TRX-2 Basic Performance Testing

Status
Not open for further replies.

Project25_MASTR

Millennial Graying OBT Guy
Joined
Jun 16, 2013
Messages
4,164
Location
Texas
Had a little free time today so I began playing with my TRX-2 on a service monitor. I really just have my scanner to play with and don't really depend on listening to anything with it. Anyway, did some basic tests such as distortion, 12 dB SINAD, 5% BER (P25) and adjacent channel rejection. Tested frequencies were 40 MHz, 150 MHz, 450 MHz, and 850 MHz. I didn't have a proper isolation transformer available for an 8 ohm load so I popped the cover off and pulled a ground referenced voltage off the positive speaker lead (after first checking to make sure it was ground referenced). Due the setup, I'm going to ignore the 7% distortion measurement taken at -50 dBm on all tested frequencies. I also left the default squelch settings in place and tested with a 1000 Hz tone at 1.5 kHz deviation for analog (all FM) measurements. Completely spaced while I was performing the test that these are AM capable as well. P25 testing was rough estimation tuned to "like" audio via my uncalibrated ear using a 1011 Hz tone patter and compared to recovered audio of a XTS5000 at it's 5% BER level. All measurements were made with a R2670 service monitor calibrated October 2017.

At 40 MHz, 12 dB SINAD (FM) occurred at -117.2 dBm (.30 µV). 5% BER is around -113 dBm.
At 150 MHz, 12 dB SINAD (FM) occurred at -111.5 dBm (.59 µV). 5% BER is around -107 dBm.
At 450 MHz, 12 dB SINAD (FM) occurred at -110.2 dBm (.69 µV). 5% BER is around -107 dBm.
At 850 MHz, 12 dB SINAD (FM) occurred at -108.7 dBm (.83 µV). 5% BER is around -107 dBm.

For all bands, a 30 dB adjacent channel rejection was measured (note, narrowband).The test was performed using a R8000B (calibrated January 2018) to generate a -50 dBm signal 12.5 kHz above the base frequency and coupled to the R2670 and scanner using a T connector. The difference between the injected adjacent frequency and new 12 dB SINAD point on frequency was used to determine this measurement.

It should also be noted, it wasn't until after I had put everything away that I realized the squelch setting could be adjusted (sorry, used to testing Motorola radios and never having to mess with squelch). As a result when 5 kHz deviation was attempted, the radio would not typically unsquelch below -113 dBm (while sensitivity was looking promising).

I'll likely add to his at a later point.
 
Last edited:

BrianG61UK

Member
Joined
Nov 13, 2016
Messages
355
Location
England
What about performance on weak signals while stronger signals are on nearby frequencies.

That's the only thing I've ever thought might be really poor on the Whistler scanners.

Try receiving a weak signal on 450MHz when there is a strong one on 470MHz or something.
 

KevinC

Big Dog...celebrating 10 years of abuse!
Super Moderator
Joined
Jan 7, 2001
Messages
11,340
Location
Home
Had a little free time today so I began playing with my TRX-2 on a service monitor. I really just have my scanner to play with and don't really depend on listening to anything with it. Anyway, did some basic tests such as distortion, 12 dB SINAD, 5% BER (P25) and adjacent channel rejection. Tested frequencies were 40 MHz, 150 MHz, 450 MHz, and 850 MHz. I didn't have a proper isolation transformer available for an 8 ohm load so I popped the cover off and pulled a ground referenced voltage off the positive speaker lead (after first checking to make sure it was ground referenced). Due the setup, I'm going to ignore the 7% distortion measurement taken at -50 dBm on all tested frequencies. I also left the default squelch settings in place and tested with a 1000 Hz tone at 1.5 kHz deviation for analog (all FM) measurements. Completely spaced while I was performing the test that these are AM capable as well. P25 testing was rough estimation tuned to "like" audio via my uncalibrated ear using a 1011 Hz tone patter and compared to recovered audio of a XTS5000 at it's 5% BER level. All measurements were made with a R2670 service monitor calibrated October 2017.

At 40 MHz, 12 dB SINAD (FM) occurred at -117.2 dBm (.30 µV). 5% BER is around -113 dBm.
At 150 MHz, 12 dB SINAD (FM) occurred at -111.5 dBm (.59 µV). 5% BER is around -107 dBm.
At 450 MHz, 12 dB SINAD (FM) occurred at -110.2 dBm (.69 µV). 5% BER is around -107 dBm.
At 850 MHz, 12 dB SINAD (FM) occurred at -108.7 dBm (.83 µV). 5% BER is around -107 dBm.

For all bands, a 30 dB adjacent channel rejection was measured (note, narrowband).The test was performed using a R8000B (calibrated January 2018) to generate a -50 dBm signal 12.5 kHz above the base frequency and coupled to the R2670 and scanner using a T connector. The difference between the injected adjacent frequency and new 12 dB SINAD point on frequency was used to determine this measurement.

It should also be noted, it wasn't until after I had put everything away that I realized the squelch setting could be adjusted (sorry, used to testing Motorola radios and never having to mess with squelch). As a result when 5 kHz deviation was attempted, the radio would not typically unsquelch below -113 dBm (while sensitivity was looking promising).

I'll likely add to his at a later point.

Did you account for your test cable loss? I'd think 12 dB SINAD should be a little better.
 

Project25_MASTR

Millennial Graying OBT Guy
Joined
Jun 16, 2013
Messages
4,164
Location
Texas
What about performance on weak signals while stronger signals are on nearby frequencies.

That's the only thing I've ever thought might be really poor on the Whistler scanners.

Try receiving a weak signal on 450MHz when there is a strong one on 470MHz or something.
That's what the adjacent channel rejection test shows. In this case it is 30 dB or in layman terms, a signal 12.5 kHz away will begin to desense the receiver when it is 30 dB (1000 times) stronger. Most LMR radios have between 60 dB and 70 dB of adjacent channel rejection. Some repeaters have 100+ dB of adjacent channel rejection.
Did you account for your test cable loss? I'd think 12 dB SINAD should be a little better.

Test cable had 0.8 dB of loss at 850 MHz. Now, the radios did appear to be doing pretty good when deviation was changed form 1.5 kHz to 3 kHz but me forgetting about adjusting the squelch ended those measurements at -113 dBm on everything above low band. Now, assume the fabled 6 dB margin for radios without narrowband filtering and that would put 12 dB SINAD with 3 kHz deviation in the locations that would be more typical of an average receiver and close to Whistler's specifications (which also specify 3 kHz deviation).
 

milcom_chaser

Member
Joined
Jul 4, 2010
Messages
979
Thanks for posting your results with that radio.
Would be interesting to see the AM Mode results in the Civilian 118-136, and Military 225-400Mhz band.
 

BrianG61UK

Member
Joined
Nov 13, 2016
Messages
355
Location
England
That's what the adjacent channel rejection test shows. In this case it is 30 dB or in layman terms, a signal 12.5 kHz away will begin to desense the receiver when it is 30 dB (1000 times) stronger. Most LMR radios have between 60 dB and 70 dB of adjacent channel rejection. Some repeaters have 100+ dB of adjacent channel rejection.

You think the figure would be no better with a wider separation?

30dB would be appallingly bad for the 20MHz separation example I gave, at least for the case of the wanted signal being the weakest receivable signal.

Also in reality the 30dB figure (a ratio) will change and the strength of the two signals changes.
 

Ubbe

Member
Joined
Sep 8, 2006
Messages
8,944
Location
Stockholm, Sweden
As a result when 5 kHz deviation was attempted, the radio would not typically unsquelch below -113 dBm (while sensitivity was looking promising).

This is with a clean sinus tone. Then concider a digital modulation using the full bandwidth with a complex signal creating much more high frequency audio components that feeds the squelch detector. It's easy to see why multimode analog/digital scanners sometimes skips over digital channels even when there are ongoing conversations.

Setting the analog squelch to a usable level will need it to have a much stronger signal when receiving a digital modulated signal, as some users have aired in the forums. As there are different squelch levels for FM and NFM in a scanner there could also be a use for an offset value for digital channels. Of course service mode should be open for access to users to fine tune all parameters in a scanner.

/Ubbe
 

Project25_MASTR

Millennial Graying OBT Guy
Joined
Jun 16, 2013
Messages
4,164
Location
Texas
You think the figure would be no better with a wider separation?

30dB would be appallingly bad for the 20MHz separation example I gave, at least for the case of the wanted signal being the weakest receivable signal.

Also in reality the 30dB figure (a ratio) will change and the strength of the two signals changes.

It may be however, adjacent channel rejection is just the standard that's used.

I was able to build an adapter to go from the speaker load I have for doing these tests with Motorola's higher tier radios to a 3.5 mm phono plug used by these radios so I can use my existing isolator. So I will likely re-perform the tests, adjust my squelch settings, and try separating adjacent channel by 20 MHz.
 

Project25_MASTR

Millennial Graying OBT Guy
Joined
Jun 16, 2013
Messages
4,164
Location
Texas
Re-Tested

Went back and re-tested.

Measured distortion at -50 dBm was 8%.

Squelch was turned down to minimum levels for testing. All AM measurements were made at 80% modulation.

27.5 MHz AM - 12 dB SINAD @ -112.0 dBm.
40 MHz FM - 12 dB SINAD @ -121.8 dBm
120 MHz AM - 12 dB SINAD @ -110.0 dBm
150 MHz - 12 dB SINAD @ -111.8 dBm for NFM, -112.8 dBm for WFM, 5% BER @ -110 dBm
250 MHz AM - 12 dB SINAD @ -111.2 dBm
380 MHz - 12 dB SINAD @ -108.9 dBm for NFM, -111.4 dBm for WFM, 5% BER @ -110 dBm
450 MHz - 12 dB SINAD @ -110.7 dBm for NFM, -115.3 dBm for WFM, 5% BER @ -110 dBm
850 MHz - 12 dB SINAD @ -109 dBm for NFM, -110.3 dBm for WFM, 5% BER @ -110 dBm

Changed how I measured adjacent channel rejection. Previously the R8000 was used to generate an unmodulated FM signal into a T adapter. In this round, a 1 kHz tone at 1.5 kHz deviation was used in conjuction with an ISO-T with a known (and remeasured) 40 dB of attenuation on the isolated leg (couldn't find the T used the previous day).

Measured adjacent channel rejection at 31 dB. Adjacent channel was also moved to 470 MHz which resulted in a rejection measurement of 58 dB.
 

Ubbe

Member
Joined
Sep 8, 2006
Messages
8,944
Location
Stockholm, Sweden
Try setting the squelch when it is scanning live on an antenna. I guess that will be the 9 or 10 setting and then measure when the squelch opens. That will be the real in life value of what it is capable of receiving and would be more useful to scanner users.

Then do the same with another scanner brand, maybe a Uniden, for comparison.

/Ubbe
 

Project25_MASTR

Millennial Graying OBT Guy
Joined
Jun 16, 2013
Messages
4,164
Location
Texas
Try setting the squelch when it is scanning live on an antenna. I guess that will be the 9 or 10 setting and then measure when the squelch opens. That will be the real in life value of what it is capable of receiving and would be more useful to scanner users.

Then do the same with another scanner brand, maybe a Uniden, for comparison.

/Ubbe

Perhaps if I have time I'll get into it some but I physically don't have anything to compare it to aside from Motorola and Kenwood P25 radios.
 

Ubbe

Member
Joined
Sep 8, 2006
Messages
8,944
Location
Stockholm, Sweden
I did some measurement at 422MHz FM using a Stabilock signalgenerator thru a 1-6 splitter Triax HTS6 with 9dB loss according to its specs.

TRX-2 set at SQ level 10 and test when SQ open with a 1KHz tone and then a 12dB sinad sensitivity test.
-104dBm 1,5KHz modulation
-102dBm 3,5KHz modulation
-101dBm 5KHz modulation
-103dBm 12dBsinad.

BCD536HP squelch set at 5.
-113dBm 1,5KHz
-111dBm 3,5KHz
-109dBm 5KHz
-109dBm 12dB sinad

TRX-2 where -1KHz wrong in frequency
BCD536HP where +1KHz wrong in frequency

If I lower the SQ on the TRX-2 from 10 it starts to open up on various empty frequencies without carriers.
The 536 can be set at SQ 2 and still only open up on valid carriers, very consistant operation.
My TRX-2 have always had worse sensitivity than my 536. Seems mine is at least 3dB worse than MCore25s but his is still 3dB worse than my 536. Remember that a scanner only tries to decode digital data if the squelch opens. I do not have any digital modes in my signalgenerator and 5KHz modulation only goes so far to prove squelch opening on digital signals.

With 5KHz modulation the TRX-2 needs a 8dB stronger signal than 536 to open squelch.
For real live antenna values add 9dB to the dBm values.

/Ubbe
 
Status
Not open for further replies.
Top