Complex Digital Demodulation Hardware Discussion

Status
Not open for further replies.

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
I am starting a new thread here which I intend to be devoted to the discussion of the design and construction of and experimentation with hardware associated with the effective demodulation of complex digitally modulated RF signals; especially that beyond the simple FM discriminator taps heretofore well discussed and understood (though, due to the limitations in terms of money and resources of those of us involved in this at the hobby and personal experimental level, I certainly would not exclude discussions concerning getting more out of the old tried and true FM discriminator tap efforts at this stage if at all possible!). This can be looked at as a companion thread to the software discussion concerning the rather new and innovative DSD demodulation software.

Some discussion of this has already started on that thread as well as Unitrunker's ProVoice demodulation thread. I would like to request that the moderators move those posts concerning this hardware discussion off of both of those threads and move them here when convenient. I feel we inadvertently hijacked both of those threads and would like to set things right as well as give us "hardware fiends" a proper thread for discussion.

The primary intent, from my point of view, of this thread is to discuss effective hardware designs and implementation to deal with the more complex forms of digital modulation wherein much of the information is contained in the phase and/or amplitude of the signals rather than just the frequency deviations as in the comparatively simple C4FM mode. This would include discussion of LSM CQPSK and any other of the more complex digital modulation modes commonly used in RF transmission.

Also, as I stated above, any good results coming from this discussion can be used to better enable the software efforts as discussed in the DSD thread.

I hope this leads to interesting and effective discussions and look forward to what the future may bring as a result of these efforts, both from a software and hardware standpoint.

Once again, I apologize on my part, for the inadvertent hijacking of those other threads!

-Mike
 
Last edited:

jcardani

Member
Joined
Jan 16, 2002
Messages
1,352
Location
Orlando, FL & Ocean City, NJ
I have an ICOM R-8500 with 10.7 MHz IF. Would the Intermediate Frequency be the place to start here?

I also have experience in using ExpressPCB and ExpressSCH CAD software so I can document what's needed. Has anyone figured out a circuit solution or at lease started?

thanks,
 

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
I'm thinking "yes" to your first question - the idea just being to investigate improved methods of demodulating a digital signal with more complexity to it than, say, a "typical" C4FM signal. Hopefully, the side effect of this might also yield improved performance on even P25 Phase 1 systems with heavy simulcast distortion.

Anyway, I started this thread just to give a place to discuss the hardware aspects of such work given that the DSD software effort has made such great strides and generated much interest. I was hoping the posts that generated my hardware interest would be moved here by the moderators but, for whatever reason, that did not occur and both of the affected threads have since been closed so I doubt the posts will move here in the future.

Anyway, my background is in the old DAMPS cellular PI/4DQPSK modulated TDMA stuff and I know we worked with discrete I/Q demodulator IC's which worked on the final IF which was somewhere in the 200KHz to 300KHz range, I think (it's been quite a while so my memory is pretty suspect here!). After that, the I and Q signals were A/D'ed and sent to a big DSP engine (again, as far as I remember). My biggest problem in terms of remembering this process is that I can't recall where the equalization was done - in that DSP or in a completely separate DSP (for some reason I seem to recall a dedicated DSP for just the equalization but that may have only been in a design discussion and not in the final result - not sure) and at what exact point in the signal chain, etc. Oh well...

From what I can gather from the little research that I have done on the web, the CQPSK modulation of P25 Phase 2 (at least, as I understand it, that is the modulation it will or is using) sounds somewhat less messy to deal with than the PI/4DQPSK stuff I wallowed in back in the day. I know we had to use a linear final amp or at least a "linearized" one for the transmitter because the signal contained amplitude components which needed to be accurately handled. It looks as though the CQPSK stuff isn't saddled with that problem, contrary to what I, at first, supposed. Also, on the receive side, apparently, there is still hope for a discriminator approach and, according to at least one published analysis I read, in some ways that may handle the signal better than an I/Q demodulator (if I understood the paper correctly - and I admit I am VERY rusty now in terms of my current knowledge of digital RF signal handling) which was a complete surprise to me!

I am really hoping that folks with more current and complete knowledge of this subject can start chiming in here and start up a good continuing discussion (MSM_Maria, KA1RBI, and ronenp are the folks I am hoping will jump in here - they were the ones who (re)sparked my interest due to their posts in the software discussion threads)!

I've given the "SoftRock" dealywonker a quick glance but I think KA1RBI's approach is more attuned to this effort. I think he has the right "bullet points" in mind in terms of what needs to be solved (I completely overlooked the AGC problems, for instance!).

As of now, I think looking at the 10.7MHz IF is a good start since it bypasses the weaknesses of the 455KHz IF filter in most consumer grade receivers.That's not to say that the 10.7MHz filter is better, per sey, but, since it is far wider than the actual signal bandwidth (of a P25 type signal) any group delay issues would not be so constraining, etc. But, of course, adjacent channel problems now come to the forefront but, for experimental purposes, given a good enough signal, we can maybe deal with it?!

I'd say, let's start simple and see what improving the LO stability and phase noise does for the quality of the decode as a first step. Use a good high quality lab grade synth if possible or at least as good an external source as possible to take the place of the consumer grade receiver's internal LO.

Then, try improving the AGC action to accommodate CQPSK simulcast signals properly and see where that takes us.

Finally, we can begin to look into different demod approaches including I/Q and maybe better forms of discriminator tapped "data slicer" approaches.

To me, it would seem that a discrete I/Q demod shouldn't be all that hard to rig up using a dedicated I/Q demod IC; they seemed ubiquitous back when I worked in the field but maybe they all got rolled into "system-on-a-chip" designs now and discrete IC I/Q demodulators are now rare?! I don't know...as I said, I've been out of the game for awhile.

Anyway, thanks for posting to this thread;-)! Hopefully the hardware folks smarter and more up-to-date than I will jump in and keep it moving!

Keep your PCB CAD software ready - I seriously hope we will need it soon!! Thanks for the offer!

-Mike

P.S.
To the Mod's - I'm starting to think that maybe starting a whole new subsection dedicated to LMR digital voice demodulation might be a good idea since it is deviating somewhat from the "Trunking Control Channel Decoding" primary focus of this section. Both software and hardware threads could be moved there. I think it would make things a tad cleaner and easier to navigate and search, etc. Just my 2 cents...
 
Last edited:

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
I'll check out those PDF's - I have a reeeeeealllllly sloooooooow internet connection so they'll take awhile to download for me. I'm guessing that any discrete demod IC is going to need at least some some filtering to work, though, which could be an issue. Once you have settled onto a final IF you will need to use a filter with good group delay aspects. And we're going to have to pick one that works with the P25 Phase one and/or Phase two signal bandwidths (of which I believe they may be more than one, system choice-wise, from some of the stuff I have read on these forums).

Cursory lookover seems to indicate that you could use the eval board directly from the 10.7MHz IF though I would be careful about leveling the IF so as not to overdrive the LNA on the board. And we really don't need the second I/Q channel (umm, at least I don't think we do....) but it won't hurt, especially for experimental purposes.

I'd setup like the test diagram on the eval board doc says at least to initially get familiar with the device under ideal circumstances before throwing it into the line of fire from a receiver's 10.7MHz IF with real world noisy signals - assuming you have the necessary lab equipment. Get to know it first and understand what you are seeing on the o'scope, etc. I'm now starting to think some decent filtering will be a necessity unless you can find a really good solid P25 Phase1 or 2 voice signal that transmits almost continuously with no nearby adjacent and alternate channel activity - kinda tricky at best! Maybe if you have access to a good (expensive!) current service monitor or lab synth that does the correct modulation that would be the way to go. I really wouldn't like to just drop right into over-the-air testing out-of-the-box just yet.

In theory, this is pretty simple BUT in practice maybe not so much - which is kinda why I suggested starting with tying down the LO phase noise-wise and working on the AGC. Nevertheless, this is good in that it shows me relatively familiar territory which is still in use; I 've been busy on other stuff but was going to search the net for demod IC's - your efforts are a good start!

Even if this is overkill - it may make a good test platform (once we get the details worked out) to evaluate this approach for demodulating P25 and other digital modes.

I'll look these over more carefully and get back to you - thanks for the information!

Working on computer networking issues right now...;-(....

-Mike
 

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
In case anyone's been scracthin' their heads wondering what my obsession is all about here is what I am envisionin', contemplatin', and generally ruminatin' aboot...

I would like to see a dedicated hardware "black box" designed to include the following:

A)
Programmable input to accommodate common wideband IF outputs of consumer (and maybe other type) receivers possibly ranging from about 5MHz to 100MHz or so. This would have wideband DC blocking capability built in to handle those IF outputs with a DC component or else it would just be DC coupled with the capability to handle varying DC bias voltages on the input.

B)
Built-in low noise and stable voltage and temperature compensated local oscillator with switchable external input to allow use of external oscillators/synthesizers as well. Ideally, the internal LO might be programmable to allow a limited selection of baseband frequency ranges.

C)
High dynamic range buffer amp and mixer stages with good IM characteristics for the downconverting stage.

D)
Switchable final IF filters with good group delay characteristics to handle various complex digital modulation modes. Various filters would accommodate various bandwidths.

E)
Programmable AGC loop to allow selectable optimization for various modulation modes and RF environment conditions.

F)
Buffered input quadrature demodulator capable of being programmed to handle various digital modulation formats including, but not limited to, CQPSK, PI/4DQPSK, GFSK, and C4FM with characteristics programmable to be optimized for each; output would be I and Q baseband which is further buffered and filtered appropriately.

G)
Capability to switch in other hardware demodulation circuitry in lieu of the quadrature demodulator above such as the more common FM discriminator with data slicer. This may also include analog demodulation modes to accommodate standard AM and FM with various IF filter bandwidth selection and programmable FM deviation optimization. This should be programmable entirely independent of mode to allow maximum experimental value.

H)
I and Q output would be analog to digital converted and coded to be included in a standard USB interface to connect to a PC for further software processing. If included, analog output from analog FM and AM modes would also be converted to digital USB for PC use as well.

I)
All programmable options (input frequency ranges, filter selection, external/internal synthesizer selection, AGC loop characteristics, buffer amp gain settings, demodulation hardware selection, etc,) should be controlled via on-board firmware on some form of memory (flash RAM or EEPROM?) with sufficient capacity to allow extensive experimentation and storing of favorite settings and be FULLY upgradeable/programmable via open source software. A USB interface strictly for this purpose and entirely separate from the demodulated output data USB interface would be built in.

This is my first iteration of my goal/desire/dream box. I won't mince words - it's a tall order but I think it's a good one. As far as I can tell, although there are I/Q receivers on the hobby/experimental market I don't think they are optimized and specialized for LMR (Land Mobile Radio) demodulation. My desire is for a box strictly optimized for THAT purpose alone.

It would give the software geniuses a good platform to work with to experiment on demodulating current and future narrowband LMR modulation modes, analog and digital, using common PC hardware and open source software.

In effect, this box would replace all of the stages following the host front end receiver's final wideband (greater than 400KHz roughly speaking) IF stage. Use of an off-the-shelf receiver as the front end reduces cost relative to a complete receiver design and allows software feature expandability and flexibility. Of course, the quality of the final output goes up or down with the quality of the front end host receiver (see below for a big issue here).

One major problem I still see, unfortunately, is that I still can't figure out a good way to compensate for the poor quality of the front end receiver's local oscillator in terms of phase noise and possibly overall stability. No matter how good this box is it cannot fix issues stemming from those problems in the front end receiver. I can only speculate at this stage that one could attempt to modify or replace the internal LO of the receiver to improve performance if it proves necessary to do so in a given circumstance - I know that kicks the cost and complexity issues up several notches but I currently am unable to think of a better solution.

Hope this starts to explain the method to my madness;-).

-Mike
 
Last edited:

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
Irritatingly, though the "Edit" function is still available on my message I cannot actually edit it given the two hour time limit so my latest thoughts must be put down here (keep in mind, this was intended to be put in the previous post following the second to last paragraph):

One final point - this box is primarily intended to provide the hardware base for an RF/IF backend dedicated to and optimized for the demodulation of narrowband LMR voice signals. It is not initially intended to deal with trunking control channel processing. Since this is fairly well handled now by existing FM discriminator tap plus PC soundcard hardware I don't see that as an issue. Of course, I also see no reason that the hardware of this box couldn't work on trunking control channels in the same way as it would on voice signals; in fact, perhaps on some oddball types it might prove superior (to current basic receiver discriminator taps; perhaps including the infamous LSM simulcast P25 control channel? And maybe the TETRA CC - I have no idea what it is like). It may even be possible to use two such boxes split from the output of the wideband IF of one host receiver front end providing the control channel is transmitting within the passband of the IF of the host receiver and all of the expected voice channel frequencies fall within that same passband. Otherwise, of course, you could have two receiver front ends and two backend boxes for a complete "dual-tuner dual receiver" approach in the most complex yet versatile and high performance case. It's up to the user and the controlling and processing software involved.

NOTE: Ahh...I see...I guess the time limit expired during the time I was writing the above so though the "Edit" function WAS there when I started it wasn't when I finished...oh well...

...funky webby digiweenie software forumeee wigglywonky stuff...;-) Sorry, you can tell - I am a analog and RF hardware weenie through and through! And, yes, I admit it - I am envious of you software gurus! You get the glory and the babes, etc....whine whine...yes I WILL have cheese with that thank you! Anyway, I've tried - every time I get past the "Hello World" stage pieces of my brain seem to liquefy and ooze noisily out of my aural orifices and I seem to have a sudden and irresistible urge to go off and do a spreadsheet analysis of third order intermod products, hug an o'scope, fondle a test lead, poke morosely at various SMD components, solder something, nuzzle up to a RF Vector Network Analyzer or some such; yup! I am seriously coding challenged! And I actually have a best friend who is a software weenie!;-)
 
Last edited:

ronenp

Member
Joined
May 8, 2002
Messages
520
what do you mean by "demodulation of narrowband LMR voice signals" ?

I myself would like to be able to decode (hear , listen to)
EDACS (ProVoice)
MotoTRBO
P25
Digital Motorola Trunked
NXDN
IDAS
DMR
Tetra (if not encrypted)
D-STAR
iDEN
InmarSat Digital Voice (i dont know the Modulation type but know it can be decode with DVDongle )

and as a benefit
Receive the DVB-T Audio signals

Im also a RF and Analog and dont know to write software (unfortunetly) but i know it can be done and in fact there is a company that have such software that decode lot of digital signals but it cost 14000 $ far then what i want to invest just for hobby
so I hope the software fellows here might do something about it...
Best Regards
Ronen - 4Z4ZQ
Ronen Pinchooks (4Z4ZQ) WebSite
 
Last edited:

grosminet

Member
Joined
Jan 21, 2004
Messages
278
Tetrapol

we could add Tetrapol . It is an 12,5 khz . France can send developpers some wav files !!
 

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
what do you mean by "demodulation of narrowband LMR voice signals" ?

I myself would like to be able to decode (hear , listen to)
EDACS (ProVoice)
MotoTRBO
P25
Digital Motorola Trunked
NXDN
IDAS
DMR
Tetra (if not encrypted)
D-STAR
iDEN
InmarSat Digital Voice (i dont know the Modulation type but know it can be decode with DVDongle )

and as a benefit
Receive the DVB-T Audio signals

Im also a RF and Analog and dont know to write software (unfortunetly) but i know it can be done and in fact there is a company that have such software that decode lot of digital signals but it cost 14000 $ far then what i want to invest just for hobby
so I hope the software fellows here might do something about it...
Best Regards
Ronen - 4Z4ZQ
Ronen Pinchooks (4Z4ZQ) WebSite
All of those with the possible exception of the Inmarsat one (it may be narrowband and I suppose it might be considered LMR but I know little about it. However, if it is within a small passband, say within 50KHz or less, then it should be fine given the right RF front end and antenna, etc - being a satellite signal and all.) are what I would consider to be Land Mobile Radio narrowband voice modes. In the US, the term "LMR", "Land Mobile Radio", is used as a catch-all term to define general public and private two-way radio communications - what we as radio hobbyists generally are interested in. I further focused on "narrowband" to differentiate from wide band signals such as direct sequence spread spectrum (DSSS) of which CDMA is a type used in the US and broadcast FM and television. This hardware (proposed) as described by me is intended for use on narrowband transmissions as opposed to wide band spread spectrum or other wide band transmissions such as broadcast FM or television transmissions.

The hardware I described would provide an ideal platform for the further development of the software solutions being discussed in these forums. The primary intent is to overcome some of the limitations that plague the current crop of consumer low cost radio receivers such as the noisy LO's, poor filtering, and limited baseband demodulation capabilities (such as relying solely on an FM discriminator as a source to process). These limitations are starting to be noticed as potential stopping points for future development of software demodulating solutions when dealing with signals wherein information is contained not only in the phase changes in the signal but also in amplitude changes and/or when dealing with extreme forms of closely spaced (in an angular sense) phase changes. Phase changes by themselves can result in vector amplitude changes especially when many phases are employed. The PI/4DQPSK modulation used in the old DAMPS cellular standard and, as I understand it, also used in the TETRA standard (I haven't confirmed this) is an example of such a mode. Also, P25 Phase II is (as I understand it, likely to be a multi-slot TDMA mode which uses a type of phase shift keying unlike the current Phase I mode (which uses a simpler frequency shift keying mode - C4FM); it uses or will use, as I understand it, a form of CQPSK (compatible quadrature phase shift keying) which, though somewhat less complex than the PI/4DQPSK mode I mentioned earlier, may still tax the capabilities of low cost discriminator tapped receivers (I am unsure here, I've read conflicting reports on this) so having other options, such as the availability of a quadrature demodulator aka an I and Q demodulator may be beneficial as well as having better tighter local oscillators and flexible AGC loop options, etc.

Certainly, a lot can be done with what we have with the amazing software being developed now and good PC and soundcard hardware. But, no matter how well the software is developed, at some stage it will run up against the hardware platform it is dealing with. I don't mean the PC, of course, I mean the interface hardware such as the sound card and the radio receiver. This is what I would like to overcome by developing a good solid common hardware RF interface focused deliberately on narrowband LMR signals to keep costs to a minimum and quality as high as practical.

As for Tetrapol, from what little I know about it, the hardware I am proposing should handle it fine. I can't recall the modulation method but if it is as simple as P25 Phase I uses in the US than it may work fine with the current discriminator tap plus sound card or data slicer methods with the right software.

-Mike
 

ronenp

Member
Joined
May 8, 2002
Messages
520
As for Discriminator as an output ...
My opinion is that we should not "invent the wheel"
mabe we have to look first in some Comercial radios (such as Icom or Motorola or Kenwood) and see if the data is decoded from the Discriminator or from BaseBand

However if we realy need a BaseBand , Than My thought is Using a Softrock Receiver (SDR) attached to an
IF ......
I have such a SoftRock and im impressed from its results specialy considering its price (cost me 45$ as a kit) and then we can get a I Q output for future proccessing

What Bother me is the Size
It will not be a Fully Portable because it will need a Pc and a Notebook is the smallest size ...
and I'd like to have a portable scanner .... but its a good start ..... specialy when i dont see any other products that are capable doing even half of what we intend to decode .

Best Regards
Ronen - 4Z4ZQ
Ronen Pinchooks (4Z4ZQ) WebSite
 
Last edited:

ronenp

Member
Joined
May 8, 2002
Messages
520
Real hardware from five years ago.

subverting the dominant paradigm in consumer receiver design
---------------------------------------------------
I have Emailed with this guy few times and even listen to this radio
currently it only decode Trunk.....

But it took him almost five years from the drawing to the final product .......


software to decode lot of modes .......
WAVECOM-Decoder W-CODE W61PC W61LAN
 

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
That receiver design is interesting (the one Unitrunker linked) and provides some good ideas, most especially the synthesizer design and implementation. However, it is overkill in some sense and falls short on certain areas I would like to concentrate on (particularly, and most obviously, the demodulator). I really don't want this to be a "Swiss Army Knife" receiver with DC-to-daylight capability - at least not in the immediate future. I want to simplify in some areas by designing for very specific applications (hence the focus on narrowband LMR) and then offset that by carefully considered complexity where applicable to gain maximum flexibility for the primary focus - a self-contained hardware box for demodulating baseband narrowband LMR digital signals down to bits for processing via computer.

I am hoping, perhaps naively, that we can get away with something that can use common consumer receivers as a "front end". The problem with that, at least the biggest one I see, is the LO of those receivers. How noisy can we tolerate them to be and get something usable in real world environments for proper decode of a wide variety of current and future complex digital modes.

Even given the above issue, assuming this "vaporware" thought exercise ever goes beyond pure thought, I still think the effort concentrated on a "back end" only design will still be useful even if we are forced into a full antenna-to-bits design.

Ronenp - I completely agree on the need to examine real radios! I was thinking today about wishing we could look at a TETRA portable radio's internal receiver design because IF TETRA really uses PI/4DQPSK as someone on here said earlier, then its demodulator should give us a good indication of a design which could handle some pretty nasty complex digital signals. That is, unless it's all wrapped up in some proprietary system-on-a-chip large scale IC which we can neither look at the innards of nor use in a flexible multi-mode analog and digital high performance sense - it might be great at TETRA and pretty much useless on anything else.

But looking at one internally or seeing a detailed schematic and/or block diagram would still be a great thing to do! I wholeheartedly endorse that investigative effort! Of course, I'd like to look at the design of the receivers and especially the final IF and demodulator portions of the Harris and Tait handhelds that do pretty much everything within the 138MHz to 950MHz LMR bands including, supposedly, future TDMA modes like Phase II P25!

I think if we could get the designs of at least the demodulators used in the TETRA portables and/or the Harris/Tait super handhelds we'd be on a good initial role...well, again, IF they aren't rolled into some large scale integrated proprietary "black box" IC. Paying boocoo bucks to Tait/HarrisMotorola/whomever for the parts and the licensing to use them doesn't sound like a practical approach to this effort to me.

As to the SoftRock - I think I am just seriously dense when it comes to the current state of SDR. I need to better understand it because something doesn't seem right to me; as I said - I admit my ignorance here. It still seems to me a better approach to use a good solid I/Q demod IC with a good filtered, stable, and reasonably phase noise free IF signal in the "goesinna" port. The SoftRock seems like a great low cost spectrum analyzer for HF use but I'm not sure how well it will work for what we want. But, hey, I'm stupid here, I admit it! I need to be enlightened. If you or anyone else has one of these and has done exactly what we want with it (take a PI/4DQPSK or equivalent in terms of complexity voice signal in the presence of a nasty real RF environment with lots of fading, phase distortion, and ugly adjacent and alternate channel interference and brought it to clean computer-processable bits or all the way to clean human-processable AF) and done so while keeping the cost reasonable, the construction feasible with hobbyist level parts and tools, and the construction and operation reproduceable without lots of tweaking I'd like to get the full details and get myself edumakated!!

As to portable usage - I'm kinda thinking that is secondary in the near term or at least will be for a while. Ultimately, at least in my "final phase of Part 1" scenario, I want to provide the design for and construction details of a good reproduceable and maybe even mass manufacturable hardware box which is as self contained as possible but which uses a separate user provided RF front end and can yield nice clean bits on the "goesoutta" to feed a hungry PC and an even hungrier software genius! If and when that stage comes to actual fruition THEN, assuming the current or future consumer scanner manufacturers haven't nailed it by then at a much lower cost than we could ever hope to hit, we can look at making it all nice and portable. Even then, a good flexible demodulator that handles phase and/or amplitude changing digital signals with oodles of phase and amplitude points in the midst of real world nasty RF environments would still make a great platform to develop new and improved software processing solutions for both current and future digital modes.

Phase 1 P25 is already off-the-shelf portable and ProVoice looks perfectly feasible to be with just new software and the current normal hardware approach. The other stuff - some might be while others, I just dunno (with easily available non-proprietary hobbiest level hardware).

Well, whatever the case - it's all a good learning exercise, at least for me!

-Mike
 
Last edited:

Mike_G_D

Member
Joined
Dec 19, 2002
Messages
1,111
Location
Vista, CA
Ya know, since we're on the subject (well, I guess I did kinda start the thread which does, by its nature, consist of a discussion of one particular subject so that actually is implied...anyway) and I know there are many skeptics, critics, head-scratchers, and general hecklers out there who think this whole idea is either impractical (due to cost, complexity, and reproduceability issues) or is completely unnecessary (because anything and everything can be done with a discriminator tap and a decent PC and the right software) I want to address them, well, at least for the first roadblock (cost, complexity, etc.).

Actually, I want to address them by addressing those who CAN do this cheaply and easily: the hardware manufacturers! Even if this idea seems too offbeat to appeal to the "Big Two" scanner manufacturers it should still appeal to the more "nitchey" manufacturers (I'm looking at you RF Space and Winradio!).

This should be an obvious product - there are no proprietary license issues at all since we are not dealing with the vocoder or any software - that's left to the user. All we need is a solid box that takes downconverted IF in, down converts it once or twice more using well designed stable PC programmable synthesizers with low phase noise, filters the result nicely with selectable high performance filters with good group delay characteristics, processes the result of that down to in-phase and quadrature baseband signals, converts those signals to digital using good quality analog-to-digital converters and spits out bits for sending to the user's PC, ideally using a USB port but I'll settle for any common interface including the old RS232 serial and parallel ports.

Your current products probably can do this but they also do more and that is reflected in the cost. Make something that JUST does the above and is targeted toward LMR digital radio signal "experimentation" and I think you will find a fairly hungry and reasonably profit-inducing consumer base. I am fairly sure you could make these boxes far cheaper and probably at better quality than we could and, given mass manufacture capability, you could sell them at lower cost compared to your current product lines and still make a profit, at least on the individual sale - I know it's "niche" but, in a way, so is your current product line!

Heck, I'll bet you could go all the way from antenna to bits and still sell for under $300! That's cheaper than today's digital scanners. Let us still provide the software and PC processing and take the heat in terms of legalities, etc. But at least the "back end" approach should be "doable" for that cost, I would think.

'Nuff said - I just wanted to throw that out there; I do not expect a response, of course, at least not from the actual manufacturers. I just want them to scratch their heads, run the numbers, and think about this.

Heck - make it a kit! Some of us still like to put things together!

-Mike
 
Last edited:

ronenp

Member
Joined
May 8, 2002
Messages
520
Again my 2 cents
I Have a Detailed Service Manual for XTS5000 Wich is APCO 25 Capable and it have there a Chip Called
Abacus III (AD9874) and it get an IF and have A-->D there so i it looks like it do the decode by digital
Certenly not By Discriminator ....

Also I have a schematic of the P25 Card that stick to the Uniden Scanners like the 296D
That are capable of P25 RX and also there the card get a baseband and not a discriminator .....

As For the SDR I Receive very well DRM (Digital Radio Mondiale) Over the HF bands And the HF Bands suffer from Phase and signal strength variations and still recieve it with a Software for DRM that expect I Q signals from the Sound card of the PC
I dont remember what Modulation is DRM but it is a digital and belive me that it is not comming good after few 1000th of KM to the receiver antenna .....
Unfortenetly there is no other Digital modulatios decode software for I-Q that im aware of that i can test ....
However im more then willing to record via this SDR receiver any type of modulation the fellows needs to have as a sample or to check if the I-Q output of this receiver is sutible inough for use as a front end for I-Q output ......
Im currently know there is MotoTRBO iDEN P25 ProVoice Signals in my RF Receive Range ..... that i can record Signals .......

one little question are you by chance a Ham radio
Best regards
Ronen
 
Last edited:

systron

Member
Joined
May 8, 2010
Messages
20
Europe Digital Voice Systems TETRAPOL / TETRA

we could add Tetrapol . It is an 12,5 khz . France can send developpers some wav files !!
we should point this to the dsdauthor....

tetrapol uses strong end to end encryption as default. sometimes not... emergency calls, out of usergroups transmission. interesting things are the control channel, data transmissions, units id, paging messages, trunking, and unencrypted voice. tetrapol uses GMSK and RPCELP voice codec.

As for Tetrapol, from what little I know about it, the hardware I am proposing should handle it fine. I can't recall the modulation method but if it is as simple as P25 Phase I uses in the US than it may work fine with the current discriminator tap plus sound card or data slicer methods with the right software.
-Mike
The most used System in Europe are Tetrapol. It should be possible to decode (CC) with a soundcard and a simple discriminator tap and 15 KHz filter.

but we need copyright protected voice codecs as software code... (AMBE,RP-CELP,ACELP)

happy hacking
 

ronenp

Member
Joined
May 8, 2002
Messages
520
I have just read your new Writing after i have post my
First of all i will ask the Softrock Group here
softrock40 : Softrock-40 Interest Group
What they think about Softrock as a Front End ....

As for your new posting .....
I think that the main issue is the decode and not the RF Because there are a lot Cheep Radios (under 100$ ) that the If can be taken for decode .... but ... when there is no decode program for Non standart modulations then it stopped ..... i would not buy a radio these days that have no capability to decode nothing other then AM-FM
but i can spend 1000$ on a digital scanner that can decode for me other digirtal then only P25
But the Scanner manufacturere dont make such radios so other solutions are being built slowly but surely ... like the TRBO decoder (unforetenetly currently for linux) I know that D-star also there is something going on ..... So in general if you think of a RF solution i think it have to come with a "decode" solution as well

Just remember that the new Uniden scanner that is capable of receiving P25 (the only digital it can decode) cost say 500 $ (although you can find cheeper models that can decode P25 also
and this Project hardware comes for the moment with no software at all so currently the Real Scanners have a big advantage .... so we must understand that the decode software is a very importent part ......

of course that if the RF hardware will be more flaxible (say using SDR technology ) then the limit will be the decode software and as long as new softwere that can decode newer or more digital signal will be it will be only replacing old software and using same hardwere ... unlike real scanner that in that case it will be needed to buy new scanner .....
Best Regards
Ronen
 
Last edited:

MattSR

Member
Joined
Jul 26, 2002
Messages
403
Location
Sydney, Australia
I Have a Detailed Service Manual for XTS5000 Wich is APCO 25 Capable and it have there a Chip Called Abacus III (AD9874) and it get an IF and have A-->D there so i it looks like it do the decode by digital Certenly not By Discriminator ....
It's been that way since the first ASTRO saber was released in 1993. Quite clever too - it uses current flow to represent I/Q values instead of voltage, as current is less susceptible to outside interference from other parts of the radio. There is also a software discriminator C4FM slicer in the DSP, and the radio constantly measures the error rates from both I/Q receivers and slicer and automatically picks the one with the lower BER. /\/\oto's engineers are smart guys!
 

KA1RBI

Member
Joined
Aug 15, 2008
Messages
581
Location
Portage Escarpment
If you or anyone else has one of these and has done exactly what we want with it (take a PI/4DQPSK or equivalent in terms of complexity voice signal in the presence of a nasty real RF environment with lots of fading, phase distortion, and ugly adjacent and alternate channel interference and brought it to clean computer-processable bits or all the way to clean human-processable AF) and done so while keeping the cost reasonable {snipped}
This has been done. "Reasonable" is obviously relative and should be viewed as a moving target.

In the OP25 project we have fully cracked the LSM CQPSK problem. The PC software requires access to the signal at a point prior to FM detection. The hardware used in this case (the USRP) is going to cost someone in the order of double the current price of a digital scanner.

In my view there are two possible paths -
a) use dedicated SDR hardware
b) modify an existing receiver to add some sort of "tap"

If using option "b" the implication seems to be that the "tap" should be as cheap as possible and let the PC software do all the heavy lifting...

Below is a constellation pic of what LSM looks like. You can't do this with a disc tap...

73

Max
 

Attachments

Status
Not open for further replies.
Top