BCD436HP/BCD536HP: BCD436HP - The Sensitivity "test" (it had to come out eventually..)

[Boatanchor]

We've heard anecdotal reports about questionable VHF performance on Uniden's 'Flagship' handheld scanner on this forum over the last nine months or so. I thought it was about time we finally nailed some sensitivity figures down and found out what is really going on here.

The following tests were conducted with a calibrated HP service monitor using a high quality RG-400 coax patch lead. The received audio was taken from the headphone socket in each case via the same audio line isolation transformer. All tests were conducted in NFM mode at 1.5Khz deviation. I should mention that my BCD396XT has had the 450Khz IF filter changed to the narrower 'G' version, which should closely mimic the filter characteristics in the BCD436HP in NFM mode. Regardless, the filter change in the BCD396XT would not impact greatly on the observed figures in these tests.

Caveat:

There are obviously slight variations in performance from model to model and even within the same batch of the same model.
The following results may have been impacted slightly by these variations. The results I obtained may be different to results others obtain.
Regardless, I am confident that in my case at least, there is a significant difference in sensitivity between my two units.

There are two fields in this test.
The first field measured sensitivity in dBm, the second in uV.
The dBm measurement is this case is more accurate, however the uV tests have been included so that you can compare the measured figures with Uniden's claimed specifications.

Comments and observations:

Firstly, it was immediately obvious that my BCD436HP does have some serious sensitivity 'deficiencies' on some bands. The 163Mhz, 420Mhz and 820Mhz bands were all 2-3dBm down in sensitivity compared to the older BCD396XT. Looking at the uV figures, my BCD436HP doesn't even meet Uniden's own specified 0.3uV level. This is very poor form IMHO.

During these tests, My BCD436HP only met the specified 0.3uV sensitivity on bands below 146Mhz.
It was interesting that the BCD436HP actually outperformed the BCD396XT on the 53Mhz (6M) band.
Another issue I have confirmed, is that the BCD436HP does in fact radiate some low level, broadband noise on the VHF bands. The noise is radiated from the vicinity of the LCD. I believe that the noise is likely produced by an internal CPU or associated system. This noise appears to degrade VHF reception slightly when using the supplied stubby antenna, or any other handheld antenna and results in weak signals not opening the squelch, even at it's lowest setting of 2..

So there you go.. For what it's worth, my results of sensitivity testing on my two Uniden scanners.

And, while these tests were conducted in NFM mode, sensitivity figures also effect weak signal digital performance too. Yes, the BCD436HP sounds great on P25, but the fact is that if you want optimum weak signal performance, it's pretty hard to beat the humble little BCD396XT.

As always, YMMV, but personally, I am a little disappointed in the RF performance of my BCD436HP.

 

[AA6IO]

Boatanchor
Thank you for the detailed analysis. I am not set up, and by profession, not really qualified to do testing with the detail that you have described. I know you have been quite vocal on the RRDB list in different threads about this issue, sometimes in response to my own posts.
I can only agree with your assessment based on my use of the 436HP, 396XT, PRO-106, and most recently, the WS-1080. For digital decode, the 436HP is a winner in the portable category. But for weaker signals, both analog and digital, I keep going back to the other portable scanners mentioned,
I have gone back and forth between firmware versions for the 436HP, but none are as sensitive as my 396XT. I really hope Uniden can fix this issue, but going now on almost 11 months since the introduction of the new x36HPs, I'm not holding my breath.
I still like the 436HP, but not for its outstanding sensitivity.
Steve AA6IO

 

[sibbley]

Boatanchor, thank you. I don't feel like I'm "nuts" anymore...

 

[jonwienke]

And how many scanners have you tested to determine if you have a defective scanner, rather than generalizing one unit's performance to all 436HP scanners?

 

[Boatanchor]

These tests were done on two recently purchased samples that I happen to own. I make a point of noting that that there may well be some minor variation between units. I do not claim that these results are representative of all units being sold.

I would be happy for others to repeat these tests to validate (or discredit) my results.

I should also point out that the BCD436HP tests were conducted using the latest V1.0.3 firmware.
I personally doubt that the tuning/gain characteristics of the RF front end circuitry is software adjustable. If so, simply changing the firmware would not alter these figures.

 

[jonwienke]

I wonder if it would be possible to organize a meeting of x36 owners to test a larger sample of scanners, and get a better idea of how much individual units vary, and what percentage do/do not meet published specs.

 

[kc5igh]

Dear Boatanchor.

Thank you! Thank you! Thank you!

Your measurements perfectly mirror my non-scientific listening experience with the 436.

There are too many of us having similar problems with this radio to chalk up this sensitivity issue to production variations, and if that were the case, then Uniden needs to do something about its quality-assurance standards/procedures.

Let's see what, if anything, Uniden can do to bring this radio up to its published sensitivity specifications!

-Johnnie

 

[Mike_G_D]

Those results are interesting, particularly in the VHF band, as you say. Besides a wider sample range of units (I realize you are limited here by what you have) I would like to see a more detailed LMR band-by-band set of tests.

For example, if you could do a VHF-Hi band test starting at 150MHz and ending at 174MHz in about 2.5MHz increments. That's about 11 frequency test points across the band (using 175MHz as the upper limit - close enough). So you'd go 150MHz, 152.5MHz, 155MHz, 157.5, 160MHz, 162.5MHz, 165MHz, 167.5MHz, 170MHz, 172.5MHz, and finally 175MHz. It would give us a little more resolution around that interesting "dip" in apparent sensitivity at 163MHz your initial results show.

Then I'd do something similar for the other LMR bands of particular major interest like 30-50MHz, 390-420MHz, 450-512MHz, 849MHz to 869MHz, etc.

Based on what you show so far, it almost looks like a low pass response with a rolloff/upper skirt in the upper half of the VHF-HI band - more resolution in that area might be more telling.

-Mike

 

[joeuser]

Very interesting, thank you for doing this & reporting your findings to the community.

 

[SCPD]

436 vs 396xt

Thank you Boatanchor,See I knew the Xt's received better,I hope Uniden makes this all better one day.Until then theres the older firmware.
I bet every 436/536hp scanner you test will be very close,they are all tuned the same and are the same radio by design.Robots make these things you know,what other tests are needed?
A test by uniden would be interesting and fun though


2-3dBm down in sensitivity,aw man.

 

[jonwienke]

Robots make these things you know,what other tests are needed?

All modern mass-produced electronic devices are "made by robots" (at least the soldering of surface-mount components to the circuit board, and the fabrication of the individual chips, resistors, and capacitors), but that does not mean that there are no unit-to-unit variations.

Also, firmware can affect how a received signal is processed, but a firmware update isn't going to correct a defective RF front end.

 

[Boatanchor]

Those results are interesting, particularly in the VHF band, as you say. Besides a wider sample range of units (I realize you are limited here by what you have) I would like to see a more detailed LMR band-by-band set of tests.

For example, if you could do a VHF-Hi band test starting at 150MHz and ending at 174MHz in about 2.5MHz increments. That's about 11 frequency test points across the band (using 175MHz as the upper limit - close enough). So you'd go 150MHz, 152.5MHz, 155MHz, 157.5, 160MHz, 162.5MHz, 165MHz, 167.5MHz, 170MHz, 172.5MHz, and finally 175MHz. It would give us a little more resolution around that interesting "dip" in apparent sensitivity at 163MHz your initial results show.

Then I'd do something similar for the other LMR bands of particular major interest like 30-50MHz, 390-420MHz, 450-512MHz, 849MHz to 869MHz, etc.

Based on what you show so far, it almost looks like a low pass response with a rolloff/upper skirt in the upper half of the VHF-HI band - more resolution in that area might be more telling.

-Mike

Good point Mike.

When I get time I will do a more detailed sweep of the VHF and maybe UHF bands.

 

[kc2kth]

Interesting this got posted today. After using my 436 exclusively for a couple of months now last weekend I brought out my pro-106 again. I thought the 106 seemed to have better receive quality, but wasn't sure and didn't want to say definitively without some additional testing. I received in the mail today a BNC Watson W-801 - I already have the SMA version on the 436.

I picked a weak weather station and compared the two. On the 106 with the Watson I can understand every word of the transmission, there is a fair amount of noise and I have a signal strength of 2.5 - 3 bars on the S meter. For reference this is about a 1.5 Bar improvement over the stock antenna.

On the 436 the same station is barely audible over the noise using the similar Watson antenna and there is no reception using the factory antenna. With the Watson the S meter shows one bar.

I did one other test. On my old BC296d the same station comes in somewhere between the two - better than the 436, not quite as solid as the 106. Either the factory antenna (for the 296) or the Watson performed about the same here, which I suppose says something for the factory stock that shipped with the 296.

I am seeing similar results as the rest of the folks here. The 436 just doesn't have the same ability to pull in those weak signals. I am curious if this is something Uniden could address in software in a future release. Interesting data either way.

 

[Boatanchor]

Presented with little comment, except to say that my BCD396XT typically exhibits sensitivity of greater than -118dBm on all of these bands..

There are some troubling characteristics to these plots..

 

[joeuser]

Are you sure this is right? Seems terrible!

 

[SCPD]

436

And what firmware is on the scanner you tested this on?

 

[sibbley]

Wow. Look at that "T" band. No wonder I have trouble in my area. Northampton County system is all "T" band 500-503mhz.

 

[AA6IO]

Boatanchor
Not a really pretty plot for the 436HP. Wonder if you could run one for the 396XT and show us the difference.
Steve AA6IO

 

[Boatanchor]

Are you sure this is right? Seems terrible!

Yes, some of those numbers are surprisingly bad.

I would love for someone to come on this forum and prove me wrong with better figures from another sample

 

[Boatanchor]

Boatanchor
Not a really pretty plot for the 436HP. Wonder if you could run one for the 396XT and show us the difference.
Steve AA6IO

Will run the same sweep on the 396XT in the next day or so - stay tuned