Following my recent amplifier build for QO-100 I decided to scrap the power hungry Spectrian board. It wasn't faulty but I had decided to put the XRF 286S transistors to better use in a 23cm amplifier.
Nervously I fired up my hotplate and raised the temperature to 235 centigrade, the trick is to turn the plate off when you drop the board onto it. If you leave the power on the hotplate it will continue to rise in temperature to a destructive level. Using a pair of ceramic tipped tweezers I removed the transistors very rapidly and put them to one side to cool naturally.
I had heard stories about the XRF 286 being damaged during removal/re-soldering in this way, so I wasn't too surprised when I found one of the three read short circuit gate to source. Luckily the two remaining devices had survived; reading fine when tested with a multimeter. My plan was to use a PCB which I bought from Jim W6PQL, fixing holes were sized in US inches but 3mm screws fitted nicely.
My engineer friend milled some 6mm thick copper for me; he also drilled and tapped all fixing points including those for mounting to the heat-sink. I planned to reuse the aluminium chassis and heat-sink from which I had just removed the Spectrian board. Using a nibbling tool I removed a bit more of the chassis to get the new larger PCB assembly to mate with the heat-sink.
So the assembly sequence is important; once the PCB is screwed to the copper heat soak It's difficult to get enough heat onto the board to solder. Initially the board is fitted to the copper with a few screws, at this point I slid the transistors into position and marked the position for soldering. After soldering the transistors to the copper and checking for any bridges I placed the PCB sections under the drain and gate tabs to check for correct alignment. When satisfied the fit was good I populated the two sections with the rest of the components including power cables, bias feed and coax. Ensuring the bias pots were adjusted for minimum/no bias I screwed the boards onto the copper heat-soak and soldered the gate and drain tabs. These tabs are thin and usually solder to the board quite easily.
Using 6mm copper allowed me to power up the amplifier prior to housing, I was able to check voltages and set the bias without generating much heat. I set the bias at one amp each transistor as this was the setting for the Spectrian amplifier, I'm still pondering whether to experiment further with the setting. Up till now I have run it at over a hundred Watts and a friend monitoring my DATV on a spectrum analyser said it was clean with spectral regrowth 30dB down.
On this occasion I only took these two pictures. You can see a good video showing the hotplate soldering process on YouTube, look for W6PQL.
This simple to construct project uses a PCB designed by Mike (G0MJW); the circuit consists of an amplifier and mixer with filtering on the input and output. While I have used the board for 4 metres; it is possible to use the same board for other frequencies with adjustment of component values. BATC members can purchase this PCB from the BATC shop. Circuit details can be found here.
For this project I have used one of the Chinese ADF4351 boards to produce a local oscillator at 400MHz. The 4 metre input at 71MHz is mixed up to 471MHz and the output fed to a MiniTiouner.
Another board available at the BATC shop is used for the PIC which sets up the ADF4351 output frequency and RF level. The board designed by Ron (G7DOE) will fit on top of the square Chinese ADF4351 boards. For this project I used the darker rectangular ADF4351 board so mounted the PIC Controller board on a small sub panel. You can see the board in the pictures below here. Code for the PIC is available on the BATC WIKI.
Local oscillator level for the mixer is specified as +7dBm but I found the +5dBm output from the ADF4351 board to be just perfect.
Builders have the option to use the mixer board with or without the PGA-103 MMIC amplifier, I included it on my board as I have no external antenna amplifier.
Below are some pictures taken during the build. The LED on the front panel is connected to the PIC board and illuminates on PLL lock.
See here the PIC Controller board top right; temporary wiring is in place to check that oscillator frequency and output levels were correct.
This picture was taken during a test prior to final assembly; note the hi-tech tinned copper wire antenna.
Now with all the plumbing and connectors in place. To the left of the PIC Controller board is a Chinese buck converter which I used to lower the input volts to a 5 volt linear regulator. This is only for the ADF4351 board as the mixer board will accept 7 to 15 volts DC.
The finished item with not so posh labels; it works a treat so I'm not too bothered about looks.
Lastly a screen grab of the output.
Several local radio amateurs with a keen interest in narrow band DATV have been carrying out tests in the six metre band. With a huge amount of help from Arthur G4CPE I now have transmit and receive equipment using modified commercial downconverters.
About six weeks ago I spotted a Continental Microwave downconverter on eBay, I knew it would be useful as Arthur (a former CM employee) had already briefed me on them. Described as new I decided to bid and got it for £35.00, it arrived with the security seals intact.
What's in the box?
The enclosure is double sided and contains a 70MHz IF board one side and a downconverter board in the second compartment. Conversion was originally 1GHz mixed down to 70MHz using the onboard 930MHz oscillator. There is an easily removable 1GHz filter on the input.
Reconfiguration for 50MHz Reception
In original form the downconverter output is connected externally to the 70Mhz IF input, this is very convenient as it allows us to retune the IF to 50MHz and reverse the process. Removal of the 1GHz filter from the converter input allows us to input 51.200MHz, mix at 930MHz and output at 981.2MHz, nicely in L Band.
Luckily the 1GHz filter is easily retuned to 980Mhz and used at the downconverter output.
Running the DATV Express board with Windows software requires a fair amount of computer power; an i7 processor just about copes. When you get it right the output from the Express is superb;It's not a cheap board but the excellent filtering justifies the cost.
My latest project makes operation of the DATV Express much simpler and uses less processing power. The system consists of the front end and software developed for the BATC Portsdown transmitter. While the Portsdown will run as a standalone multiband transmitter it also has the capability to drive the DATV Express board. Linux based; the software in conjunction with the Raspberry Pi 3B + makes a fast and compact front end.
After taking the various components out on portable operations I decided to construct a complete package to tidy things up. Today I finalised what I will refer to as the head unit. Some tips for my build came from Noel G8GTZ who also saw the benefits of this system.
I will show some pictures and describe what I've done. The DATV Express and associated component parts (Base Unit) will be discussed and posted in a later blog.
Given the increased touch sensitivity and clarity of the 7 inch Element 14 screen I decided to use it instead of the 3.5 inch Waveshare used in my Portsdown Transmitter. Noel G8GTZ pointed out a very nice case for the screen which can be got from MODMYPI. In the picture below you can see the 12 Volt power conversion on the back. I've also drilled the case and fitted an additional audio output as the unit won't fit my shelf with a jack plug sticking out the top.
So there are brightness issues with some LCD screens when viewed from above and I understand the 7 inch Element 14 screen is no exception. Not sure how much truth there is in it but I'm lead to believe the manufacturer decided to invert the screen to solve the problem. Whatever you'll find the picture is upside down in this case; unless you do as I did and use a software solution to put it the "right way" up. If you use it as is; the rubber feet will be on top and you'll need to make a bracket to hold it upright. In my situation the "Head Unit" will be above my eye level so no problem having the case round the right way. More on the software fix later.
In the next picture you can see the contents of the box I fitted to the back cover of the case; it consists of of a Chinese made DC-DC buck converter board to allow 12Volt input. There are a couple of additional filter components and a 5 watt zener diode crowbar at 5.6Volts. The Chinese board is rated to 3 Amps but gets scary hot; there are vent holes top and bottom of the box and it has run for many hours.
My completed head unit will require just one cable to connect to the base unit; a good quality USB cable. At the base unit the USB cable connects to a 4-way powered USB hub. The USB cables I use are the same brand as recommended for use with the BATC MiniTiouner; Lindy Cromo.
Two software mods were made to my Portsdown installation; neither have thus far been undone by the routine Portsdown updates. The first is the fix to put the LCD up the right way; you don't need this if you're happy to use the MODMYPI case upside down. Find the fix here on Github rpi-touch-display-fix
My second software mod allows command line adjustment of the Raspberry Pi sound level; see my blog entry here.
As usual this item was rushed; if I missed something please ask using the home page contact menu.
This 2 metre BPF was a design I found in the Radio Communications Handbook page 5.30. Having recently committed my existing filter to a breadboard project I decided to go ahead and construct a replacement.All of the critical dimensions are shown in the handbook so construction was quite easy.
Since I completed the filter it has been aligned by a friend Arthur G4CPE. Setup was carried out on Arthur's Wiltron RF Analyser. Bandwidth is 5MHz and loss is under 2dB at 146.5MHz.
This is an inexpensive project using basic tools, coils are wound on a 9.5mm drill bit and the piston capacitors were found on eBay. There's not much to it as you can see from the pictures below.
On test this week my mk2 transceiver with power amplifier was producing just over 1 Watt. One of my friends; Arthur G4CPE suggested checking output directly on the pa as he wasn't convinced the connectors were ideal. In my haste to get the job done I had grabbed what was in my connector stock and used this open type. I should have considered suitability at 5.6GHz.
With the test gear connected directly to the power amplifier I was amazed at the difference and saw 35.9dBm on the meter. Arthur produced an SMA to N bulkhead connector with a short lead and urged me to do a quick swap.
Wow what a silly mistake I had made and what a difference. Next I changed the receive antenna connector for the same type. When I get out portable in a week or so I hope to see a big difference.
Well the Res Ingenium THV15 VHF pallet amplifier arrived from the Netherlands as expected. I quickly unpacked it to check it was all okay, it had been described as brand new old stock. Sure enough the amplifier was in the original packing and sealed in an anti-static bag....great.
What was needed now was a few hours work to make it usable.
Following a visit to Arthur G4CPE I left with some useful tips and a gift of a chunk of heat sink; off I went to make the pallet a home. I should mention that the heat sink and its base compartment was over a foot long, time to fit a new blade to the hacksaw and locate the tin snips. An entire morning later I had a smaller chunk of metal and a roughly fashioned base. I don't have a metal folding machine so the base was reshaped by brute force on the end of my workmate. It looks great from one side.
There is a solder pad on the PCB, a point you can ground to remove the bias for power down but I decided not to use it at the moment, for now I'll switch the 26volts. So not too much soldering was required just SMA in /outs, power supply and fan. It probably took longer to test and set up than it did to build, I couldn't take risks with this valuable amplifier.
The amp which is rated at 15 Watts has 30dB gain and Arthur had found his was most linear at around 4 Watts. My driver amp was to be the 7 Watt Minikits amplifier which up until now had been run as a low power PA. Some padding reduced the drive power to the 6dBm I was looking for.
In a hurry to get the project finished and working I didn't take many pictures of the build, below are a few just before the fan was wired.
On Thursday this week the PA was hooked up to my 2m DATV kit and it performed superbly.Our next test will be over a longer path to Ian G3KKD in Cambridgeshire.
Last week I located a class A power amplifier at a good price, it will be ideal for 2 Metre DATV. Rated at 15 Watts this is the same type used by Arthur G4CPE. Some work is required before I can use it as it's a pallet and needs heat sink and connectors.
My amplifier is coming from the Netherlands; its new old stock hence the good price. At the moment I feel like a kid at Christmas; looking out the window for a courier.
I'm hoping to use my new power amplifier during the next BATC activity weekend in June. First though a temperature test was needed as it's not very efficient. Testing power consumption a couple of weeks back it was using 85 Watts to generate 10 Watts RF output.
Following discussion with a knowledgeable friend I added a small Papst fan found in my spares cupboard. Advice was to have it cut in at about 60 degrees C; so I bought and fitted a normally open bi-metal device.
Initially I thought about buying some of the adhesive temperature strips and couldn't believe the cost; luckily I was loaned a measuring device with a probe which tucked nicely between the temperature sensor and one of the heat sink fins.
Starting at 20c which was shack temperature at breakfast time I fired up the amplifier with loads on input and output; at the same time running a stop watch. From an earlier brief run I had suspected the sensor and fan would be a safety feature only, I figured I would be OK with short overs.
It took 20 minutes and 40 seconds for the fan to kick in and the probe showed that it was exactly 60 degrees. After 2 minutes with the fan running the probe temperature dropped to 54 degrees; and 51 degrees after 5 minutes. Confident the amplifier is protected I'll run it in June.
There is one more issue to deal with; and that's where and how to mount the heavy lump without using long cable runs. Maybe the subject of another post.
While out and about with my 5GHz ATV portable set-up my modified survey tripod has been a talking point. As standard they have a 60mm hole in the centre of the head; a gimbal fitted with a 3/8" threaded screw is used to secure the measuring instrument.
What I needed was a pole mount to secure either a dish or WiFi grid; I removed the gimbal and frankly I had no idea how to make the required change. It was later the same week I recalled another club member had some brackets specially made by a company in Nottinghamshire; I gave them a call. Photographs and measurements were emailed and within a few days I received a drawing.
In this view of the underside of the tripod head you can see the peg on which the gimbal was mounted.
Below is a picture of the completed modification; which was created without making any changes to the tripod head. Apart from refitting the gimbal there's nothing more required to restore the tripod to its original condition.
If you're looking for any special brackets for your antenna installation I recommend you have a chat with Brian; see link below.
This simple On Screen Display uses a board intended for use with a drone or quadcopter. The project was devised by Mike Stevens G7GTN and published in issue 12 of CQ-DATV.
Video quality from the finished item is excellent and ideal for use with amateur TV output. You should understand output is monochrome and single page.
Two PCB's are used in the project but only one is permanently installed, the other is used for programming. Unfortunately there are no mounting points on the OSD board so a little fabrication is required if you plan to box it up. I soldered some push pins to the grounds on mine and soldered the other end of the pins to copper clad board. If you want to run from the shack 13.6v supply a tiny regulator board can be added. See picture.
Out of the packing the OSD board is pre-loaded with a font set suitable for aircraft use e.g. an artificial horizon, not useful for amateur TV. Mike has provided a font set which can be downloaded and installed using the Arduino IDE software and a simple to use terminal programme. All instructions and software links are available on the CQ-DATV site so I will just outline my build.
Before you connect up the board there is a small amount of soldering needed. The OSD board has two isolated sections. At one end video output at 12v and at the other digital input at 5v, the 12v input feeds a 5v regulator which we are told is prone to overheating. For this project the two sections are linked and run with a single 5v supply. There are link points on the board so the alterations are a five minute job with the right tools.
Digital input to the board is serial so the second board you'll need to source is an FTDI USB to Serial interface. These are cheap as chips from eBay; mine was less then £3.00 delivered. Most will come with a set of leads. Beware as some Chinese sellers advertise as FTDI and supply an alternative which in my case didn't work.
Pictured is the genuine board which is only required during programming. Mine can be re-attached if I want to make changes.
Prior to boxing up; the parts cost less than £10 and it worked great. I built two of these with different talkback frequencies; a paged unit is planned using several boards in one enclosure.
OSD with no video input. Monitor TV is under a shelf causing some shading at the top.
Links. Working at time of writing. There are other suppliers for the hardware.