12 dB SINAD vs 5% BER

General P25 discussion
Post Reply
User avatar
Project25
Wiki Admin
Posts: 244
Joined: Fri Jan 06, 2017 4:30 am
Location: Somewhere where RF is present

12 dB SINAD vs 5% BER

Post by Project25 » Sun Jul 08, 2018 10:13 pm

Performed an experiment about six months ago and figured I'd finally stop sitting on the audio. This test compares the audio quality of 5% BER versus 12 dB SINAD (i.e. the points at which we spec analog and digital receiver sensitivity to) as it is reproduced by a 7/800 MHz Motorola XTS5000 with firmware release R20.50.10 on 8CALL90 (851.0125 MHz). All audio was generated by a calibrated Motorola R2670A service monitor.

The 7/800 MHz XTS5000 was chosen for several reasons:

1. It was the only easily accessible Astro 25 subscriber I had available with current firmware at the time.
2. I tend to use Motorola as the standard to which I compare everything else to.
3. I did not have access to an APX subscriber at the time.
4. The R2670A I have will support Astro (VSLEP) or P25 with it's IMBE vocoder.
5. I hadn't honestly seen my XPR6550 in almost a year at that point.
6. Unlike most APX and XPR SU's I've tested, the Astro 25 line typically hits 5% BER within 1 dB of 12 dB SINAD (not that it really matters for this experiment).
7. This one had been aligned several weeks prior.

The R2670 was chosen for several reasons:

1. The P25 encode/decode functionality of the Freedom R8000B had recently been released and my company had not purchased the upgrade yet.
2. The R8000/8100 does not have encode/decode capability for NXDN and DMR, just store and playback which made it impractical to obtain consistent base audio recordings for comparison. Maybe we'll see that in the R9000 (*cough* Doyle Wofford @ Freedom CTE *cough*).
3. I own the R2670A and my employer owns the R8000B.

P25 was chosen as the digital medium to test for several reasons.

1. See above equipment that I had access to.
2. Since FirstNet squashed the development for Phase 3 (which was focused on mobile data) and Phase 2 (TDMA) operation only applies to trunked systems, C4FM will remain the modulation method for conventional P25 operations.

A baseline recording was made using a Sure SM58 microphone, recorded at a sample rate of 192 kbps and stored as a WAV file (i.e. this is about as close as I can get to broadcast quality audio). The file was then replayed and fed into the External Modulation input of the R2670 and done so for all tests. All reproduced audio recordings were taken by directly mic-ing the speaker of the XTS5000 with the SM58. The baseline recording can be heard here.

Just as another reference point, the test was run on analog at a level that would be easily considered full quieting (~-50 dBm) at 4 kHz max deviation. Note, the frequency response has clearly changed due to the ~300Hz to 3125 Hz pass band in the radio.

Following the service manual procedure for performing 12 dB SINAD using an RLN64xx test set, 12 dB SINAD was found to occur at -120.8 dBm at 851.0125 MHz. The meter output was cabled directly to the R2670 and not an external SINAD meter. The baseline audio was then fed into the external modulation input and recorded at the 12 dB SINAD signal level (again, 4 kHz max deviation).

Finally, using Astro 25 tuner and a O.153 test pattern, 5% BER was found to occur at -121.2 dBm at 851.0125 MHz. The baseline audio was encoded via the external modulation input at 2.83 kHz max deviation (default) with C4FM modulation and the resulting recording provided.

In conclusion, it's all personal preference. I can't stand noise and would rather listen to bit error. This is simply a comparison that can help bridge the gap between some of the common misconceptions about digital radios.
User avatar
NavyBOFH
CTF Mod
Posts: 214
Joined: Wed Jan 11, 2017 3:56 pm
Location: USA

Re: 12 dB SINAD vs 5% BER

Post by NavyBOFH » Mon Jul 09, 2018 6:46 pm

I have always liked digital modes for exactly what you illustrated. With permission, I might include these clips and test data in my next presentation to a CERT group that is adamant against going digital for the sake of the old-timers not wanting to mess with "those new modes only cops use". It may be personal preference, but I was able to essentially make out a full copy of your transmission right at 5% BER while analog sounded like static with a hint of voice.

Another example which I might record to add shows in our area the ability for digital modes to squash out other analog sources. We have a VHF itinerant repeater up on a mountain, around 920' between mountain and tower elevation. On analog the repeater just washes out within a few miles due to other itinerant users in the towns below. Digital, however, can be heard for a good 17-20 mile radius... on an XTS5000! With a mobile (even my Astro Spectra+ without the high sensitivity RX mod), I was able to reliably talk on the repeater up to 48 miles "as the crow flies" from the repeater site. Analog? Can't even key up the repeater from that distance.
Editable Google map for P25 systems. Google P25 Map
To join the "CTF Team" and help support the forum, click "Donate" below!
[paypal]Donate[/paypal]
User avatar
Project25
Wiki Admin
Posts: 244
Joined: Fri Jan 06, 2017 4:30 am
Location: Somewhere where RF is present

Re: 12 dB SINAD vs 5% BER

Post by Project25 » Mon Jul 09, 2018 9:00 pm

NavyBOFH wrote:
Mon Jul 09, 2018 6:46 pm
I have always liked digital modes for exactly what you illustrated. With permission, I might include these clips and test data in my next presentation to a CERT group that is adamant against going digital for the sake of the old-timers not wanting to mess with "those new modes only cops use". It may be personal preference, but I was able to essentially make out a full copy of your transmission right at 5% BER while analog sounded like static with a hint of voice.

Another example which I might record to add shows in our area the ability for digital modes to squash out other analog sources. We have a VHF itinerant repeater up on a mountain, around 920' between mountain and tower elevation. On analog the repeater just washes out within a few miles due to other itinerant users in the towns below. Digital, however, can be heard for a good 17-20 mile radius... on an XTS5000! With a mobile (even my Astro Spectra+ without the high sensitivity RX mod), I was able to reliably talk on the repeater up to 48 miles "as the crow flies" from the repeater site. Analog? Can't even key up the repeater from that distance.
Permission granted. In fact, anyone who wants to use the info may freely do so as long as credit is to this site.
Post Reply