The official "I want LSM to work properly in my scanner" thread

[iMONITOR]

Maybe they know something we don't. By the time they invest time and money into resolving the problem, it very possible everything will be encrypted, or go to one of these new WiFi type private networks like AT&T is rolling out (forget what they call it), and it will be game over for them, and us.

 

[KA1RBI]

Any idea what information the amplitude component conveys? Is it strictly for synchronization of the data in the phase modulation? That might explain why an FM demod sort of works up.

See US Patent 6,061,574. I can confirm that this method is in use on LSM via direct observation of a local /\/\ LSM system.

73

Max

 

[UPMan]

Would the mfgs have to abandon all their circuit designs and start over?

Pretty much, yes.

 

[KA1RBI]

Pretty much, yes.

haha - the RF front end, tuner, microprocessor and memory chips (etc)., xMBE speech codec chip, audio amp, speaker, battery charging circuit, USB interface circuit, even including the illuminating lamps all MUST be thrown away and redesigned from scratch in order for Uniden and Whistler to produce a proper LSM receiver. The cost to license this advanced technology (known as PSK) would by itself add HUNDREDS of dollars per unit, none of which could be recouped by Uniden or Whistler as all of it would be passed through to the patent consortium that owns this (passe by the 1960's) intellectual property...

73

Max

 

[KA1RBI]

After setting up and running a Raspberry Pi with an SDR dongle and OP25, it's been proven to me that the solution to LSM is having the right demodulator for CQPSK.

Many thanks for this report.

OP25 using the CQPSK demodulator and an omni-directional antenna gives clear decoding on two local simulcast systems, better than my 436hp hooked to a directional antenna.

One of the ironies in all of this is that there _is_ a circuit point inside the scanner at which a "clean" copy of the LSM digital signal exists - I've tested this on a RS PRO-106 by tapping the final IF (at 455 KHz) and sampling the signal using an SDR. The SDR receiver that was used for this test (unlike the common RTL) can tune to the 455 KHz frequency. OP25 was used at the software receiver. Anyone with the proper equipment and know-how can duplicate this.

It is concluded from this test that most of the scanner's circuits are usable for LSM without modification. The 455 KHz ceramic filter is going to cause some group delay distortion and the scanner's AGC circuits were certainly not optimized for the LSM waveform.

The Pi setup doesn't have all the features of a scanner, but for roughly $70 worth of equipment (not counting antenna), it does pretty well.

Have you tried the new OP25 HTTP web console yet?

73

Max

p.s. I am firmly in the camp of not purchasing any new scanner unless LSM is fixed

 

[fredva]

Have you tried the new OP25 HTTP web console yet?

I noticed some postings about that in the SDR forum thread but I haven't tried it yet. I'll take a look when I get a chance.

Many thanks for the work of the OP25 developers.

p.s. I am firmly in the camp of not purchasing any new scanner unless LSM is fixed

Yeah, I have no interest in another scanner without good LSM decode either.

 

[UPMan]

haha - the RF front end, tuner, microprocessor and memory chips (etc)., xMBE speech codec chip, audio amp, speaker, battery charging circuit, USB interface circuit, even including the illuminating lamps all MUST be thrown away and redesigned from scratch in order for Uniden and Whistler to produce a proper LSM receiver. The cost to license this advanced technology (known as PSK) would by itself add HUNDREDS of dollars per unit, none of which could be recouped by Uniden or Whistler as all of it would be passed through to the patent consortium that owns this (passe by the 1960's) intellectual property...

73

Max

I caught that this was mostly tongue in cheek. Others might not. Not quite that drastic, but still substantial. Coupled with that is the challenge that some very basic radio things are done completely differently when using an I/Q type receiver (things like AGC, Squelch, FM and AM demodulation, and more). It would be a bigger technological change than the move from crystals to PLL, affecting many more functions and systems.

Not saying that it would never be done, but, it would be a very big undertaking.

 

[cpetraglia]

I caught that this was mostly tongue in cheek. Others might not. Not quite that drastic, but still substantial. Coupled with that is the challenge that some very basic radio things are done completely differently when using an I/Q type receiver (things like AGC, Squelch, FM and AM demodulation, and more). It would be a bigger technological change than the move from crystals to PLL, affecting many more functions and systems.

Not saying that it would never be done, but, it would be a very big undertaking.

That pretty much answers everyone's questions, now doesn't it.

 

[RFI-EMI-GUY]

Many thanks for this report.



One of the ironies in all of this is that there _is_ a circuit point inside the scanner at which a "clean" copy of the LSM digital signal exists - I've tested this on a RS PRO-106 by tapping the final IF (at 455 KHz) and sampling the signal using an SDR. The SDR receiver that was used for this test (unlike the common RTL) can tune to the 455 KHz frequency. OP25 was used at the software receiver. Anyone with the proper equipment and know-how can duplicate this.

It is concluded from this test that most of the scanner's circuits are usable for LSM without modification. The 455 KHz ceramic filter is going to cause some group delay distortion and the scanner's AGC circuits were certainly not optimized for the LSM waveform.



Have you tried the new OP25 HTTP web console yet?

73

Max

p.s. I am firmly in the camp of not purchasing any new scanner unless LSM is fixed

Yes!

Some resourceful experimenter could probably design/build a drop in board that has an I/Q demodulator at the IF frequency used by Uniden (or whistler) and detects and converts the CQPSK to a data stream and outputs an equivalent C4FM waveform to feed into the discriminator tap. All that would be needed is to break the connection between the final IF stage and the discriminator.

There are chipsets already made for these waveforms.

The radio should perform as otherwise designed, scanning and receiving etc, since the whole process normally begins at the discriminator.

Regarding comments from UPMAN (welcome back!), AGC and bandwidth might need to be tweaked. But only for the narrow FM/p25 signal path. There are no shortage of engineers with that sort of expertise in the hobbyist realm.

There is a company called Sherwood who made their money selling upgrades to popular ICOM receivers. The effort and market are no different in this case.

I would buy one for my BCD536HP if I could. Hey, maybe Uniden can take it on as a side project and make money upgrading their installed base?

Sent from my SM-T350 using Tapatalk

 

[marcotor]

That pretty much answers everyone's questions, now doesn't it.

You're kidding right? It wont be long before some "expert" who knows more about manufacturing R&D than Whistler and Uniden combined will say this is a "trivial" undertaking; the scanner manufacturers wont do it because they are lazy, incompetent, or some other reason.

This "expert" will then point to the Unication pagers, and say "see? It's possible!!!" while conveniently forgetting that is a purpose built device. Hey, my Motorola handhelds never have a problem with LSM but then they also cost $5500

 

[RFI-EMI-GUY]

See US Patent 6,061,574. I can confirm that this method is in use on LSM via direct observation of a local /\/\ LSM system.

73

Max

Thanks. I have been looking for that, interesting that it leverages FLEX paging simulcast technology.

So not only does the amplitude modulation convey synch, it also contains some error correction code (still sorting that out in my head), And it blanks the power output of some distant transmitters during the intersymbol period thus enabling FM capture of transmitters having tighter/better time differential.

The last bit has me stumped, because it would seem that in a mobile environment, the variables are such that determining which sites are undesirable requires prior knowledge of location of a receiver.

Fascinating stuff.

Sent from my SM-T350 using Tapatalk

 

[slicerwizard]

Any idea what information the amplitude component conveys? Is it strictly for synchronization of the data in the phase modulation? That might explain why an FM demod sort of works up.

Max has covered this topic in the past, so this is largely a rehash.

The purpose of the amplitude modulation is to reduce neighbouring channel interference (power drops as the transmitter frequency moves away from the channel center and towards the channel edges), but more importantly, it reduces the time period where differing signals from multiple simulcast sites are being received. At the center of each transmitted symbol, all transmitters are sending at the channel's center frequency and that is what a receiver sees. In the diagram, http://forums.radioreference.com/ge...properly-my-scanner-thread-5.html#post2879997 the symbol centers are where the upper trace (FM demod) briefly sits at the center line.

By having all transmitters running at maximum power only when near those symbol centers, the signals at those points in time can rise above any delayed signals that are still transitioning from one symbol value to the next; this enables a properly designed receiver to reconstruct the original datastream.

If an FM receiver were to only look at the signal during the symbol centers, it would see nothing, as the signal is always at the channel's center frequency at those points. Instead, an FM receiver has to look at the periods between symbol centers, where the transmitter frequencies ramp up-down or down-up to effect phase changes. Problem is, during those time periods, all of the simulcast transmitters are running at greatly reduced power and since their signals reach a receiver with differing time delays, an FM receiver sees multiple weak signals, all sending at different frequencies. This leads directly to poor symbol recovery.

With a PSK receiver, the phase change from one symbol center to the next is what is examined and the RF power levels at those centers is maximized, which promotes good symbol recovery.

So to answer your question, the amplitude doesn't necessarily convey that much information, but its presence enhances PSK demodulation and cripples FM demodulation.

So not only does the amplitude modulation convey synch, it also contains some error correction code (still sorting that out in my head), And it blanks the power output of some distant transmitters during the intersymbol period thus enabling FM capture of transmitters having tighter/better time differential.

The last bit has me stumped, because it would seem that in a mobile environment, the variables are such that determining which sites are undesirable requires prior knowledge of location of a receiver.

This sounds like tweaking the TX timing to maximize coverage in certain areas. Receivers near a transmitter will effectively hear just that transmitter, while receivers that are between transmitters, but not necessarily equidistant, can benefit from favourable timing settings. If RX is suffering due to a near low site and a far high site, timing can be retarded on the near site.

That pretty much answers everyone's questions, now doesn't it.

Indeed it does. When Paul finally loosens up and tells us how difficult something (like DMR decoding, for example) is, it means most of the bugs have been stomped on and the product release date has firmed up.

You're kidding right? It wont be long before some "expert" who knows more about manufacturing R&D than Whistler and Uniden combined will say this is a "trivial" undertaking; the scanner manufacturers wont do it because they are lazy, incompetent, or some other reason.

Pretty sure that ship sailed ages ago.

 

[kayn1n32008]

I caught that this was mostly tongue in cheek. Others might not. Not quite that drastic, but still substantial. Coupled with that is the challenge that some very basic radio things are done completely differently when using an I/Q type receiver (things like AGC, Squelch, FM and AM demodulation, and more). It would be a bigger technological change than the move from crystals to PLL, affecting many more functions and systems.

Not saying that it would never be done, but, it would be a very big undertaking.

Yet Unication and every manufacturer of P25 equipment that can be used with CQPSK have been doing it for YEARS now. Since the 396/996 scanners have come to market users have brought forward the issue of deciding CQPSK, yet Unidentifiable has refused to design a receiver to properly decode CQPSK.

In fact Uniden has released 2(3?) lines of scanners still using a discriminator tap knowing full well that CQPSK wave form is being used more and more, and that these scanners, with out serious fiddling with antennas and setting, will have serious issues decoding the CQPSK wave form.





Sent from my iPhone using Tapatalk

 

[KA1RBI]

Thanks. I have been looking for that, interesting that it leverages FLEX paging simulcast technology.

So not only does the amplitude modulation convey synch, it also contains some error correction code (still sorting that out in my head), And it blanks the power output of some distant transmitters during the intersymbol period thus enabling FM capture of transmitters having tighter/better time differential.

I'm not sure it's an error correction code per se, as much as a method for reducing ISI (inter symbol interference) that occurs due to delay spread. The reduction in ISI helps the existing ECC's that are built into P25 (hamming/golay/trellis/bch and so forth).

It is possible (and I've done so) to take complex data symbols at the output of an LSM demodulator at twice the normal rate. For proper results these complex I/Q symbols must be taken at a point prior to the differential demodulator. Assuming the symbols are all spaced evenly in time and that the symbol clock recovery is operating as usual, the result would be an alternating sequence A...B...A...B...A...B..... where A is the sample at the symbol center and B is the sample halfway between symbol centers. In any ordinary type of modulation the B values would be 100% uninteresting, but in LSM the B values (both in amplitude and phase) are really quite interesting indeed...

The last bit has me stumped, because it would seem that in a mobile environment, the variables are such that determining which sites are undesirable requires prior knowledge of location of a receiver.

And in a simulcast system consisting of, say a dozen radio towers, doesn't the system controller (or whatever the terminology is) already know that? I.e., based on which towers can hear the mobile (and if so, how strong). During the interval when the mobile is transmitting the infrastructure tracks which tower(s) are hearing the mobile, and when the mobile is receiving the infrastructure adjusts the LSM secret sauce accordingly?

The above must be understood as speculation - I'm not an authority on LSM...

Max

 

[JT-112]

And in a simulcast system consisting of, say a dozen radio towers, doesn't the system controller (or whatever the terminology is) already know that? I.e., based on which towers can hear the mobile (and if so, how strong). During the interval when the mobile is transmitting the infrastructure tracks which tower(s) are hearing the mobile, and when the mobile is receiving the infrastructure adjusts the LSM secret sauce accordingly?

For 1:1 systems (e.g., cellular), yes the system works very hard to keep track of where the mobile is.

But for 1:many systems, you have the vast majority of participants in a talk group being silent, and therefore the system has no real idea of where they are. Therefore it cannot make any adjustments to favor them - any adjustments are necessarily going to have a negative impact for other passive listeners.

Put another way, you don't know that Big Chief Six Actual is listening in from Farsville when all of the action is in Neartown. Adjust for Neartown, and Chief can't hear any more. No es bueno.

However, in reality, no matter my location - even dozens of miles away from the specified coverage area - I always get a great signal when passively listening to any LSM system. They've done a great job of knocking down ISI with LSM to the point where it's just not an issue anywhere in the receivable footprint.

Some of this is likely smart engineering on site selection, antenna patterns and ERP adjustments, but it all works very well. Topology and topography combine favorably as well, at least where I listen the most.

At one point I had documentation/presentations on LSM from Motorola - I'll see if I can't find them again, they made it quite understandable.

 

[RFI-EMI-GUY]

What sixtytwo said is right. Simulcast systems transmit to entire fleet at once, so you cannot direct a correction toward a specific subscriber without fouling the signal for someone else.

The action of reducing specific transmitter power levels to correct ISI, should by that action affect all radios. Let's assume this is a preplanned action to blank certain distant transmitters. I wonder how that would affect receivers at the fringe? On the other hand, it should be beneficial to crude scanner discriminators as well, yet apparently it does not. From the reading, I get a sense that the amplitude is an additional information channel adding a timing mechanism at a minimum.

Sent from my SM-T350 using Tapatalk

 

[JT-112]

I only see amplitude as being a way of massaging the symbols width so to reduce ISI.

I'm not sure there's any other function there. KISS is in play here - get too tricky, too dynamic, too configurable, that's when things bite you in the butt.

I could be wrong... :^)

 

[KA1RBI]

- any adjustments are necessarily going to have a negative impact for other passive listeners.

fair enough, but the adjustments ARE there... Another possibility that I speculated about back in 2012 was that these "variations" in the signal (halfway between the symbol centers) might have to do with adaptive equalizer training at the mobile unit. Back in those days I made a writeup about it, showing some of the fascinating patterns in the data.

LSM Gallery - Page 2

73

Max

p.s. and If you're curious, Page 1 can be found at

LSM Gallery

 

[MTS2000des]

Yet Unication and every manufacturer of P25 equipment that can be used with CQPSK have been doing it for YEARS now.

and the Unication is within the same ball park of what modern high end digital scanners sell for, and are professional tools with superior performance on LSM.

How can Unication, a relatively small company with a narrow product line build a quality receiver and sell it for a profit in the same price range and make a profit but scanner manufacturers can't? Or is it they just won't?

 

[RFI-EMI-GUY]

and the Unication is within the same ball park of what modern high end digital scanners sell for, and are professional tools with superior performance on LSM.

How can Unication, a relatively small company with a narrow product line build a quality receiver and sell it for a profit in the same price range and make a profit but scanner manufacturers can't? Or is it they just won't?

I spoke with a Unification rep at IWCE and she says currently they have no plans for a "scanner" type product. She said the focus is on releasing P25 phase 2 code.



Sent from my SM-T350 using Tapatalk