Home brew - old projectsThis page will gradually get more things added to it as I keep coming across old stuff in boxes! It provides something of a history of my interest in home constructed electronics.
My first oscilloscope
This was based on a couple of designs that appeared in Practical Wireless (which made me want to build one) and Practical Electronics July - August 1967 ("Investigator Oscilloscope") which persuaded me to have a go. I couldn't afford much as a schoolboy - certainly not a mumetal screen (so the power supply with its magnetic-field generating transformer is in the separate box on the left) and not even a tube base - the connections to the CRT are soldered to the pins! I still have the pages from the magazine, but they didn't put dates on them then so the link above came via Google. Note the lack of knobs on the less important controls, and the elegant wooden boxes.
Here's the trace shown better - illustrating the lack of flyback suppression and the not too linear timebase waveform at the left hand side. Even with little faults like these and its relatively limited bandwidth, this was a very useful instrument to have and was used to check out most of the projects that followed.
Here it is in the nude - the front end of the CRT is supported by a bit of foam rubber in the end of a tin can, and you can make out the soldered connections to the CRT tube pins. It's an all valve design of course; and many of the components were salvaged from old TVs etc.
Here's the under chassis view - it's all quite simple really. I used to love working with valves, they were much easier to get working than early transistor designs. I think that's why I now like MOSFETs!
When I started in amateur radio I took the licence conditions very seriously, as we all should; here are a couple of the wavemeters I built to measure output and check for harmonics. The wavemeter is a lovely simple bit of test equipment.
Grid Dip Oscillator
A bit more sophisticated than the wavemeter, the grid dip oscillator is a signal generator and a device for checking resonance. My version uses a transistor oscillator (I'll have to take the lid off to check the details when I put a battery in - it's not been used for a long time) and it had (has still, I hope) 5 coils mounted on salvaged octal valve bases. The meter isn't very sensitive - it was all I could afford at the time but it was driven with a transistor amplifier.
As digital electronics became more available the options for test equipment expanded. I thought it was wonderful to be able to have a crystal controlled frequency meansuing device. This is built with 74 series TTL so the innards get warm - the 74LS series wasn't available at first (to say nothing of 74HC etc). This is still very much in use.
2m FM transceiver
I eventually got round to taking the RAE and getting my Class B (couldn't use HF then!) licence early in 1978. I worked on a 2m FM rig, crystalled, based on circuits in the old RSGB VHF/UHF Manual whilst waiting for the licence to come through.
The rig worked well, though from my flat in the bottom of deepest Morningside I found I had to make most contacts through the GB3CS repeater. However being portable (getting about 0.6W RF out of 8 AA dry cells) it could go to the tops of hills and do a bit better. Best was into GI and EI from the island of Islay (GM). It was normally used with the whip shown (closed - opened it is 1/4 wave) though I have also used it with a quarter wave groundplane in the loft in the current flat.
Construction was all home made double sided pcb, etched one side, with screening constructed from the same stuff. The loadspeaker is actually one from a Philips Interphone Engineer set - had only one function but with brothers to play with it was a substitute for a walkie talkie. Battery boxes are at the bottom (also DC in connectors for the external PSU), with 4 crystals installed for S20, S21, S22 and R6. I eventually replaced R6 with S23.
In the view below, the receiver board is on the left, and the transmitter board centre, just to the left of the battery box. Other boards include mic amplifier and AF amplifier.
Recently I found the hand dawn circuits, layouts. Here's the RF part of the receiver ...
and here's the IF and audio part of the receiver.
The transmit side crystal and trimmer has been lopped a bit in scanning - I didn't believe in wasting paper!
All good fun. I have the layouts as well but they are very messy having been extensively scribbled on during construction.
The second set was built for communication over short distances for our "megalithic astronomy" trips, and was also finished in 1978; it just had a crystal for S22. Guy, G3ZHL, was on these trips so we were able to communicate over long surveying paths legally.
Direct conversion 80m receiver
Working with G3ZHL got me enthused about 80m, and this was my first step towards getting going on HF. My usual sturdy wooden case to contain it! The joke in my family from Lancashire was always that "it's built like a wagon" because my grandfather was a carpenter for British Railways who built railway wagons.
It's a nice simple design, the four boards visible in the photo below being, from right to left, RF stage (single transistor amp), local oscillator (the tuning knob is geared which makes this easier - and calibration was of course done using the frequency counter above), product detector, and AF amp. Pretty well all junk box stuff even when it was built.
Simple it may be but it worked pretty well. I fear that bits from it may now be salvaged for things like QRP antenna matching units ...
My first real computer
As I became more interested in computers I wanted one at home; microprocessors were starting to make this possible and I had been trying to find a student-affordable way of doing this when I was an undergraduate, without success. By the time I was a postgraduate (1975) microprocessor chips were becoming cheaper and more easily available, and so it was not too long (I don't know exactly when, will have to dig out ancient notebooks) before I got one. This is my Z80 test rig; it was basically a very simple computer with a Z80 processor and a couple of 2114 memory chips (bottom row to the left of the buttons). The observant will have noticed that there is no Z80 - this was removed for use in a Z80 second processor which I built for my BBC micro, and subsequently ported CP/M to it - that's another story.
This computer was programmed by DIP switches (the joke was that it was programmed with a screwdriver since this makes it easier to set a DIP switch). The switches set memory location and data, the buttons control operation, and the row of LEDs (in the photo above) showed address and data information. This wasn't an unfamiliar concept to me as I had learned to boot a PDP-7 using the switches when playing in the computer lab in Cambridge.
Here's the circuit diagram in its draft form (click to get the full
size version); this dates from about 1979. I wasn't meticulous about
dating notebook pages in those days as I am now!
I have found a slightly cleaner version in the same notebook, and
also have pages of hand code, but getting it going is easy and left as
an exercise for the student! Here's the better version (click on it to
get the full size one):
Somewhere I have the circuits for the BBC
microcomputer "second processor" version as well - and a whole notebook
devoted to writing a BIOS, then porting CP/M onto it!
As I became more interested in computers at home, I was also using them a lot at work, and with the emergent network we had dial-up access available to the University system. I took advantage of this by building a modem (it used a single AM7910 chip, details to follow when I find my notes) and I built an acoustic coupler for the telephone handset with a microphone and loudspeaker connected to the wires you can see. The coupler was built using a couple of old headphone ear mufflers, cardboard etc. and fell to bits eventually so is no longer available to photograph. I wrote modem software for the BBC micro which enabled me to use it as a dial-up terminal. Happy days!
Here's the circuit, complete with crossings out as it was wired up.