OP25 Nooelec Smart XTR not decoding.

[MCWKen]

I purchased a new Nooelec Smart XTR dongle, and for some reason, it fails to decode the control channel. Under GQRX, I do get a sufficient signal, and even under OP25, the FFT plot also shows a good signal. Attached will be the logs and plots from both the old Noolec and the new XTR. Perhaps I am missing something.

First the command line: ./rx.py --args 'rtl' --gains 'lna:49' -S 2560000 -T trunkmusc.tsv -2 -w -X -U -v 5 -l http:127.0.0.1:8080 2> stderr.2

Old plots:
New plots:

In both instances, the same USB port and antenna was used. Thanks in advance.

Ken Schmidt

 

[boatbod]

Based on the FFT plot it looks like your PPM might be off. What do the raw and tuned mixer plots show for each dongle?

 

[MCWKen]

First the old:

Now the new:

Hope this helps.
Ken

 

[boatbod]

Mixer signal does not look so good.
I did see an oddity in the "new" log where LNA gain is being set to 99.

gain: name: LNA range: start -1 stop 42 step 2
gain: name: IF range: start 3 stop 56 step 1
setting gain lna to 99

Not sure what's up with that, but you'll see from the line above that max gain is 42, so try lowering the gain in your config.

 

[MCWKen]

Setting gain to --gains 'lna:42' Here are the resultinig FFT plots; new: and old:.

 

[boatbod]

The new dongle plot still looks off-centered and the signal level is considerably lower than the old dongle.
I'm not sure what else to suggest - maybe play with the sample rate to see if that makes any difference. Usually op25 only needs a sample rate of 1Mhz to work well.

 

[wgbecks]

Setting gain to --gains 'lna:42' Here are the resultinig FFT plots; new:View attachment 169356 and old:View attachment 169357.

It appears from Post # 5 that you don't have sufficient signal to noise (SNR) to support good decoding. The signal level seen in the "OLD"
FFT plot is marginal. The noise floor seen in the "NEW" FFT plot is higher than the "OLD" while the signal level is even lower.

Are you (switching) using the same antenna and antenna placement between the two devices when testing and capturing these plots?
Mixer and Constellation plots of the "NEW" would be most helpful to determine frequency accuracy (PPM) and to evaluate the SNR.

Although it's been a number of years since running comparative tests, I want to recall that SDR's using the E4000 tuner did not perform
nearly as well on 7/800 MHz as those with the R820T2/R860 tuner. Additionally, I've encountered one out of a every so many identical SDR's that have a worse SNR performance that may have been attributable to noisy LNA or mixer component(s).

 

[wgbecks]

Please disregard my question about use of the same antenna for your comparisons. I hadn't caught that you indeed were using the same
USB port and antenna to perform your tests. My general impression is that the new XTR isn't performing nearly as well as your old SDR does.

 

[MCWKen]

Please disregard my question about use of the same antenna for your comparisons. I hadn't caught that you indeed were using the same
USB port and antenna to perform your tests. My general impression is that the new XTR isn't performing nearly as well as your old SDR does.

AI have been doing some comparisons in GQRX between the two. I had noted as you said, there is a difference in the SNR (last time I even thought about that was in 1975, playing with shortwave antennas and feeds). Reading you comment made me go back and look up how to figure SNR. I am going back to the original Nooelec device.

Thanks to you both, Ken.

 

[wgbecks]

Ken,

I wish that I still had an XTR's (E4000 tuner) to test for noise figure and etc. I've had very good success with the SMArt's but have observed
one or two of them out of many owned by myself and friends that have had higher internal noise levels that diminish SNR, especially up in
the 7/800 MHz bands. It's no surprise that device quality can vary from batch to batch when you consider the low cost of these devices.

I realize the E4000 has a higher tuning range and whereby all other specification appears to be equal with the R820T2/R860 devices, yet I am
almost certain that I didn't particularly care for them unless I had plenty of site signal in reserve because of higher internal noise temperature.

Bill

 

[MCWKen]

Ken,

I wish that I still had an XTR's (E4000 tuner) to test for noise figure and etc. I've had very good success with the SMArt's but have observed
one or two of them out of many owned by myself and friends that have had higher internal noise levels that diminish SNR, especially up in
the 7/800 MHz bands. It's no surprise that device quality can vary from batch to batch when you consider the low cost of these devices.

I realize the E4000 has a higher tuning range and whereby all other specification appears to be equal with the R820T2/R860 devices, yet I am
almost certain that I didn't particularly care for them unless I had plenty of site signal in reserve because of higher internal noise temperature.

Bill

Although I doing the testing on the North Scott site, 20 miles away, It also did not funtion on the local site 2 miles away. Signal strength was there, but it did not decode.

 

[wgbecks]

Ken,

It's very likely the NAC of the local site is different from the more distant or adjacent site. Try editing your trunkmusc.tsv file and set
the NAC field to 0x0. Doing so in boatbod op25 will allow the trunking logic to accept the first NAC decoded.

Bill

 

[boatbod]

That big dip in the signal at the center of the tuned signal is very curious; kinda like the inverse of a DC spike.
In op25 try configuring an offset to get that out of the way. (rx.py command line parameter "-o 15000")