Open source hardware mods?

[LoyalServant]

Anyone doing anything of this sort?

I did some projects 20 years ago and over the last few months I have brought those back to life.
Just wondering if there are any hardware hackers out there.

I redesigned my 20 year old PLL controller on an FPGA and added an OLED display to replace the channel display, for example. I did the same thing 20 years ago to an old Midland I had with an 02A PLL.
Recently re-wrote it in C for an AVR (like an Arduino)

Will make a video about it... and thinking about putting the code out there after I finish tweaking it for an Arduino - that's what I am doing now to try to make it easier on hobbyists.
It's also a challenge.. make the code small enough to use 2k of sram and 31k of flash
I have flip-flopped on it a bunch.. trying to do it on various MCUs and FPGAs...
Was running it on a Spartan-III then a Cyclone II and then an ARM core MCU like the ATSAM3X

Had a lot of fun doing this project.. and an old radio with an OLED channel display will look cool too

 

[RFI-EMI-GUY]

Actually I am interested in controlling the channel or frequency selection of 40 channel CB's using an Arduino or similar. The newer radios pose a bigger challenge than the old BCD switch type. Still the ability to select 1 of 40 channels "on the fly" so to speak is what I am interested in. Think FHSS for the poor man!

 

[LoyalServant]

It's gone wayyyyyyy beyond an arduino but it can still run on one.
It's actually somewhat trivial.

(I use AVR Studio when doing Atmel MCUs or MPLab for Microchip MCUs)

This is where I am at now.... excuse the mess... engineers with clean work surfaces are not working
http://i.imgur.com/9DeXGzA.jpg

What I have working... PLLs MB8719, D858, MC145106, MC145170
It has several roger beeps.. frequency counter with 1Hz resolution and can distinguish between
LSB, USB, AM, FM, etc.
Some other stuff in there as well.

Been a fun project... the first version is in that Cobra 48 still but that never made it past being on a protoboard.
Code for that is rummaging around someplace.

Counters are easier on a pic than an AVR but I wanted to try it.
It also needs a TCXO for the counter because the regular oscillator is not quite good enough.
I got it working accurately but only with clever code.
It also requires a prescaler where a pic16 or the like can just count straight off a conditioned signal.

To do serial control... I am not sure if the Arduino Uno would be the best choice.
I would go to a Arduino Mega 2560 because that AVR has a lot more space.
The LCD graphics totally kill space.

I have a smaller version of those LCDs already installed in several radios.
They fit where the old 7 segment display went.

(Sorry, but in this day and age using 7 segment displays whether 2 or 3 digits is a complete and utter engineering FAIL. They cost more than an LCD now and are so limited......)

My field trials using the LCD in a Cobra 148 and a SS 3900 have been very successful.

Your timing is ironic.... I am about to send gerbers off to get a run of test boards made.

 

[RFI-EMI-GUY]

My idea is for a frequency hopping add on board. Bendix King had a VHF radio that would hop random channel sequence, and used a collection tone to synchronize at the beginning of each transmission. Initially I was thinking of using GPS timecode and 1PPS to synch, but the Bendix King method is simpler.

Radios operating in the AM mode could be eavesdropped using a wide band AM detector, but FM and SSB radios (provided PLL settling time is short) would go undetected. Yes you could set up 40 receivers and hear the traffic.

The trick is finding a popular contemporary radio design (interface standard) that would make the idea attractive.

 

[LoyalServant]

An interesting idea.....

The PLL lock time on these radios is fairly instant. That all occurs probably within a millisecond so that's not the problem. I have heard of the method you describe that was used by Bendix King but I am not familiar with it intimately.

If you wanted to obscure SSB comms like that the only issue I see is that unless there was some shared
information between operators anyone with such a board... or like you said 40 receivers could still listen in.

Seems like this could be made to be compatible regardless of radio.
The MCU is going to have to know what PLL you have in order to program it.
Those capabilities could be relayed to all parties to have an election of sorts.
If someone had a straight 40 channel radio then your choices are limited once you remove the channels
you would not want to hop to. (9,19, etc)

That is, if you wanted to deal with more than 2 parties.

I like it... because it's hard

 

[RFI-EMI-GUY]

There are military radios called SINCGARS that use HAVE QUICK technique. Those use a high stability clock to synch the radios. More importantly they use a coding scheme to seed the pseudorandom sequence with NET information so that different groups can share the channels without interfering with each other. My thought is to have a selectable NET code for different groups of radios so that they hop a sequence orthogonally from others.

So instead of telling the net to meet on 27.865 or some other illegal channel, they would meet on "NET 2005" which would be a subset of the 40 (0r 38) channel sequence.

Theoretically you could operate on all 40 channels without interfering with channels 9/19. But I could see where if dozens of different networks were operating, this could start to be a problem.

 

[RFI-EMI-GUY]

The radio interface would require a ribbon cable with channel information, receive and transmit audio (for the synch tone sequence) and likely a PTT and possible COR indicate. Modification of the radio might entail disabling or delaying lock detect control of receive audio mute and transmit PTT inhibit functions. And the front panel channel selection would have to be overridden.

I have been pouring over The CB PLL Data Book Paperback – 1995 by Lou Franklin but realize that is kind of dated and the design of newer radios have superseded that publication. I am an LMR guy, so my CB hands on is a bit rusty. Personally I was going to import some FM CB's from Europe (that share US bandplan) and cobble something up. The question would be what would be marketable either here in US or UK?

Would FCC become annoyed if "out banders" suddenly started confining themselves to SSB in the legal 40 channels using an obscure military "anti-jam" technique?

Would UK buyers be inclined to spend many Euros for an add on that would give them a sense of privacy for their congested CB band?

Just food for thought.

 

[LoyalServant]

I am not aware of any rule that says you cannot use any form of anti jamming technique.
It's not remote control if you are operating the radio and it's control apparatus...

If you were operating in the 40 channels and transmitting small data bursts .. is that breaking any rules tho?

Why would the lock detect need to be disabled?
I would think that in this case of trying to be as close to legit as possible the lock detect becomes all that more important so we don't send a divider sequence that it cannot lock and then try to transmit with a free running VCO. (and splatter all over the place)

In fact it's not a terrible idea for my MCU mod to monitor LD and warn the user if the PLL has unlocked.

Now.. marketability.... because that's a touchy subject

You have to be careful because the so called 'custom radio' crowd tends to go scorched earth.
What I am doing will never see the light of day because I already had a scorched earth campaign waged against me. That's why I started looking at a public release for the code/design because any commercial viability was essentially destroyed by the golden screwdrivers and co called 'custom radio' community.

I have an inbox full of hate mail

Now, this project I am aware of no other so it's possible that it's accepted.
Me on the other hand.... no way will they allow a refresh to the UI of a 30 year old radio when they can
sell inferior ones for twice the price. (literally)

 

[RFI-EMI-GUY]

My comments below in (parens)

I am not aware of any rule that says you cannot use any form of anti jamming technique.
It's not remote control if you are operating the radio and it's control apparatus...

(I think we are good here. If you compare type acceptance certifications for most products against what they market, you will find a lot of variance between the radio initially certified and later "enhancements" by the manufacturer. In the LMR field, adding a function board has never to my recollection been a call for re-certification.)


If you were operating in the 40 channels and transmitting small data bursts .. is that breaking any rules tho?

(As long as you are transmitting in an approved emission designator you are OK. The only bursts would be a collection tone at the beginning of transmission, like a roger beep but with the purpose of synching the receiver)

Why would the lock detect need to be disabled?
I would think that in this case of trying to be as close to legit as possible the lock detect becomes all that more important so we don't send a divider sequence that it cannot lock and then try to transmit with a free running VCO. (and splatter all over the place)

(The only problem would be if the time constant was such that the transmitter dekeys or the receiver mutes during the hopping sequence. Try this on a radio, when you manually switch channels while transmitting does the TX turn off momentarily? In receive does in mute for 100 msec while switching channels? If the PLL locks up in 10 milliseconds, you can have adjust the hop rate and dwell time to be pretty fast, you don't want LD circuit unnecessarily slowing you down.)


In fact it's not a terrible idea for my MCU mod to monitor LD and warn the user if the PLL has unlocked.

(Just adds a wire and some code)

Now.. marketability.... because that's a touchy subject

(It always is)

You have to be careful because the so called 'custom radio' crowd tends to go scorched earth.

What I am doing will never see the light of day because I already had a scorched earth campaign waged against me. That's why I started looking at a public release for the code/design because any commercial viability was essentially destroyed by the golden screwdrivers and co called 'custom radio' community.

I have an inbox full of hate mail

Now, this project I am aware of no other so it's possible that it's accepted.
Me on the other hand.... no way will they allow a refresh to the UI of a 30 year old radio when they can
sell inferior ones for twice the price. (literally)

(I think the answer might be an add on board that works with popular contemporary radio models, not necessarily "export models", but something Midland or Radio Shack sell, an add-on module that the average "hacker" can install. The user interface could be a small module that velcros onto the radio and connects via ribbon connector. What I am thinking is the module when switched on, would blank the 40 channel selector and would ask the user to input into its memory one of several NET codes. ) Ideally it should work with AM models, AM/SSB models and AM/FM models (EU Market)

 

[prcguy]

SINCGARS and HAVE QUICK are two completely different systems.

Why do you have to hop specific channels when you can hop within a channel? I have some Harris military rigs with both wide and narrow frequency hopping. The wide can cover the whole HF band but narrow hops within a few KHz of a center frequency and its very effective.

You could have a controller diddle a few padding capacitors across the reference osc crystal or if the radio synthesizer has 100Hz or 1KHz steps you can hop in those increments.
prcguy

There are military radios called SINCGARS that use HAVE QUICK technique. Those use a high stability clock to synch the radios. More importantly they use a coding scheme to seed the pseudorandom sequence with NET information so that different groups can share the channels without interfering with each other. My thought is to have a selectable NET code for different groups of radios so that they hop a sequence orthogonally from others.

So instead of telling the net to meet on 27.865 or some other illegal channel, they would meet on "NET 2005" which would be a subset of the 40 (0r 38) channel sequence.

Theoretically you could operate on all 40 channels without interfering with channels 9/19. But I could see where if dozens of different networks were operating, this could start to be a problem.

 

[RFI-EMI-GUY]

I am not a SINCGARS expert, I have read a bit on HAVE QUICK. I see now that they are different. The HAVE QUICK uses a clock Time of Day, TOD and a Word of Day (WOD) and NET code.

I had a wild idea to use cheap GPS modules to synch FHSS with some UHF LMR radios or CB radios. Further research leads me to believe that using a collection tone (BK Radio, ARRL Spread Spectrum Experiments) would be sufficient.


I recall seeing the narrow band frequency hopping software on a Spanish site. I thought they were using for data transmission. Hopping the channel steps is fine, reference oscillator padding caps would require some fiddling.

The CB band is channelized, so you are allocated to operate on a "channel" center, and the emissions bandwidth and frequency stability is constrained. The thought was to directly manipulate the channel selection and hop most or all of the 40 channels.

THOUGHT EXPERIMENT: You and I share a table of 1 to 40 random channel assignments to be utilized every 10 minutes while holding a net on the CB radio. Every 10 minutes we QSY. Nothing illegal. Now speed this up 6000 times, by hopping every 100 milliseconds. Still legal? Now let a computer do the work. Still legal?

By the way, DOD contends that this is not encryption, it is anti-jam technology. I agree, because with 40 receivers I can hear the network from 1 of 40 speakers. But, if ten networks are going??

On some really old radios, a BCD channel selector could be emulated in the MCU and use a look up table of pseudorandom channel assignments. Some newer radios use up down switches and knobs which is a challenge. The goal is a KISS approach that anyone smart enough to solder a few wires could install.

There might be a a particular current model radio chassis that would be easily hacked, if so the approach should exploit that advantage. It would be shame to have a great mod that works, but only on old dinosaur radios or something otherwise out of production. If it is a popular model chassis that many millions were produced, that has an advantage as well.

SINCGARS and HAVE QUICK are two completely different systems.

Why do you have to hop specific channels when you can hop within a channel? I have some Harris military rigs with both wide and narrow frequency hopping. The wide can cover the whole HF band but narrow hops within a few KHz of a center frequency and its very effective.

You could have a controller diddle a few padding capacitors across the reference osc crystal or if the radio synthesizer has 100Hz or 1KHz steps you can hop in those increments.
prcguy

 

[LoyalServant]

That would be various Cobra radios... 25,29,148, etc along with their Uniden branded cousins.
All of those being AM except for the 148/Grant XL which are AM/SSB radios.

All to my knowledge were produced in the millions and use binary encoders.
In fact my controller would interface with any of them in a matter of a few minutes.. just by writing the code to talk to that PLL.
The 25/29, etc have rom locked PLLs but were talking about staying in the 40 channels (minus 9,19, etc)
so it is not a problem.

I don't think MCU controlled radios are going to be in the cards and the issue there is that MCUs are cheap, and momentary encoders are even cheaper than the old 40 position n bit binary/BCD encoders they used.

But... Cobra 25,29,149 and their cousins... are mainstream radios.
Readily available and are actually really good radios to boot.

Now, as far as the lock detect... your right... they may very well mute.
I will throw a scope on it and do a single shot to see if it changes state when the PLL is programmed.

 

[RFI-EMI-GUY]

Check to see if the TX is inhibited when changing channel as well. The fix is probably simple, adding a cap to delay LD from inhibiting the TX.

That would be various Cobra radios... 25,29,148, etc along with their Uniden branded cousins.
All of those being AM except for the 148/Grant XL which are AM/SSB radios.

All to my knowledge were produced in the millions and use binary encoders.
In fact my controller would interface with any of them in a matter of a few minutes.. just by writing the code to talk to that PLL.
The 25/29, etc have rom locked PLLs but were talking about staying in the 40 channels (minus 9,19, etc)
so it is not a problem.

I don't think MCU controlled radios are going to be in the cards and the issue there is that MCUs are cheap, and momentary encoders are even cheaper than the old 40 position n bit binary/BCD encoders they used.

But... Cobra 25,29,149 and their cousins... are mainstream radios.
Readily available and are actually really good radios to boot.

Now, as far as the lock detect... your right... they may very well mute.
I will throw a scope on it and do a single shot to see if it changes state when the PLL is programmed.

 

[LoyalServant]

Changing channels does fire LD.... very quickly but it's fired...
It never hits zero tho.... which is why I find this very strange but this is what I got....

 

[JayMojave]

Hello LS: Yes there are a few hardware hackers out there:

Adding in the computer controlled channel circuit via a program controlled function is out there. A older CB radio such as the Tram D201A uses a specially made channel switch to add all the different crystals to make the Receive LO and Transmit frequencies. Problem is these special (many switch function) switches after time go intermittent basically disabling the radio.

The fix for me has been a short piece of rubber hose lodged between the channel select knob and lower lip of the radio, applying a small amount of pressure on the channel select switch.

But a new switching circuit will be needed, or better yet a modified PLL Circuit installed, if you want to keep this old great radio on line. I have seen on U-Tube such modifications and are looking into them.

QST / QEX Magazine has had several articles on this, giving the designers circuit and web site. QEX has also had articles on modified "Older Receivers with big Dials and S-Meters" with these newer DDS circuits, with highly modified receiver hardware with much better IF Filtering and noise reduction with DSP and such. Very Impressive.

Impressive scope measurement display, good going.

Jay in the Mojave

 

[LoyalServant]

Hello LS: Yes there are a few hardware hackers out there:

Adding in the computer controlled channel circuit via a program controlled function is out there. A older CB radio such as the Tram D201A uses a specially made channel switch to add all the different crystals to make the Receive LO and Transmit frequencies. Problem is these special (many switch function) switches after time go intermittent basically disabling the radio.

The fix for me has been a short piece of rubber hose lodged between the channel select knob and lower lip of the radio, applying a small amount of pressure on the channel select switch.

But a new switching circuit will be needed, or better yet a modified PLL Circuit installed, if you want to keep this old great radio on line. I have seen on U-Tube such modifications and are looking into them.

QST / QEX Magazine has had several articles on this, giving the designers circuit and web site. QEX has also had articles on modified "Older Receivers with big Dials and S-Meters" with these newer DDS circuits, with highly modified receiver hardware with much better IF Filtering and noise reduction with DSP and such. Very Impressive.

Impressive scope measurement display, good going.

Jay in the Mojave

The display is actually a function of my Rigol DS2202a DSO.
DS2202A 200 MHz, 2 Channel Digital Oscilloscope | Rigol - Beyond Measure

I have been experimenting with the MC154170 PLL as a replacement for some rom locked PLLs or even a case like your Tram.

Call me crazy but interfacing some modern controls and old radio gear while trying to maintain the stock appearance has been some of my coolest projects.
The stuff I do as a consultant isn't anywhere near as cool.... or fun.

 

[prcguy]

I don't believe frequency hopping will be legal on CB because it would take on a different FCC modulation designator which is not allowed.
prcguy

I am not a SINCGARS expert, I have read a bit on HAVE QUICK. I see now that they are different. The HAVE QUICK uses a clock Time of Day, TOD and a Word of Day (WOD) and NET code.

I had a wild idea to use cheap GPS modules to synch FHSS with some UHF LMR radios or CB radios. Further research leads me to believe that using a collection tone (BK Radio, ARRL Spread Spectrum Experiments) would be sufficient.


I recall seeing the narrow band frequency hopping software on a Spanish site. I thought they were using for data transmission. Hopping the channel steps is fine, reference oscillator padding caps would require some fiddling.

The CB band is channelized, so you are allocated to operate on a "channel" center, and the emissions bandwidth and frequency stability is constrained. The thought was to directly manipulate the channel selection and hop most or all of the 40 channels.

THOUGHT EXPERIMENT: You and I share a table of 1 to 40 random channel assignments to be utilized every 10 minutes while holding a net on the CB radio. Every 10 minutes we QSY. Nothing illegal. Now speed this up 6000 times, by hopping every 100 milliseconds. Still legal? Now let a computer do the work. Still legal?

By the way, DOD contends that this is not encryption, it is anti-jam technology. I agree, because with 40 receivers I can hear the network from 1 of 40 speakers. But, if ten networks are going??

On some really old radios, a BCD channel selector could be emulated in the MCU and use a look up table of pseudorandom channel assignments. Some newer radios use up down switches and knobs which is a challenge. The goal is a KISS approach that anyone smart enough to solder a few wires could install.

There might be a a particular current model radio chassis that would be easily hacked, if so the approach should exploit that advantage. It would be shame to have a great mod that works, but only on old dinosaur radios or something otherwise out of production. If it is a popular model chassis that many millions were produced, that has an advantage as well.

 

[RFI-EMI-GUY]

I believe the emission designator would remain the same. in Frequency Hopping Spread Spectrum, the bandwidth of the base band modulation is what counts, not the bandwidth of the hopping space. You would be correct for Direct Sequence Spread Spectrum where a wide bandwidth spreading code is "ANDed" with the modulation code.

The FCC may be persuaded that this is a more efficient use of spectrum.

 

[RFI-EMI-GUY]

I am guessing from the scope that the LD voltage is high normally at about 3.9 V and drops to about 1.8V when out of lock. Maybe it is buffered and is biasing a switch downstream? Most importantly, does the audio mute circuit activate when switching channels? And does the transmitter become inhibited when switching channels?

Changing channels does fire LD.... very quickly but it's fired...
It never hits zero tho.... which is why I find this very strange but this is what I got....

 

[RFI-EMI-GUY]

My old Lafayette COMSTAT 25A had such a crystal mixing scheme. And worked well until a crystal died taking out about 6 channels. I had to replace it twice. Great old rig.

I have been thinking that a nice low noise FM stereo broadcast tuner could be made using such a mixing scheme.It would take about 25 crystals to make it work on 200 KHz channel spacing.

Your old tram could benefit from an outboard PLL, or better yet, you could build a switching board and some logic to convert a standard rotary switch to the sequence required to step the mixing crystals. I am thinking some small reed relays and some TTL chips to drive them from a BCD rotary switch. Of course a PIC could be programmed to do this switching nicely.

Hello LS: Yes there are a few hardware hackers out there:

Adding in the computer controlled channel circuit via a program controlled function is out there. A older CB radio such as the Tram D201A uses a specially made channel switch to add all the different crystals to make the Receive LO and Transmit frequencies. Problem is these special (many switch function) switches after time go intermittent basically disabling the radio.

The fix for me has been a short piece of rubber hose lodged between the channel select knob and lower lip of the radio, applying a small amount of pressure on the channel select switch.

But a new switching circuit will be needed, or better yet a modified PLL Circuit installed, if you want to keep this old great radio on line. I have seen on U-Tube such modifications and are looking into them.

QST / QEX Magazine has had several articles on this, giving the designers circuit and web site. QEX has also had articles on modified "Older Receivers with big Dials and S-Meters" with these newer DDS circuits, with highly modified receiver hardware with much better IF Filtering and noise reduction with DSP and such. Very Impressive.

Impressive scope measurement display, good going.

Jay in the Mojave