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

[Boatanchor]

This thread is like fly paper! Anyone else?

OK! Anyway, Boatanchor - will you be testing the 396 (or are you)? Curious to your results, very curious!

It's coming

Too many other jobs to do & not enough hours in the day.

Just have to transfer my scribbles into a spreadsheet now.

 

[joeuser]

It's coming

Too many other jobs to do & not enough hours in the day.

Just have to transfer my scribbles into a spreadsheet now.

Understandable. Appreciate your work in this!

 

[Boatanchor]

Updated figures..

Firstly an apology..

When I ran the initial Sinad tests last week, I was using a line isolation transformer between the headphone jack on the scanners and the audio input on the service monitor. As it turns out, the isolation transformer must have been introducing a little audio distortion into the equation which degraded the Sinad readings by a couple of dB in all cases. This was possibly due to some audio impedance mismatch between the scanner's high impedance headphone amplifier output circuitry and the transformer.

Anyway, I have re-run the tests using a direct connection between the scanners audio output and the service monitor. The difference between the two scanners sensitivity performance on the various band remains the same, however, it should be pointed out that the 2dB or so increase in sensitivity figures now places my BCD436HP at or better than the specified sensitivity levels on more bands than before.

There still appears to be some question about sensitivity on the bands above 490Mhz, but as others have pointed out, whats a few dB between friends?

At this point, I am probably more concerned about the BCD436HP squelch operation.

Looking at this chart, it just goes to show how hot the receiver really is in the BCD396XT.

 

[KevinC]

Firstly an apology..

When I ran the initial Sinad tests last week, I was using a line isolation transformer between the headphone jack on the scanners and the audio input on the service monitor. As it turns out, the isolation transformer must have been introducing a little audio distortion into the equation which degraded the Sinad readings by a couple of dB in all cases. This was possibly due to some audio impedance mismatch between the scanner's high impedance headphone amplifier output circuitry and the transformer.

Anyway, I have re-run the tests using a direct connection between the scanners audio output and the service monitor. The difference between the two scanners sensitivity performance on the various band remains the same, however, it should be pointed out that the 2dB or so increase in sensitivity figures now places my BCD436HP at or better than the specified sensitivity levels on more bands than before.

There still appears to be some question about sensitivity on the bands above 490Mhz, but as others have pointed out, whats a few dB between friends?

At this point, I am probably more concerned about the BCD436HP squelch operation.

Looking at this chart, it just goes to show how hot the receiver really is in the BCD396XT.

Are you accounting for test cable loss at the higher frequencies?

 

[Boatanchor]

I'm running 2ft of RG-400 coax.

According to the coax loss calculator at:
Welcome to Times Microwave | Coaxial Cable - Attenuation & Power Handling Calculator

The coax/connector loss would be about 0.3dB at 800Mhz.
At 500Mhz, the coax loss drops back to 0.2dB

Feel free to add that loss figure to the tables, but it won't make much difference in the grand scheme of things. The production and component variations in the scanners themselves would overshadow such small coax patch lead losses.

 

[KevinC]

I'm running 2ft of RG-400 coax.

According to the coax loss calculator at:
Welcome to Times Microwave | Coaxial Cable - Attenuation & Power Handling Calculator

The coax/connector loss would be about 0.3dB at 800Mhz.
At 500Mhz, the coax loss drops back to 0.2dB

Feel free to add that loss figure to the tables, but it won't make much difference in the grand scheme of things. The production and component variations in the scanners themselves would overshadow such small coax patch lead losses.

Ok, just checking...you never know.

 

[joeuser]

Firstly an apology..

When I ran the initial Sinad tests last week, I was using a line isolation transformer between the headphone jack on the scanners and the audio input on the service monitor. As it turns out, the isolation transformer must have been introducing a little audio distortion into the equation which degraded the Sinad readings by a couple of dB in all cases. This was possibly due to some audio impedance mismatch between the scanner's high impedance headphone amplifier output circuitry and the transformer.

Anyway, I have re-run the tests using a direct connection between the scanners audio output and the service monitor. The difference between the two scanners sensitivity performance on the various band remains the same, however, it should be pointed out that the 2dB or so increase in sensitivity figures now places my BCD436HP at or better than the specified sensitivity levels on more bands than before.

There still appears to be some question about sensitivity on the bands above 490Mhz, but as others have pointed out, whats a few dB between friends?

At this point, I am probably more concerned about the BCD436HP squelch operation.

Looking at this chart, it just goes to show how hot the receiver really is in the BCD396XT.

Flagship... Well this is revealing! Its disappointing... I appreciate you doing these tests, perfect or not - for a layperson such as myself the point is made.

Out of curiosity, and I won't quote anyone on it, but is it reasonable to assume the 396XT would be quite similar to the 996XT in this regard? Best guess?

 

[zz0468]

Flagship... Well this is revealing! Its disappointing... I appreciate you doing these tests, perfect or not - for a layperson such as myself the point is made.

I wouldn't read too much into those results.

First off, I am NOT doubting the accuracy or veracity of Boatanchor's measurements.

What I am commenting on is just how significant the measured difference is. It's clear that the 436 is inferior on the particular test that the OP made. But taking the worst case measurement at 820 MHz (?), it looks like the spread is -115 dbm for the 436, and -118 dbm for the 396.

That's about 0.4 uv vs 0.28, a spread of about 1/10th of a microvolt.

Now, take any receiver and inject a signal down near the threshold, and vary it + or - 1/10th of a microvolt. In a blind test, you probably wouldn't notice a difference. Watching and expecting a change, you will barely perceive it. I stand by my earlier comments that the actual sensitivity in microvolts is not what's causing the perceived poor performance on the 436.

Addressing the out of spec 0.4 uv sensitivity for a moment, there's a couple of things that need to be considered. One is the calibration of the instrumentation. There's a signal generator and a sinadder somewhere in the mix. There are a lot of complex factors that come into play when determining accuracy, and absolute accuracy down to 1/10th of a microvolt resolution is 'iffy' outside of a metrology lab. Differences in input impedance of the receivers will cause errors between the two receivers, and in overall absolute accuracy. I guarantee that the input impedance to these receivers is NOT anywhere close to 50 ohms resistive. It's a complex impedance that will vary with frequency. Differences in the IF filters will affect the accuracy of the SINAD measurement. And so on.

That's NOT to disparage the measurements in any way. I'm just trying to point out some of the measurement uncertainty that will ALWAYS be there. The closer the tolerances you're looking for, the more those minor details will add up.

BTW, 0.4 uv for 12 db sinad for a 12.5 KHz receiver bandwidth translates to a 6.2 db noise figure. If one is familiar with some of my previous technical rantings about scanner sensitivity, I have mentioned repeatedly that the typical scanner noise figure (the measurement that REALLY counts) is around 6 db. So, in other words, this is comparable to everything else out there, within a pretty close margin.

To the OP, Boatanchor... Please don't misread my comments. Your measurements are good, and quite interesting. I've made it a sideline of my career to understand receiver sensitivity measurements and receiver performance, and even when using metrology grade cables, adapters, and instruments, there is always a margin of error, especially when making power measurements. A receiver sensitivity test is a power measurement, in a very real sense.

So, in the end, I'd call the .4 uv on the 436 a wash, if a bit sloppy on Uniden's part. Usually a manufacturer gives a bit of wiggle room on the specs, and may even issue two specs, one as "guaranteed", and one as "typical".

 

[Boatanchor]

All very true of course, however, factor in another 3 or 4dBm before the dodgy 436HP squelch opens and the situation looks even less rosy.

I know that people will say that signal levels below -115dBm probably don't matter much, but when you are trying to monitor distant signals or weak signals from a relatively inefficient stubby antenna and the squelch refuses to open until signals are 3 or 4dB stronger than the 12dB Sinad point on the band you are listening to, it sure does matter.

 

[zz0468]

All very true of course, however, factor in another 3 or 4dBm before the dodgy 436HP squelch opens and the situation looks even less rosy.

Yes, that effectively reduces the receiver's sensitivity even more.

I know that people will say that signal levels below -115dBm probably don't matter much, but...

The factor in a couple of db for a substandard feedline, a couple more on a cruddy antenna, and a few more for desense from the cell site a mile down the road, and all the little bits add up. Which is why, even though a couple of db here or there may not be noticeable, they all add up when you're chasing the weak ones.

Pile a lousy noisy VCO on top if it all, and you wind up with an expensively irritating new toy. :evil:

Good thread! I'd like to see you do a couple more tests.

I'd be curious what the 20 db quieting sensitivity is on both radios on the same frequencies you previously measured. I'd also like to see a test where you measure the quieting in 10 db steps, to see what level it actually stops quieting. Use the unsquelched audio as your reference level, and then see where 20 db quiet is, and then maybe the quieting level at 5uv and 10 uv, and still higher if you feel like it.

 

[Mike_G_D]

I'd also like to add (gee, Boatanchor don't you like all this extra work we're asking of you?;-)) another test where you go for a relative difference between the 396 and 436 with a 6dB pad on the antenna port of both receivers to better damp out receiver input impedance effects. This is more for seeing what the variation across the band(s) is relative between each receiver with return loss to the generator improved (and maybe some stability improvement in the receiver's RF amp stage - you never know!); better if you can add another pad on the output of the generator also.

zz0468's comments about using an IsoTee to test local noise effects (in this case noise possibly generated from the receiver itself getting back through the antenna in portable use) sound good to me also and I went looking for a good on-line tutorial about that. I found this which I think is pretty well written and gives a good description of what an IsoTee does and how to properly use it: Every toolbox needs one of these | Test and Measurement content from Urgent Communications.

In that case, I would test the receivers with an IsoTee first with a 50 dummy load as the tutorial describes and then with a typical portable antenna connected (like the standard "ducky") and see what the relative difference in noise is. I'd try it also with a known good matched antenna (one tuned for the test frequencies with a good ground plane or matched end-fed half wave, coaxial dipole, etc.) as well.

-Mike

 

[ratboy]

I would like to see this test redone with several units, as I don't think any of the 436's I've used would even remotely come close to the claimed sensitivity. The usual VHF stuff I listen to, MED and PD/FD, is for the most part, unlistenable on the 436. Maybe that's the wrong term, it's pointless to listen to it, as you just don't hear anything at all on about 90% of the VHF freqs programmed in, and the ones you do hear are either hissy (These are pretty close to full quieting on every other radio I have here), or barely there (Squelch is chopping the audio up). Even with the old firmware(s), there is still a problem with VHF deafness compared to any scanner I own, and I still own a lot of scanners. Some of them are pretty much parts radios, and they all outperform any of the 436's I've used here on VHF. And then comparing a 436 to my Yaesu VX-170 is a total joke. On 161.070, the local NS road freq, the VX-170 is talking away, the 436 is nearly silent, and when the squelch breaks, all you hear most of the time, is static. A $500+ plus radio that really can't get VHF is pretty much of a joke. These radios remind me of the old brass HO locomotives that looked great but were really bad runners.

 

[joeuser]

This reminds me that new & improved ain't always improved. Seems like every generation of a device hits a plateau & then the older stuff becomes the best of that generation. Until a new generation comes out.

See, I debated over the 436 v. 396 before I went ahead & got the 436. Now, I wish I had just got the 396 to begin with. This may be the real gem in this line. You guys know what I mean. I do own a 396 & a 996 XT, boy they do really well... I have the 396T also & wish the XT had carried over the look & feel. Off track...

So, I read all posts & will anyone speculate if the 396 finding would most likely apply to the 996? I assume they will but... Just looking for confirmation.

 

[kc5igh]

What now? (And suggestion for Paul and the Uniden team)

Okay. I'm following all this at a bit of a technical distance, especially when compared to the likes of Boatanchor and ZZ0468.

Insignificant as the technical sensitivity measurement differences may be, my 436 (like Joeuser's) just plain performs miserably on VHF and UHF compared to my BCD396T, BCD396XT, PRO-106, PSR-800, and PSR-500. In fact, Boatanchor's chart shows dips in sensitivity in precisely the VHF-hi and 406-420 MHz ranges that constitute the bulk of my local monitoring efforts.

It's gratifying to know that my (and others') 436 performance disparity may be due more to squelch issues than rf receiver sensitivity differences and that modifying the squelch tables for some of those ranges (by means of a firmware update?) might be possible.

Suggestion for Paul Opitz and the Uniden team: If and when you solve the Siren and analyzer problems, many of us who live in weak-signal environments would appreciate a firmware upgrade (if indeed, it's possible) that addresses the squelch issues that currently limit our ability to monitor analog VHF and UHF signals. The people who are currently happy with the performance of their 436's could opt-out of that particular upgrade if it hampers the radio's performance in their environment?

In any event, I'd also like to hear Boatanchor's and ZZ0468's thoughts on the internal noise issue Boatanchor detected, because that might be the ultimate, hardware-driven showstopper for any improvement to the current model of the 436 (not that I intend to ever buy another of these radios without first learning how well it works in the hands of real people).

Thanks for listening!

-Johnnie

 

[Mike_G_D]

In any event, I'd also like to hear Boatanchor's and ZZ0468's thoughts on the internal noise issue Boatanchor detected, ...

This is where the use of the IsoTee to test local noise with a portable antenna might prove relevant. We'd perform the same test on both the 396 and the 436 on the same test frequencies with the same antenna(s) and see what kind of noise might be radiated out of the unit and back into the antenna.

Of course, that's not the only way internally generated noise can get into the receiver chain but since many do see improvements when using an external antenna (in terms of relative apparent weak signal reception between the 436 and the 396) it's worth a look, I think.

-Mike

 

[Boatanchor]

This is where the use of the IsoTee to test local noise with a portable antenna might prove relevant. We'd perform the same test on both the 396 and the 436 on the same test frequencies with the same antenna(s) and see what kind of noise might be radiated out of the unit and back into the antenna.

Of course, that's not the only way internally generated noise can get into the receiver chain but since many do see improvements when using an external antenna (in terms of relative apparent weak signal reception between the 436 and the 396) it's worth a look, I think.

-Mike

I'm afraid the Iso-Tee idea is beyond my testing capability for one simple reason.

To conduct such testing would require the use of a shielded room (Faraday shielded environment).
This is something I do not have access to.

If this test were conducted in the typical RF polluted environment, it would reveal nothing about radiated noise from the scanner and more about radiated noise from other nearby noise sources such as computers, monitors & SMPS etc.

If anyone is really interested, they can use a sensitive spectrum analyzer and place it's resonant VHF antenna near to the LCD screen on the 436HP. They should see a rise in broadband noise which disappears when the scanner is switched off.

As I alluded to before, any radiated noise from the scanner itself is likely to degrade the S/N ratio of weak signals being picked up when using an attached stubby antenna, but I am not in a position to accurately measure this. At this point in time, it is purely speculation on my part.

B

 

[Mike_G_D]

Actually, not so much - this is why I wrote of doing a "relative" test between the 396 and the 436. If you do a local noise test between both and one shows considerably more than the other then you have a relative difference data point.

Here's what I would try...

Make a practical attempt to quiet the local RF environment as much as possible. I would shut down whatever is in the room that may radiate RF that isn't absolutely needed such as the test equipment.

Run the IsoTee test on the 396 and then the 436 and record the results.

Compare the two - if the 436 is really radiating a lot more gunk than the 396 then the relative difference between the two should still be obvious given the same test environment.

If you really want to you could also construct a small Faraday Cage test box - we aren't talking really large here so this should be doable.

Remember, we aren't doing a receiver noise floor test here using a calibrated noise meter we're only doing a relative site noise test wherein the bulk of the "site" noise is, we think, coming from one of the receivers under test.

In any case, I would like to see a re-test of the sensitivity with pads on the receiver antenna inputs if at all possible to negate sig gen return loss effects. May not make any difference but it would be good to check; depending on the design of the RF amp in the 436 it might make the sensitivity variation across the band improve and maybe even the noise due to potential stability issues with reactive source impedances.

-Mike

 

[zz0468]

If the 436 is radiating that much crud, then it's a good possibility that the internal crud is getting into the receiver internally, in which case there is no good easy test to see what's going on. In that eventuality, the receiver sensitivity is whatever it is, degraded by it's internal noise. That would show up in the over all receiver sensitivity and noise figure tests, but there would be no way to break that component out of the test results.

What can be done, however, is the OP's HP service likely has a reasonably sensitive spectrum analyzer. I'd put a probe, maybe a half inch of exposed center conductor on a piece of coax, and probe around the radio while the spectrum analyzer is in peak hold mode. A control run could be made with the radio off to see what the probe picks up off the air.

I'd check all frequencies that the 436 is capable of receiving on.

My uneducated and uninformed guess would be that the radio is cleaner than you might think, with radiated noise at a low level confined to very close proximity to the receiver.

 

[Mike_G_D]

True, hmm, well it had to meet radiated emissions to pass Part 15 so it shouldn't be more than that, of course. But the primary reason I still wonder are the reports of the 436's apparent sensitivity being much closer to if not on par with the 396 when attached to an external antenna. That "seems" to indicate some radiated noise issues which manifest themselves when using a portable antenna - or some other issue stemming from the use of the portable antenna (like some RF amp instability cause by a poor reactive match from the portable antenna). Of course the reports are scant on details and we need to note exactly what antenna is used under what conditions, etc.

It's frustrating to me as I could run all of these tests and more if I had access to the equipment - so many ideas I want to try! And I am very thorough and quite persistant. Oh well.

If it's primarily the synth phase noise (as the main culprit in terms of what is causing the noted issues) then that's not going to show up so much (directly) as a radiated issue, of course. But I'm still curious given some of the reports concerning the 436's performance with external antennas.

Sniffing with a probe and a spec an would quite telling! If such is available then absolutely I would love to see those spectrum plots (run the same on both the 396 as well as the 436)! I would look at the spectrum with the antenna of the receiver unterminated, terminated in a 50 ohm resistive load, and terminated in the stock duck. Just to see if any oscillation/stability issues pop up (I've noted something like that on some of my scanners - self quieters pop up with open termination or poor termination and greatly reduce if not disappear when terminated properly).

-Mike

 

[stingray327]

So then which radio has a better receiver? The 396XT or the BCD 436? What are the differences between these two scanners? Is it the pre-program feature (Home Patrol) on the 436 that makes it different than the 396XT?