“Those who hear not the music, think the dancer mad”
Jaws Mk2

APRS Transmitter

APRS TX - "Specifications and Objectives"

Thought some more about building an APRS TX to replace the £200 Kenwood HT stolen with Simon's Enfield.

Suitable for Simon's Side Car APRS
144.800 single freq TX - 36.200 2x and 2x multiple, like Fredbox? But, somehow avoid using a £17 custom crystal
FM only - Varactor / LED! All a bit unknown.
Could a PIC directly mod it?
Compact and rugged - Diecast box PA to make 25 watts Use my junk box VHF power transistor, Very low duty cycle 3 secs every 2 mins - Reduces PSU and heat issues
Use a TinyTrak for the APRS functions, although as such a reduced level of function is required, perhaps a PIC and PD code could be used. Temperature sensor ? GPS - Serial

Use as few "special" parts as practicable
Vaguely reproducible design(!)
Simulate, fully document and completely understand the design.


Had fun today with oscillators and frequency doublers. Got the 36MHZ crystal perculating nicely at 72Mhz (Fredbox style). But could I get the next doubler to work? Oh no. The 144 odd target remained - well a target. Yes, despite much fiddling with the tuned circuit it stayed at 72. So, I thought it was time to dig out a GDO and see what was happening. I had, I thought, a nice shiney LOWE GDO. Plugged everything in and swept about. Nowt. Even checked the battery and read the instructions. Strange there is no power to the oscillator.. Infact where is the oscillator. Infact... Ah... it is a frequency absorbtion meter. Ah. that would explain it. Took a while to get there!

Of the other GDO's I have one has no coils and the other has no coils above 30 Mhz. Back to the drawing board. There may well be a harder way to 144.800 (APRS) it is just that I have yet to figure it out!  But, I have learnt alot.

Oscillators - GM1SXX, G4BYE, G8KDF and G4KGP - Thank you very much for your thoughts! I am still struggling on with the Fredbox design. Still stuck at 72Mhz. Nice clean output, but nothing much out of the next stage at all. I think the key is might be the driving and how well the collector load is working. I am now revisiting Class A amps and looking closely at the tank circuit, as people suggested. Plus, I now have a Dip Meter! I think I will end getting it all working, then abandon it to build a tripled version based on a 16 Mhz crystal. I can still use the 12Mhz crystals for grinding / penning practice. Should be fun.

I was considering "Blogging" my meanderings, however progress is slow enough as it is..  I might be tempted to try and build a 30m APRS TX at some point for Simon's up and coming European trip.

From the Dunstable rally I purchased some 2SC-1947 transistors, which might form the basis of a 2.5 watt PA and some Zener Diodes.

“Results! I have gotten a lot of results. I know several thousand things that won't work.”

Still not got my 144 Mhz oscillator running. Perhaps doublers are just against the laws of nature? I might have another look later. I did have another think about the Crystal. I am going to have a go at a little crystal grinding. Starting off with a 12Mhz clock crystal and see if I can lift it 66khz(!)  No idea if it will work, but it is a bit of hacking I have not tried. Even Tom had not tried it and Simon thought it was rather hard core. So we must be on the right track then..

I had said to Simon the project was intended to be reproducible and not use exotic components... Providing you have a scope, can select components on test and grind / pen crystals it will work fine! Oh, and you need a 6mm bolt - to wind the coils on. Tom wondered why I did not want to use the cambion coils as he has loads of them. That - would be way to easy, and besides it is a rather exotic component!

Crystal Oscillator Components
APRS Oscillator Experiment One - 144.890

09/07/10 - AOE One

APRS TX - Well, I have finally got a carrier on 144.890 - Nearest crystal I have. One of the keys was using a "gimmick" twisted wire capacitor so I could couple up the scope and see what was actually happening. GM1SXX had reminded me of the loading/coupling issues associated with measuring. Other issues included widely varying transistor Hfe - need to check those junk box parts carefully - bias that alters under load, collector voltages and a dead transistor.

“You don't want to go loading the 'Slithy Toves'. They don't like it.”

The TX strip that G4KGP built provided great help for comparing measurements and construction ideas, although judging by the state of my expermiment one board you might not think so!

Cut apart a 12Mhz crystal. As I had seen illustrated elsewere, there was no way the crystal wafer was coming out of it's mount. But, it was interesting to take it apart. It might be possible to "pen" one of the slightly higher frequency crystals down to 12.06667

12/07/10 - AOE One

Unsucessfully attempted to open a crystal case using a Dremel. A finer disc cutter and a whole lot more precision is required. Still at 144.890

19/07/10 - AOE One

Finally managed to hacksaw a crystal apart without completely wrecking it. Gripped the solid metal base in the vice jaws, then very carefully hacksawed around the case parallel to the base and few mm above it. Very carefully removed the case then, using a CD marker pen carefully "inked" part of one side of the crystal, waited for it to dry and measured the frequency after the doubler stage. After repeating this procedure a few times the frequency had dropped by 10Khz. This test crystal was way to high, but the technique certainly shows promise for a crystal closer to 12.06667

Further "ink all over" experiments resulted in the crystal frequency actually going up by 200Khz at a reduced amplitude.

          Crystal Oscillator Closeup
Crystal Closeup

01/08/10 - Crystal Penning

Sawed the case off another crystal this afternoon. Finally got the hang of it. Did some more penning experiments, quite surprising how robust they are. It was even possible to remove some of the pen with foam cleaner, which would be handy if it had been moved to far. Accidently penning the metal plates stopped oscillation.

Although the case was now a couple of mm shorter, there was still enough clearance to refit it over the crystal.

Still no closer to 144.800, but the experiments were interesting.