Weak Signal Mode APRS

Over the past few weeks I’ve been experimenting with APRS Messenger, a great PC based application that uses weak signal modes for Automatic Packet Reporting System (APRS). Traditional HF APRS uses 300 baud packet data for APRS messaging which isn’t really suited to HF communications.

APRS Messenger is capable of decoding PSK63, PSK250, GMSK250, MFSK4 and MFSK16 and then sending them to the internet (iGate operation).  These modes are much more suited to HF than 300 baud packet.

Gerard, VK2IO (Sydney) and myself (Melbourne) ran some tests between our stations on 40 metres (about 700km) using the various modes with latitude and longitude encoded.  Using PSK63, I was able to lodge quite a few location reports with Gerard while only transmitting 500mW RF out.  Of course this was not consistent and in fact there were some times when we could not communicate with 20 W, although most nights this has been fine.   Wade, VK1MIC also popped up on PSK63 with 5W from an FT817.

PSK250 was not too far behind PSK63 in performance.  A PSK63 position report takes about 12 seconds, where a PSK250 report takes about 4 seconds.  At my location I have a high level of noise and also an electric fence ticking away and these modes coped well with it.

Of more interest was constructing a light weight HF APRS station for remote area hiking.  If there were a few HF weak signal mode iGates around the country, this could be viable method of remote area APRS.
Recently, David, VK3IL and myself did some remote area hiking and David had set up a VHF packet APRS to HF APRS repeater in his car.  This worked pretty well considering the terrain we were hiking in.  See his blog for details on the hike.

An even better modulation mode for weak signal APRS is MFSK16.   I came across this interesting project from Mike, N0QBH.  This was a low power, standalone APRS tracker using MFSK16 mode.
I ordered some parts and built one up on a breadboard.  For the PA, I just used a temporary 2N3866 amplifier that was producing about 60-70mW out from the DDS drive of about 1mW on 40 metres.
The unit was calibrated and put on air.  To my surprise, I received an email from Gerard, VK2IO to tell me he was decoding the occasional packet in Sydney!  Conditions on HF have not been all that good recently either!

Here is a photo of the rather messy lash up of the beacon.  When the 5 Watt PA / LPF is finished and working, I will do a PCB and try and keep it as small as possible.  The GPS receiver, processor and DDS are all running from a 3.3V supply.  Current drain is between about 60 and 120mA at 3.3V.  The PA is running from 12V.
Mike’s software shuts down the DDS and GPS between beacons to save power.  At present I have not wired up the GPS shutdown, so it is running all the time.
With a switching regulator, the whole thing should be quite low current at 12V or 3 cell Lipo.

It’s a mess but it works!

Here is the breadboard lash up of the beacon.  It’s certainly not pretty, but it works.  The GPS gets a fix pretty quickly indoors.

Posted in Hiking, Projects, radio communications | Tagged , , , , | 3 Comments

Yaesu VX8-DR GPS Revisited

A while back I made a GPS receiver module add on for the VX8-DR so that it could be used for APRS while on SOTA hikes.
The module worked well although it was not that practical for portable operation because it had to be attached somewhere (usually rubber banded to the rear of the radio) and this made the radio difficult to carry.

After a recent SOTA expedition where we were using APRS, I thought it was time to try and repackage the GPS module.  One of the drawbacks of the previous module was that it had a small rechargeable LiPo battery which had to be charged.

The target is for a GPS receiver that plugs into the top connector of the VX8-DR and uses the power from the radio.  The unit should also be of a similar size to the original VX8-DR module.

The old GPS module had the GPS receiver and the protocol converter board in line.  The protocol converter board was moved and placed on the back of the GPS receiver.  A Supertronic PP12N hobby box was cut in half to house the module and converter as this size was almost perfect.  

Plugs for the VX8-DR are hard to come by at a reasonable price, so a USB programming lead was sourced from Ebay and the radio plug cannibalised.  The USB plug lead doesn’t bring out the power wire, so the plug was trimmed down to the point where the pins were exposed.  A U bracket was folded from Aluminium and a hole drilled in the bracket to take the plug as a tight fit.  Polyurethane glue was used to fix the plug into the bracket.

It turned out reasonably well and compact as well (compared to the old one).

Some work needs to be done to make it robust such as a bracket like the original Yaesu one that provides support to the module from the antenna mount.

The photo to the left shows a button like mark on the box.  This is where I was experimenting with milling from the inside and overheated the plastic.  Luckily I have more of these boxes so can put another one together.  A cable tie is holding the lid on for the moment.

The GPS receiver obtains a fix very quickly and works pretty well inside the house.

From this diagram,
VCC goes to Pin 2 on the accessory connector,
Ground to Pin 6,
Module RX to Pin 4 and
Module TX to Pin 5.

Showing connections to VX8-DR socket, viewed from top

Things to do when time permits:

  • Make a PCB for the Atmel protocol converter.  Use an SMD chip.
  • Make a support for the module back to the antenna mount.
  • Maybe do a 3d printed box for a better package.


2018-03-10:  Changed the interface PCB for an Arduino Pro Mini.  This is a very compact unit that saves some space.  I came across another VX8-DR GPS project from 4X1MD, and thought I would give it a try.  It worked well and gets around the issue of garbage characters to the radio screen while the GPS is acquiring signal.  This software uses a single serial port compared to the previous software that had a second software implemented port.  There is no real need for two ports as the GPS sends to the protocol converter and the converter sends to the radio.

Arduino Pro Mini

Posted in Projects, radio communications, SOTA | Tagged , | 1 Comment

Sheet Metal Bender

This metal bender project arose out of the need to often make small enclosures for various radio related projects.  Some time back, I purchased one from Ebay but it was not very good and would flex while bending even thin aluminium.

The angle iron was purchased from Bunnings hardware, it was about $20 for a 2 metre length of 40mm x 40mm x 5mm thick.  The pins for the hinge were also from a length of 10mm black iron rod sourced from Bunnings.  Unfortunately I had to buy a 2 metre length when I only required 60mm or so.

The design is quite simple and like many others that can be found on the internet.  The work clamp, base and folding part are all made from the same 40mm angle iron.  I wanted to make sure it was very rigid, so I decided on a fully welded construction.  This is a bit risky as things tend to move as they cool when welded.  The parts were clamped very tightly as they were welded and fortunately, it came out pretty well.  The same can’t be said for the quality of some of the welds though.  I probably should have practiced a bit more on some scrap to get the welding current correct before starting on the actual project.

The fixed part of the bender (rear angle iron) and pivoted part (front angle iron)

The pivoted section of the bender is at the front of the photo.  This angle iron section is about 5mm shorter in length than the fixed part at the rear.  Two short 30mm lengths of 10mm diameter rod are welded in a notched out section of the pivoted section.  The rods are centered into the bending axis.

The fixed section of angle is 350mm long and the pivoted section 345mm.

The fixed section is also notched out at each end to allow for the hinge rod.

Side supports

Each side support (upside down in the photo) are made from the same angle with a bit of angle added on to obtain a clearance height from the mounting surface.  Rather than buy a different size of angle, I just cut a flat piece out from a bit of angle and welded it on.

Note that the photo is upside down!  The side supports will need to have hole precisely drilled out for the pivot.  To precisely locate the hole, it was drilled as accurately as possible and then the whole side support moved so that it was on the bending axis.  Once located, the side support was tack welded and then finally welded.

The side support welded to the rear fixed angle of the bender

This is shown in the photo to the left.  It wasn’t too hard to keep it all clamped while the side support was welded, and thankfully nothing moved.

The next step was to mount the clamp for the material to be bent.  This is much less critical as the fixing holes are oversized so that the clamp can be adjusted before the clamp is tightened down.  10mm bolts are used to hold down the clamp angle section.

The nuts of the bolts are tack welded to the base so that they wont move when the clamp is tightened.

Handles on the pivoted bending section are simply 200mm 3/8th bolts from Bunnings.  Two nuts for each bolt are welded to the underside of the pivoted bending section.

Completed bender. It needs a rustproof and a paint!

It works pretty well.  I have tried 1.6mm aluminium and there is not noticeable flexing anywhere while it is bending.

One improvement is that the clamp section of angle has a slightly rounded edge.  For a nice tight bend, this should probably ground for a slight chisel edge, although I am not sure how to do this accurately at the moment.

First go at bending, looks fine

Another improvement is to cut some slots into the clamp section so that box sections can be made, the previous bends can be recessed into the slots in the clamp.

All of the angle iron cuts were made with a hacksaw (2 blades in total).  Originally, I set up an angle grinder in a cut off saw table and just as I started the first cut, the angle grinder burned out 😦

All painted to protect from rusting

Clamps were made by welding some 8mm rod to long bolts and then some heatshrink added to the handles.

The clamp piece leading edge was ground with an angle grinder to allow a sharp bend.  This had to be done very carefully to keep the edge straight.

Posted in Projects, radio communications | Tagged | 7 Comments

QCX QRP Portable Transceiver Build

This compact “QRP” or low power radio kit from QRP Labs in UK was too good to resist.  At $49.00 USD, it packs a lot of functionality into a small package.  The radio is based around an IQ phasing receiver for CW (morse) and an Class E efficient transmitter.

The transceiver includes built in alignment software and hardware, a WSPR beaconing mode, CW decode and internal memory keyer.  See the QRP LAbs site for the details.

The software alignment is most impressive as is the well thought out menu structure.  I chose the 20 metre band variant when ordered.

The radio looks ideal for lightweight SOTA operations.  As the kit is supplied without a case, some thought is needed to package it efficiently.  This plastic enclosure from Jaycar fits quite nicely with a little space to fit a 3 cell Lipo pack.  There is still some work to do around fitting the two pushbuttons, volume and tuning encoder.

A shaft extender will be used from the encoder shaft through the top panel.  The volume control and two push buttons will be top panel mounted and cabled back to the board via a connector.  I had to cut down the supplied board stand-offs to about 8mm long with a Dremel so that the display was the right height to sit behind the top panel.

Nice fit of QCX and battery

Top view without controls










RHS panel with BNC and paddle socket

LHS with earphone socket







A shaft extender for the rotary encoder was turned up on the lathe.  I could have probably bought one, but this one is small and light.  A bush from an old pot might be placed on the top lid of the case for a bit more support of the shaft.  It doesn’t feel to bad though.

Shaft extender

The controls have now been fitted without problems.  It was nice that there are connector pads for off board controls.

The radio looks a little off balance with placement of controls now, probably not helped by the extra space above the display.  The DC screw terminal connector had to be removed as it interfered with the side panel of the enclosure.   A barrel type 2.1mm DC bulkhead socket is fitted to the side panel for power in.

A bezel of some sort is needed for the display to protect it and improve the appearance.  For internal batteries, the 3 cell Lipo fitted ok, although I have settled on using 3 x 18500 round cells instead.  These cells take up a bit less space and are more robust.

Controls Fitted


It is still work in progress.

The RF power output had a tendency to immediately go to about 1.5 watts and then slowly rise to just under 3 watts from cold.  A quick re-ptt saw the power go about 3 watts on 14.020 MHz.  If the rig was rested for 10 seconds or more, the power would go to 1.5 watts and slowly rise again.

As suspected, it turned out to be capacitors in the low pass filter.   One by one the capacitors were replaced and the RF output behaviour checked.

There was no change when the outer 2 x 180pF (C27+C28) caps were changed, but a noticeable change when one of the inner 390pF (C25) was changed and then perfect when the other 390pF (C26) was changed.
I changed each cap with a 250V ATC ceramic chip cap.

The output power is now very stable and the supply current has dropped.  Output is 3.5W with 13.8V supply.
I took a turn off both L2 and L3, so L3 is 15T and L2 is 16T and output is about 4W.

I also experimented with the 30pF capacitor from the drains of the finals to ground.  The efficiency went up slightly without it.  I put a 30pF trimmer in its place and found best efficiency at almost no mesh of the trimmer.

1500mAh 18500 cells

This photo shows the 3 x 1500 mAh 18500 lithium cells fitted.  A balance connector is connected to the cells.

After running a WSPR test, the power output and current consumption became erratic.  Sometimes it would draw 1000mA and transmit only half a Watt and other PTTs it would draw 450mA and put out 4 Watts.  Q6 was also getting very hot.

Q6 was replaced and it completely fixed the problem.  Q6 is on the limits of its ratings.  It looks like Q6 has been replaced with an MPS751 is later versions of the kit.


I had noticed that the first PTT after power up resulted in a high current draw and low power output. This behaviour was completely repeatable after every power up.  Others on the forum (Olgierd SQ3SWF) also had some issues with RF output changing from PTT to PTT as I had previously experienced.  Olgierd had noticed some RF or possibly oscillation on the gate of Q5 when the radio was transmitting low power out and consuming high current.  I put the scope on my QCX and noticed that also, on my first PTT high current, low power out condition there was some RF on the gate of Q5.  It seemed that Q5 was shunting RF into the front end in this condition.

A 4k7 resistor was placed from +12V to the drain of Q5 and in my case the first PTT issue was completely resolved.  About 30 power cycles were tried with no recurrence of the issue.

4k7 resistor on back of PCB


See photo for placement of the resistor.  Also note the SMD low pass resistors that were replaced.




Programming the processor with new firmware:

  1. The .hex file is downloaded from the files area on QRPLabs groups.
  2. Download Extreme Burner for AVR.  Its a bit tricky to find where to download this on the website.  There seemed to be adverts everywhere.
  3. Obtain a USBasp programmer from Ebay or similar.  Mine didnt come with a lead from the USBasp programmer header that fitted the QCX programmer header, so I had to make one up.

    Lead from USBasp to QCX


    4. Install the Zadig drivers.  I can’t remember doing this, it must have already been installed on the PC.

    5. Put the header on the QCX and plug the 10 pin end into the USBasp and plug the USB connector into the PC.

    6. Start Extreme Burner.  Choose ATmega328P under “Chip”.

7. Go File>Open Eprom File and browse to the downloaded QCX hex file.

8. Click “Write All” and that’s it.

Posted in Projects, radio communications, SOTA | Tagged , , | 9 Comments

CW Paddle for the KX2

I came across this KX2 paddle from the QRPGuys recently.  The price was right at USD 25.00, postage was USD 5.00, so quite a bit cheaper than the original one.  It’s made from soldered PCB material so it won’t be as robust, but it seems to work just fine so far.

It took about a week to arrive from ordering, I understand that they do their mailing on a Wednesday.

The instructions downloaded from the website were clear and straightforward, construction took about an hour.  Right at the start, it’s important to line up the PCB rear and sides nice and square before tacking together with solder.  Apart from that, the only fiddly bit was getting the screws, washers and spacers in place to hold the paddles in place.

At first it felt a but strange to use with the flexible paddles, but you get used to it pretty quickly.

I will report back when it has been used on a SOTA activation.  

It probably needs a cover for the paddles when it’s not in use to protect it from backpack rash.
For the moment, it is wrapped in some EPE foam.

Posted in Projects, SOTA | Tagged | 4 Comments

Toorongo Range VK3/VT-026 Again

With freezing temperatures although clear skies, it looked like this would be one of the last times to get to a winter bonus summit for a while (or so I thought).   The ice was removed from the car and I departed for Toorongo Range.  I have been to this summit several times before, it is not really one one of the favourites although access is very straight forward. There is just a short walk up the Toorongo Track from the closed gate and then a bush bash scramble to the summit.

The drive there is through Noojee on the Mt Baw Baw Tourist Rd and before Tanjil Bren, take the Toorongo-Tanjil link road to the left.  Some six km in, take the Mundic Rd. and then the Toorongo Track.  This time of year, there is a gate about 1km short of the highest part of the road where you scramble through steep bush to the summit.

On turning into Mundic Rd, snow appeared on the road which gradually became heavier on the Toorongo Track.  The road is pretty good for a track classification.  The car was parked about 100m short of the gate where there is a bit of space.  The weather was holding out and things looked quite nice with the snow, much more interesting than the previous summer activations.

The bush bash up to the summit was straight forward, without slipping over like last time.  The summit is on a rocky outcrop with a small cairn.  Just behind the summit, there is a sheltered spot with some good tree cover.  I used this spot once before during high winds and ran some coax out to the antenna.

Contacts were made on 40 and 10 metres in great sunny winter conditions.  As expected 40 metres was behaving pretty badly.


Summit Cairn and Elecraft KX2 Station


Sheltered area behind summit


Sheltered area


Behind the summt


The short log!


Posted in Hiking, SOTA | Tagged | 4 Comments

Takatoriyama JA/HG-252 SOTA

Following on from the previous post, the other close by summit was Takatoriyama.  This one looked very close on the map.  Occasionally, we had seen hikers going past Meagan’s house so it looked like this was the on a popular hiking route.
The SOTA radio gear was put in the pack and we headed up the road.  At the end of the street, it looked like a dead end.  I was about to turn around, when a really old guy hunched over yelled out and pointed to some steps at the end of the street.  The steps looked like they were on someones property.  He shuffled over and gave directions is Japanese (that I could not understand), but I got the general idea with hand actions.  It must have been my hiking gear that made me look like a lost hiker.   I thanked him and we continued.

Park with Sakura

We went up the stairs and then to a lane and at this point, my mapping application (Maptrek with Open Street Maps) showed a dotted line.  The track continued to a small park with a playground and then to the right there was a hiking trail.

At the start of the trail a sign warned of the dangers of wild boars in the area.  We didn’t come across any but there was evidence of digging along the track, not sure whether the boars were responsible!

Wild Boars!

The walk was reasonably steep with some really great views along the way.  We saw many other hikers on the way and practiced our greetings as they went past.   About a km in, there was a sign to the summit with the altitude in metres.  We turned left up there.  The track went past a small clearing (an ideal SOTA location) and then to a small walled area at the back of the shrine.

The trail to the summit


I went back to the SOTA spot and tied the squid pole to a tree.  For the antenna, only the 15 metre section of the End Fed Half Wave was rolled out due to space constraints.
The noise floor was very low which was encouraging.

SOTA location

After a CQ as JP1PBQ, a few JA stations came back.  Shortly after I worked ZL1BYZ and  then VK7CW which was a nice distance.

I also worked 9M6BOB in Borneo with very strong signals both ways, it must have been the sweet spot for propagation at that time.

Weather was holding out, so the station was shut down and we went over to the first shrine we saw and had a look around.  The guy care taking the shrine gave us lots of information on other hiking trails and the shrine itself.

It was in Japanese, so relied on translations from my daughter.

Shrine before the summit

This guy must get cold

The hike back was uneventful, although just as we got back to the street, the rain came down, so another nicely timed activation!

Posted in Hiking, SOTA | Leave a comment