Saturday, 28 October 2017

More UHF APRS Testing

I tried the E32-TTL 1 watt 70cm LORA module a bit further away from home, again using APRS software to send the position to the internet.
The location I tried first was some high ground just off the A65 to the west of Settle (grid reference somewhere around SD786669) on the narrow road above the Austwick TV mast. This has a good take off towards the west.
Instead of the mobile antenna I put a short antenna directly on top of the module and taped the thing to a mop handle so it could be as high as possible while operating. This also gets rid of any coax cable loss.

 Video camera also fixed to mop handle

 View to the west from near Settle


 But did it work? Yes it did, at a distance of 37.7 miles (60.7km), with more than one transmission being received as I moved around the parking area.
I tried another two locations further south, which weren't on high ground, but received nothing. These were at the Lancaster (Forton) M6 motorway service area and just outside the boundary of the Charnock Richard M6 motorway service area.

My next experiment was to compare the performance of the LORA module against normal 1200 baud packet radio using an FM mobile transceiver.
Using the same antenna as in my last blog post (it's a small Moonraker dual bander, about 40cm long, on a mag mount), I transmitted APRS beacon messages from my Yaesu FT-8900 on its 20 watt setting. The frequency was the same as I had used for the LORA test.


See the results of the LORA test below


Result? A lot worse than the 1 watt LORA transmissions. Because the car APRS tracker has no logging function, I could only guess where any failed transmissions were from. I know that it beacons every 110 seconds so I've put red Xs in rough locations of where it should have worked but didn't.
The blue O shows where it actually worked but no location was sent (my tracker is programmed to send some other information every few beacons instead of the location). I had stopped the car there, which would have helped as I wouldn't have been passing through any multipath dead spots while transmitting.
So with 13dB (20x) more power on the same frequency, traditional packet radio is still worse than LORA when used mobile. Even if you compared VHF packet with UHF LORA, I don't think 20 watts VHF (to a 0dB gain antenna like the Moonraker) would get much further than 1 watt UHF.

And here is Dales Radio's Ingleton antenna (103.0 MHz), covering the area further west of Settle, with Settle itself on 104.9 MHz.




Thursday, 19 October 2017

433 MHz LORA

Experimenting with a couple of 1 watt 410 - 441MHz LORA based modems.
The part number is E32-TTL-1W (there are also 100mW versions available).

Manufacturer's website and data sheet

Only about £12 each on the usual sites


With most electronic equipment shipped from overseas, the specifications are usually quite 'optimistic' so I wasn't really expecting 1 watt from them. Especially as the 100mW version looks exactly the same. But they really do put out around 1 watt.



They fit nicely inside a Choc Box once you cut out the bit which is supposed to go through the terminal block. The cable is a USB to TTL serial adaptor, they were less than £1 each.
The power is supplied from the USB cable (red wire), with the current on transmit 670mA (I couldn't actually measure this using a USB tester, maybe the quick variation in transmit power confused it).
Although they draw quite a high current from the USB port, I had no trouble using them on a PC USB3 port and my phone's USB-C connector (through an On The Go cable).

Programming software. These settings are for the highest power output and slowest transmission rate (greatest range). There are 65536 addresses available, but #FFFF can be used as a broadcast address.



Only 1 MHz channel steps are available, and the transmission is 125 KHz wide, centrered on the MHz.

How far do they go? LORA is supposed to work with very weak signals at the lowest transmission rates. I connected one modem to the car antenna and the other to the base antenna on my chimney. The data I transmitted came from an Android App on my phone APRSDroid and was received by APRSIS32 in Windows.
To these applications and the APRS internet servers, it looked just like I was using packet radio TNCs and the normal APRS frequency (in the UK) of 144.800 MHz.

Transmitting at least once every 2 minutes, I drove to a point over 8km away. At this distance there was no line of sight. It even worked in Dalton town centre, behind a hill. Over the route there were only 2 locations where the signal was lost, see the Xs on the map below
These locations are at the back of a hill, I would have expected it to be worse in the town centre than there but at least I know where to test in future.
Range would have probably been well over 8km if I'd carried on further east along the A590.
These are good results for 1 watt, probably as good as 10 watts on VHF APRS using normal 1200 baud packet radio.

This would be interesting as a weak signal mode when propagation is good on UHF. Using terminal software you can just type and they transmit, much like PSK31 but with error correction.

Thursday, 28 September 2017

WSPR House vs Car

A test comparing my home and car antennas on 14MHz using WSPR

At home, I used 5 watts into an end fed wire that was run along the gutter, with the feed point just inside the upstairs window. This isn't cut to any specific length except to fit along the front of the house and isn't much more than about 7 metres long.

This is fed through an LDG auto tuner

In the car I used a Watson Multi Ranger 9 band whip on a small Sirio mag mount. The mag mount should really be bigger for the HF bands (except 28MHz - this type of mag mount was also sold with a Sirio CB antenna). The transmitter was a SOTA Beams WSPRLite, with 200mW output.
The car was parked on the road outside my house.

Both the wire and the car are on the north side of the house, with the axis of the wire about 20/290 degrees. To the north there is a field across the road, then more houses around 100 metres away.

During the test, both transmitters activated simulatneously 3 times in their random sequences. There was some spacing between the frequencies so they didn't overlap, around 50Hz. 

With equal gain on both antennas, there should have been a 14dB difference between the house and the car. Most reports had less than 14dB, showing the car antenna was performing better. In some cases the car was better, even with 14dB less power. The best performance of the mobile antenna seems to be towards the north (OH - Finland, LA - Norway) and worst towards the south (EA - Spain). This makes sense with the open area towards the north allowing a low angle of radiation and the house to the south blocking anything in the direction of Spain.
From those results, it looks like the mobile antenna would be best for all directions in an open area.

There were some stations who I was received by only when using the wire antenna and 5 watts, including those in the USA and Canada. Without the extra power, none of these stations would have received me as there was much less than 14dB margin above the limit of WSPR reception. The 200mW to a mobile antenna has got to the USA quite a few times before.

Friday, 22 September 2017

FT-847

Borrowing an FT-847 and getting back into using some HF data modes at home. Quickly put up a HF antenna along the house under the gutter, with about 8 metres of wire fed through 9:1 UnUn and LDG auto ATU.
 Data modes interface made from 12v relay board and a few other components.

 As the FT-847 HF data connection is a 3 pole 3.5mm jack, transmit audio and PTT share the same input. I used an isolating transformer for the transmit audio. Because of the low current needed to operate the PTT circuit, and the transformer, I had to choose quite a low capacitor value for the audio coupling. Otherwise the capacitor would charge up every time the PTT contacts opened, drawing current and keeping the transmitter on. With the first capacitor I had handy, a 150uF electrolytic, it took about 15 seconds to change between transmit and receive, ending with a bit of relay chattering from the 847.
I changed to a 470nF which I got out of an ADSL filter. The FT-847 manual shows a recommended 2.2K resistor to pull the PTT circuit low. The 470nF and 2.2K would give a cut off frequency of around 150Hz as a high pass filter, just about right.

 Here is the circuit. It's not ideal because the relay board has a 'high' and 'low' level input selectable by jumper. The high level input just connects the input signal directly to the relay coil, which is too much of a load for a serial port to drive. The idle state of a serial port RTS line is negative, so what I've done here is use this to keep the relay held on except when I need to transmit. If the PC shut down or the 12v power to the relay board was lost, the radio would change to transmit. But with SSB data modes, unless there is audio AND PTT stuck on (which couldn't happen if the PC shut down), the worst that could happen is a few milliwatts of carrier and noise leaking out.
The transformer made a difference on transmit, without it, there was quite a lot of noise along with the wanted data signal.

Thursday, 21 September 2017

Baofeng BF-888 Charger

Inside a Baofeng BF-888 charger

The output is around 5 volts through a resistor, 2.2 ohms. Probably turns red LED off and green LED on when voltage across this falls below transistor junction voltage (fully charged cell around 4.2v). Green LED comes on whenever a radio is put into the charger.
Not too many components and no need to pack them in too tightly like in a plug top USB charger.
Not the worst separation between high and low voltage sections that I've seen.

Shaver adaptor with 1 amp fuse. The mains cable is the same stuff as you would get on the low voltage side of something like a battery charger. If anything did go wrong inside the charger, I wouldn't like to have that cable on an unfused outlet on a 32A ring circuit. The input goes through a low value resistor as it enters the circuit board, which would burn out very quickly with the full mains voltage across it though.



Wednesday, 20 September 2017

Cheap Handhelds vs Outdoor Antenna

 Tested these radios on an outdoor antenna - Comet GP-15 6m/2m/70cm, for both strong signal handling and sensitivity.

VHF
The Baofeng GT-3 dual bander was completely overloaded right across the 137-174MHz band. I don't think it was paging, even the paging on 153MHz sounded like it might have been getting some interference. I'm guessing strong broadcast signals (FM or DAB) as it was non-stop desensing of the receiver. It's hardly worth going outside with this radio on VHF or connecting anything but the stock antenna, it was the worst of the lot. Receiver sensitivity is as good as any other handheld though. S-meter isn't just 0 or full, but seems to be linked to the squelch level (with squelch of 0, there is no meter reading on any signal).

Puxing PX-777 VHF was much better than the Baofeng at coping with strong signals, even though it was one of the first cheap handhelds on the market (they came out in about 2007 for £35). At 165MHz, there were no problems, but near the bottom of the frequency range there was some pager breakthrough on mid-level signals. I wouldn't take this very close to a transmitter site and expect decent contacts on it but it's definitely an improvement on the more modern Baofeng. I've had problems with the battery life in cold weather so maybe not one to rely on for a hill top activation on it's internal battery only.

Uniden Bearcat UBC3500 is a handheld scanner covering 25 - 1300MHz and despite the wide frequency coverage, this didn't have any great problems with strong signals anywhere in the 137 - 174MHz range. The 3500 isn't quite as sensitive as the other handheld radios but in everyday use this doesn't really matter.

UHF
UHF isn't usually as hard on radios as VHF for strong signal handling, apart from Airwave around 390MHz, most transmitters are low powered and on roof tops rather than hill top masts.
The UBC3500 had no problems with overloading, but you could definitely tell the lower receiver sensitivity. This could have been better, as some of the other handhelds were much better on weak signals without any interference from nearby transmitters.

The UHF version of the PX-777 and the Retevis RT-3 (TYT MD-380 rebadged) were both able to cope with any strong signals and more sensitive than the Uniden scanner. The S-meter on the PX-777 wasn't realistic, giving full scale on a slightly noisy signal which was about S1 on my FT-7900 mobile radio. I hadn't used the RT-3 on FM very much on an external antenna and it's hard with DMR to tell if a signal has either gone below the noise or has been wiped out by something else. The RT-3 has a thing with the S-meter where it goes up to full scale for a few seconds after changing channel, which makes it look a bit like there is some interference in the background.

Once again, the Baofeng GT-3 was worse than other handhelds, and around 456MHz was being overloaded by something which didn't affect the others, making it struggle with the weakest signals. Lower down the UHF band it seemed OK, so may be usable on UHF even when VHF isn't.

The Puxing PX-2R (photo below) was as good as the others on UHF, it's only 2 watts though so not something I'd really be connecting to an external antenna any time. There is a few seconds delay on the S-meter, which usually reads full scale or nothing. It has a Nokia type charging socket and the battery looks like a mobile phone battery.
Last, and least, was the Baofeng BF-888. But what did you expect for £10? The slightly noisy signal on the other radios was non-existent on this. Without any display or tuning dial it isn't easy to tell whether it was sensitivity or strong signal handling. They do work - around a site or into a nearby repeater they are OK and no great loss if someone breaks one but unless you only have £10 or really must have no display, buy something else.

Tuesday, 19 September 2017

Inside a HF SDR

 Inside the £25 RTL-SDR receiver with dual inputs. Low pass filter for HF is a coil on the PCB.
 The original antenna socket is left on the board


Monday, 4 September 2017

Walton Hill

A couple of weeks ago we went for a little 3 night tour of the Midlands, stopping at Oldbury, Droitwich and Leicester.

Droitwich from the hotel car park

On the Sunday, I had a walk up Walton Hill (SOTA G/CE-002) in the Clent Hills South West of Birmingham, parking in the large car park at SO942802.
I just had the Baofeng GT-3 handheld with me but as the hill wasn't very far from Birmingham I knew I'd still be able to make a few contacts. I managed the 4 QSOs for SOTA but there was a problem, even with its own 20cm long antenna, on 145MHz the GT-3 was being overloaded by some strong signals. At one time it would probably have been paging transmitters that were the strongest things near that frequency, but many pager sites have been closed now and the interference could have been from anything on VHF. Even on strong signals that were perfectly readable, the squelch was being closed by some noise in the background and I ended up keeping the squelch open.
At home on my outdoor antenna, the GT-3 is completely wiped out on VHF and I'm over 10km from any mast with powerful VHF transmitters (but the closest does have high power DAB and FM broadcast so this may be what is causing the problems on hill tops).

Mast at Romsley to the East of Walton Hill
 Malvern Hills

Clee Hill



 Brierley Hill
 Towards Birmingham

Thursday, 24 August 2017

HF SDR

After getting rid of any HF equipment last year, I'm slowly putting together what can't really be called a HF station. I got a WSPRLite to transmit WSPR with at a massive 200mW, and have been using that into a mobile antenna on the car. I thought it was time to get something to receive with, so bought one of the cheap HF software defined radios, based on the usual TV tuner devices.







This has 2 SMA antenna sockets, one for below 30MHz and the other for above that. There is data/power through a Mini USB connector (just when you think nobody uses them any more). So a bit tidier than a TV tuner 'dongle'.
On VHF/UHF, it seems about the same as a standard RTL TV tuner, using SDR# software. On HF and below is where it is a bit strange. I thought it might have some kind of up-converter to allow reception below 30MHz (they can tune to about 24MHz normally). No, it doesn't use any converter, you just select "Direct Sampling - Q Branch" in the settings (even if you aren't James Bond).
But, as the sampling frequency is fixed at 28.8MHz in this mode, it can't receive anything above 14.4MHz. If there is a signal on 14.3MHz, it will also appear to be on 14.5MHz etc. Close to 14.4MHz there is a lot of noise from the 28.8MHz oscillator, so it doesn't really work on the 20m amateur band either.
It does work OK below 10MHz, best on the 3.5 and 7MHz amateur bands and I've had it receiving WSPR on those once I worked out the frequency error (it's about 50ppm out - even at 3.5MHz that would be too much to leave uncorrected for WSPR).
SDR tuners seem to be quite fussy about the power from the USB port - on my laptop this wouldn't work at all without being plugged into an external powered USB hub, and even the standard TV tuner dongle refused to work with a short USB extension cable.

So if you're looking for something covering all HF bands, this isn't for you and you should spend a bit more on one of the better software defined receivers.

I've been using it with a little active antenna built by RA0SMS