable, then signal drops after 10 seconds. Would love to keep those bars in the green!

Hmmm… Maybe an receiver input saturation possible. Such problems are not easy to detect. I know in the old days, you could get crosstalk between channels that would lower SNRs but not signal levels. It’s possible that, if receiver input amplifier biases take time stabilize and clip the input signals, then it could explain this. But, this is just pure speculation at this point. I don’t know enough about the new receivers to troubleshoot this possibility. Generally, lowering the input signal levels should help the SNRs if this is the case. I don’t have an attenuator to test my setup, so for now it is just speculation.

This topic was automatically closed after 100 days. New replies are no longer allowed.