APRS

Automatic Packet Reporting System Discussion

Re: APRS

Postby VK7HH » Thu Jun 19, 2014 9:22 am

Joe 7JG has been in the desert the past month or so with a mixture of HF/VHF APRS that seems to of worked well.

http://aprs.fi/#!mt=roadmap&z=11&call=a ... ail=604800

(see last 7 days track).

Our OT3 running on VK7RML seems to be stable now and working fine.
I've found that it get's upset with certain audio configurations fed into it and doesn't decode everything. Sometimes it can lock up too.
I've got in the script a reset command to run every 12 hours to stop against lock ups and it seems to work fine.

The radio we have is a PRM80. I think from memory I removed a SMD resistor in the RX audio line on the board and fed this from the DB15 connector on the back of the radio (pin 10)
VK7HH
Forum Diehard
 
Posts: 232
Joined: Thu Sep 08, 2005 11:18 pm
Location: Southern Tasmania

Re: APRS

Postby VK2OMD » Thu Jun 19, 2014 9:45 am

VK7HA wrote:...
I've found that it get's upset with certain audio configurations fed into it and doesn't decode everything.


AFSK can be unforgiving of overdrive of the transmitter. The problems are obvious if you look at the audio spectrum of demodulated signal.

Some types of demodulators are moderately tolerant of the fautly signal, some most intolerant.

Listening on air, some transmitters are overdriven, some underdriven, few just right.

A long term problem of AFSK in ham hands.

Owen
VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK7HH » Thu Jun 19, 2014 10:22 am

VK2OMD wrote:
AFSK can be unforgiving of overdrive of the transmitter. The problems are obvious if you look at the audio spectrum of demodulated signal.

Some types of demodulators are moderately tolerant of the fautly signal, some most intolerant.


Owen



That's correct Owen, but that is out of our (digipeater owners) control. Some users have the "all knobs to the right" philosophy. But from what 2XSO has described is an issue I had with that same unit with a correctly driven APRS transmitter. Another issue with those units would be after digipeating 15 mins or so, it would lock up and stop receiving. I discussed it at length with Scott from Argent.
Another issue was powering the devices. Powering it via the USB connector was more unstable then via the DC supply input. Also move the radio away from the Tracker unit. RF *may* upset things. We also use TNC-X's with X-Digi's on most of our digipeaters.
VK7HH
Forum Diehard
 
Posts: 232
Joined: Thu Sep 08, 2005 11:18 pm
Location: Southern Tasmania

Re: APRS

Postby VK2OMD » Thu Jun 19, 2014 11:06 am

VK7HA wrote:...
That's correct Owen, but that is out of our (digipeater owners) control. Some users have the "all knobs to the right" philosophy.


Yes, APRS is seen as a turnkey tecnology and no technical knowledge is needed to properly configure a transmitter on a shared channel.

VK7HA wrote: But from what 2XSO has described is an issue I had with that same unit with a correctly driven APRS transmitter. Another issue with those units would be after digipeating 15 mins or so, it would lock up and stop receiving. I discussed it at length with Scott from Argent.
Another issue was powering the devices. Powering it via the USB connector was more unstable then via the DC supply input. Also move the radio away from the Tracker unit. RF *may* upset things. We also use TNC-X's with X-Digi's on most of our digipeaters.


Sounds like a device to be avoided... but if you are afflicted with them, understanding their decode limitations might help maximise the benefit you obtain.

I think it uses a DSP based decoder.

In the old telephone modem chips, different techniques were employed for the demodulator.

The Exar chips IIRC used a PLL demodulator, and it performed very badly if the low tone was high in level compared to the high tone. The effect of the preemphasis followed by the limiter is to effectively wipe out the preemphasis on the transmitted signal if the input reaches the limiter threshold, and the receiver de-emphais results in the high frequency end being attenuated more than the low frequency end... delivering the worst demodulation signal for the Exar type chips.

Of course, overdriving the limiter creates a whole bunch of intermod products (see 3 above) which may degrade the DSP demodulator.

Now some people obtain the same transmitted signal by bypassing the preemphasis (like using the high speed data input on some ham radios).

afsk6.jpg


Above is a spectrum capture of a few signals. None are actually good, the first is low in deviation and will work ok, but the later ones are the result of classic overdrive of a LMR mic input.

You could experiment by deliberately overdriving a transmitter, monitoring spectrum with another receiver, and observing whether the digi decodes the packets successfully... and learn about its possible limitations.

Owen
VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK7HH » Thu Jun 19, 2014 11:20 am

VK2OMD wrote:Of course, overdriving the limiter creates a whole bunch of intermod products (see 3 above) which may degrade the DSP demodulator.

Now some people obtain the same transmitted signal by bypassing the preemphasis (like using the high speed data input on some ham radios).

Owen


Yes and sometimes (particularly with narrow band radios) the maximum deviation is likely 2.5-3KHz. Interface your APRS radio onto one, trying to set the deviation on the high tone for around 3.2KHz or so doesn't work. So the TNC level is wound up to max, causing the clipper to come in and all sorts of issues arise.

There are quite a few old digi's with the demodulator chips you speak about Owen around. The TNC-X is pretty forgiving (uses a CML MX614 Bell 202 modem chip). I feed that with audio directly from the discriminator bypassing any de-emphasis. It seems to work. VK7RTC is configured that way and it has heard plenty of different stations at distance (including some overdriven) - http://aprs.fi/info/a/VK7RTC.

I set the audio level out to be 3.2KHz for the high tone and let the low tone fall wherever it wants to. I did a mod to the PRM80 from 7DB on the high level audio input which makes the premphasis more closely to 6dB per octave, it's about 5.5 or so from memory. I think the low tone sits at about 1.8KHz deviated or so. In any case the proof is an on air test - again if you look at the Stations which heard VK7RTC directly on radio tab for last month VK1RGI-1 heard it on a Sporadic E opening.

*EDIT* I've also been working on using my HP8924c test set and the external mod input to it to feed in properly adjusted levels and test the decode rates at low signal levels. I've managed to get good decode rates down to 1uV (-107dBm) of signal. This method is good because little adjustments can be made. Argent I think provide a CD with test files for this too.
VK7HH
Forum Diehard
 
Posts: 232
Joined: Thu Sep 08, 2005 11:18 pm
Location: Southern Tasmania

Re: APRS

Postby VK2OMD » Thu Jun 19, 2014 12:08 pm

The radios are designed to use pre-emphasis and de-emphasis.

AFSK is intended to pass through the ordinary audio path of the transceiver at both ends.

Over the years I have heard all sorts of magic recipes for making AFSK work, but none work as well as appropriate drive and a bulk standard audio chain.

If you try to test this stuff with a test set that does not include pre-emphasis, you are wasting your time as the results do not apply to standard radios.

123.png


It is not hard to find examples of bad setup. Above is one I just captured here, there is almost no traffic at the moment so I guess that is the digi beacon (very short packet every few minutes). Note the content below 1000Hz... IMD products. Look at the comb created during the (slightly excessive) sync period, energy should not extend below 1000 or above 2400Hz.

In my experience, digis commonly set a bad example to follow... and naturally end users adjust their modulation to sound as loud as these overdriven things.

If you have a deviation meter, on a radio with properly functioning pre-emphais, adjust the DEVIATION to the LESSER of 3kHz at 220Hz input (the high tone) or clipping. It MUST NOT clip. Checkout the purpose of MIC GAIN and DEVIATION controls on radios with both, they aren't simply interchangeable.

If the 2200Hz tone is 3kHz devn, the 1200Hz tone should be 1.6kHz devn if pre-emphasis is correct.

Taking pre-emphasis off the transmitter sets another bad example and can be expected to compromise weak signal performance on a receiver with deemphasis.

Owen
VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK7HH » Thu Jun 19, 2014 12:22 pm

VK2OMD wrote:The radios are designed to use pre-emphasis and de-emphasis.

AFSK is intended to pass through the ordinary audio path of the transceiver at both ends.



Correct. I've just found that running un-demphasized audio into the TNC seems to yield better results with incorrectly adjusted transmitters.


If you try to test this stuff with a test set that does not include pre-emphasis, you are wasting your time as the results do not apply to standard radios.

My test set does have switchable pre and de-emphasis.


It is not hard to find examples of bad setup.


You can pick them on air.

In my experience, digis commonly set a bad example to follow... and naturally end users adjust their modulation to sound as loud as these overdriven things.


Depends who set them up in the first place.

Usually I make sure that the radio does not clip at 2200Hz. I can check this on the scope in the test set. You see the peaks flatten out at the top when over driving it and clipping occurs.
If it does clip, turn the level down from the TNC until it doesn't, or fix the radio's deviation so it can handle it (usually 5k) then readjust the level from the TNC to 3 odd KHz for the high tone.

If the 200Hz tone is 3kHz devn, the 1200Hz tone should be 1.6kHz devn if pre-emphasis is correct.


Nearly correct.... 2200Hz not 200Hz. Not all radios have perfect pre-emphasis curves though.

Granted most users don't have access to test equipment, which is probably the main reason behind incorrectly setup systems.

http://www.febo.com/packet/layer-one/transmit.html
http://www.febo.com/packet/layer-one/receive.html
VK7HH
Forum Diehard
 
Posts: 232
Joined: Thu Sep 08, 2005 11:18 pm
Location: Southern Tasmania

Re: APRS

Postby VK2OMD » Thu Jun 19, 2014 12:30 pm

VK7HA wrote:...
Nearly correct.... 2200Hz not 200Hz. Not all radios have perfect pre-emphasis curves though.

Granted most users don't have access to test equipment, which is probably the main reason behind incorrectly setup systems.



You captured that in about 2s before I corrected it!

You don't need expensive equipment, an SSB receiver can be used to calibrate an FM modulator if you understand FM and know what you are doing.

Yes, radios are designed to have imperfect pre-emphais... it is one of the ways that manufaturers cheat on specs by tailoring off the high end end to end so that their sensitivity specs look a dB or two better.

If you understand FM and PM, you know that what characterises PM is that phase deviation (known as the modulation index in FM) is proportional to modulator input voltage (irrespective of modulating frequency), and when you feed modulation that is pre-emphasised 6dB/octave of the entire voice band to a 'FM modulator', you actually have PM.

The fact that land mobiles have frequency deviation proportional to modulating frequency across the voice band says they are actually PM... and of course the much claimed FM noise triangle does not apply.

Owen
VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK2OMD » Thu Jun 19, 2014 1:35 pm

VK2OMD wrote:You don't need expensive equipment, an SSB receiver can be used to calibrate an FM modulator if you understand FM and know what you are doing.


http://youtu.be/V9hccM7557A

VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK7DB » Thu Jun 19, 2014 4:32 pm

VK7DB
Frequent Poster
 
Posts: 144
Joined: Fri Mar 18, 2011 12:21 pm
Location: Stowport, Tasmania

Re: APRS

Postby VK7HH » Thu Jun 19, 2014 6:07 pm

My last post down the bottom Dion :)
VK7HH
Forum Diehard
 
Posts: 232
Joined: Thu Sep 08, 2005 11:18 pm
Location: Southern Tasmania

Re: APRS

Postby VK2XSO » Thu Jun 19, 2014 8:18 pm

VK7HA wrote:That's correct Owen, but that is out of our (digipeater owners) control. Some users have the "all knobs to the right" philosophy. But from what 2XSO has described is an issue I had with that same unit with a correctly driven APRS transmitter. Another issue with those units would be after digipeating 15 mins or so, it would lock up and stop receiving.


I'd be happy just to understand why the OT3m doesn't work where other TNC's in the same place work fine. Including OT2, Tiny2, Foxdigi, and KPC3.
All of these other TNC's working fine points the finger at the OT3m. There aren't a lot of knobs to set right on a digi. Volume and squelch and both of these have a lot of tolerance even on the OT3m when running with other OT's.

I use a foxtrak for testing since it's cheap and common, but that's the point. While it's deviation is set up correctly, I haven't checked it recently. I doubt anything has changed.
They work with every digipeater in the state that isn't an OT3m.

I can't remember where I was up to with testing. I'll have to get back onto it. I'll probably start from scratch with an off air test of the tracker audio fed directly into the OT3m and see if it can
decode the audio. I do remember on air bench tests result in failure. I've brought two of the OT3m digis in today. They weren't doing any good on air.
There's a lot more variables to play with on the trackers but I don't remember any improvements from adjusting them in the past.

With a packet TNC, I'd turn on the passall command or something similar so that I could at least see what the errors are on incoming packets.

I have also noticed lockups with both the OT2 and OT3m. I've not had much luck with the otwincfg software. I've always used the shell to configure the TNC's.
Other guys in the club use the otwincfg program. We'll be loading the reset scripts as we get the chance.
I've been looking at loading the scripts remotely, but I don't think it can be done.
Trash - vk2xso
Yes I am an agent of Satan, but my duties are largely ceremonial.
User avatar
VK2XSO
Forum Diehard
 
Posts: 665
Joined: Mon Jun 18, 2007 2:50 pm
Location: Parramatta

Re: APRS

Postby VK7DB » Fri Jun 20, 2014 1:58 pm

I've had good success with the TNC-X and Xdigi combination. I have a TNC-Pi with APRX running a tri-band digi, I'm keen to get one of those on a mountain and do some tests with 6m APRS.. Either gate 6m -> 2m or 6m -> APRSIS if we get the net on the mountain.
VK7DB
Frequent Poster
 
Posts: 144
Joined: Fri Mar 18, 2011 12:21 pm
Location: Stowport, Tasmania

Re: APRS

Postby VK7DB » Fri Jun 20, 2014 2:03 pm

I also think a Dire Wolf multi-port digi on an Odroid U3 would be worth a try too..
Last edited by VK7DB on Fri Jun 20, 2014 2:20 pm, edited 1 time in total.
VK7DB
Frequent Poster
 
Posts: 144
Joined: Fri Mar 18, 2011 12:21 pm
Location: Stowport, Tasmania

Re: APRS

Postby VK2OMD » Fri Jun 20, 2014 2:15 pm

VK7HA wrote:The TNC-X is pretty forgiving (uses a CML MX614 Bell 202 modem chip). I feed that with audio directly from the discriminator bypassing any de-emphasis. It seems to work. VK7RTC is configured that way and it has heard plenty of different stations at distance (including some overdriven)


By bypassing de-emphasis, you are putting a 5dB slope or twist on the modem passband if the transmitter properly used pre-emphasis.

Have a look at the MX614 modem specs, they specify a maximum twist of 6dB... and you have used up 83% of its twist tolerance (if someone used one to demodulate your sigs) for an 'improvement'???

If it did improve things on an objective test, it hints the root cause is somewhere else.

Owen
VK2OMD
Forum Diehard
 
Posts: 1097
Joined: Tue Oct 10, 2006 8:34 am

Re: APRS

Postby VK2JDS » Mon Jun 23, 2014 6:35 pm

Just a note about digi's
I run a TT4 and a IC22S on a mountain at 950m, solar powered, with a long diamond vertical on a 20 foot mast.
The rx audio is taken off the hot side of the volume pot.
It used to crash every few days to a week, for years, and I had to drive up there and depower/repower it. The yellow carrier detect led would be locked on.
Since i asked Kim vk2asy to paint the digi on aprs.fi and I turned off the beaconing its been working without a crash now for several months.
Something to think about to improve reliability on others.
The unit idents itself with packets going through it anyway.
73 from Dave
User avatar
VK2JDS
Forum Diehard
 
Posts: 435
Joined: Wed Feb 15, 2006 8:17 am
Location: qf46pv nsw central tablelands

Re: APRS

Postby VK2FAK » Tue Jul 22, 2014 8:17 pm

Hi all...

I have been playing with the APRS and doing some reading ....not an easy thing to go through...

I was wondering with using the standalone Direwolf software is it possible to show an IGate on the map....or is one of the additional add-on programs required...?


John
Regards

John Vk2FAK
VK2FAK
Forum Diehard
 
Posts: 972
Joined: Thu Apr 15, 2010 6:28 pm

Re: APRS

Postby VK4CRO » Wed Jul 23, 2014 7:14 am

hi, I just acquired a pk88and want to set up for aprs .
What free software available for aprs ?
ciao Ron
Cheers
Ron cro
VK4CRO
Frequent Poster
 
Posts: 105
Joined: Wed Aug 20, 2008 8:26 am
Location: 282 Boston Rd.Belmont ,

Re: APRS

Postby VK2FAK » Wed Jul 23, 2014 11:37 am

Hi all....

I downloaded the latest version of Direwolf that has the option to put yourself on the map with a beacon to the Igate server only.....

But has anyone noticed that the new version has quite a bit more cpu usage....?

P.S.. fixed the problem...

John
Regards

John Vk2FAK
VK2FAK
Forum Diehard
 
Posts: 972
Joined: Thu Apr 15, 2010 6:28 pm

Re: APRS

Postby VK2XSO » Fri Aug 22, 2014 7:18 pm

VK4CRO wrote:hi, I just acquired a pk88and want to set up for aprs .
What free software available for aprs ?
ciao Ron


It's not a TNC that has any APRS support, so they're kind of useless as a stand alone unit.
BUT
There are always things that you can do !

The most obvious is just to run the TNC in kiss mode on an igate or as a digi if you're on the edge of the network.
As kiss modems they're very useful.

It is possible to turn one into a tracker quite easily too.
The simplest thing is just to set them up with some specific settings, put them into converse mode and then connect a GPS directly too them.
This will work and the APRS network can handle raw NMEA data, but it's a bit clumsy and if the modem power drops, you have to connect a PC to it and put it back in converse mode.
I think there might be a setting that automatically puts the PK88 into converse mode at power up.

Another option is to use an external PIC or arduino etc to control the TNC and compress the GPS data.
When I was first shown APRS, rather than go out and buy some sort of tracker, I read the APRS white paper and programmed up a PIC to do exactly this on a PK88.
Actually, I think one of my friends might still be running it. I was just looking for the source code, but I think I've lost it.
Trash - vk2xso
Yes I am an agent of Satan, but my duties are largely ceremonial.
User avatar
VK2XSO
Forum Diehard
 
Posts: 665
Joined: Mon Jun 18, 2007 2:50 pm
Location: Parramatta

Previous

Return to APRS

Who is online

Users browsing this forum: No registered users and 1 guest

cron