With the help from ZoZo got PCB's made up for the "Corsair" transmitter

This will save on build time and make them look alot neater!

Left hand side board is 6KHz compressor/limiter. Right hand side is the Corsair

You got those pretty quick, JLC does produce good quality PCB's. I'm waiting for Dave Martins thought's when he see's it

That's quite amusing. The audio processor was my design from many years ago! If built as described, the audio filter applies a bit of "mid boost" to brighten the audio a bit, and then there's a 6kHz lowpass filter to keep the modulation reasonably tight. The compressor part is pretty much a "cookbook" circuit using a couple of op-amps and an FET as the variable gain element.

There is a small improvement that can be made to the compressor, to reduce distortion - connect a 1MΩ resistor in series with a 100n cap between the drain and the gate. This applies a bit of the audio to the FET gate, and helps to reduce the distortion (particularly at around 6dB gain reduction).

I always used to throw these processors together on a bit of Veroboard. I think that I've got the layout here somewhere - I'll post it when I find it.

Incidentally, it's NOTHING like an "Optimod". The Orban processor uses multi-band processing and aggressive clipping to give a really tightly controlled modulation level. They always sound HORRIBLE!

That's interesting regarding the modification for reduced distortion, maybe that will have to be implemented on a new revision.

Sadly I no longer have the original schematic I was sent to work from. I think Gigahertz will have to post that to see if it's a 1:1 copy of yours.
All I do know is that it was called "A Very Poor Mans Optimod", hence the AVPM anacronym. There was no other name(s) to credit onto the PCB from the schematic.

Be nice if you can locate and share your original Veroboard design.

Great work Gigahertz

I have never really delved into the realms of AM? But this looks interesting?

I am guessing the frequency of operation is determined by the crystal selected?

Nice work GHz - be interested to know what power you get out of that. I always used a 74HC240 to drive the gate of a FET for SW tx rather than the (I think BFY51) that the Corsair uses. I found you got more power, and less heat. I'll have to dig out a schematic and post it here.

Rev

sinus trouble wrote: ↑Mon Apr 26, 2021 12:41 am Great work Gigahertz

I have never really delved into the realms of AM? But this looks interesting?

I am guessing the frequency of operation is determined by the crystal selected?

Yep uses a crystal but from what research I have done, people have also made VFO circuits and Dave did share the design somewhere. I remember it did use TOKO can which may be obsolete now, if I find the schematic I’ll post it on here.

reverend wrote: ↑Mon Apr 26, 2021 9:17 am Nice work GHz - be interested to know what power you get out of that. I always used a 74HC240 to drive the gate of a FET for SW tx rather than the (I think BFY51) that the Corsair uses. I found you got more power, and less heat. I'll have to dig out a schematic and post it here.

Rev

Thanks Rev, but I don’t want to take all the credit for this! ZoZo helped turn this idea into reality.

I have built a number of Dave Martin’s transmitters on copper cladding. All using genuine parts and with the Corsair get 10w carrier. The only time I’ve notice excessive heat on the BFY51 is when I built the Commando circuit.

I’ve seen 74HC*** being used in the LuLu but not studied the circuit as of yet.

So far the best distance with a Corsair (was weak signal but people where able ID the track) East coast of the States. So not bad low power.

Please share your circuit when you come across it.

It's quite close to the old medium wave circuit I used to build. I used a TDA2030 as the series modulator, and got roughly 12 Watts carrier from a car battery supply. I used a simple 3-chip PLL with a 4.608MHz rock (a standard frequency cheaply available from Farnell or CPC) instead of the VFO. These little rigs - into a "tower block sloper" aerial - would cover a significant area in the daytime!

I'll dig out the Veroboard layouts......

Evening Necks,
This project got put on the back burner due to other commitments, so this weekend I finally had the time to sit down and populate
both boards and put them under test. Works extremely well and sounds pretty good especially with the matching 6KHz compressor limiter.

If your interested in a set of boards (only 4 left) let me know via pm. I can also supply BF245a & BFY51 if required and be ordering some more boards soon.

Many thanks to Dave Martin for sharing the design and thanks to ZoZo for the PCB layout

Looks smashing that GHz.

Loving some of the recent projects on here recently, and it makes a nice change as it went a little quiet at one point.

I have a little PLL oscillator design here that works extremely well with this type of transmitter circuit. There are two versions - one using 74HC logic and the other using the MC145106. Both use the same Colpitts oscillator circuit and provide a nice buffered sinewave output.

I've also used "huff & puff" stabilisers for tunable rigs. If you design for (say) 6.2 - 6.4 MHz, you can make the controlled VCO tune in 100Hz "jumps" - you need a counter to set the output frequency, but it's "set and forget" until you restart the transmitter for the next broadcast. I've found this to be useful to avoid "QRM" from other stations, fax transmissions and other data modes.

Shortwave AM is interesting - I've done it intermittently since the late 60's - and in all that time I only "lost" about sixteen rigs!

Nice boards, by the way GHz!

Can I get hold of these PCBs anywhere? Am interested in making my own Corsair Transmitter.

The circuit for the audio processor is: There are a couple of minor updates in the circuit above. I've added an input pull-down resistor to prevent transmitted clicks and pops when things are plugged in, a 1n feedback capacitor added to the compressor sidechain amplifier (improves stability), and the distortion reduction trick (the series resistor and capacitor between drain and gate of the FET). I've found my Veroboard layout for the circuit in an old notebook - I'm re-drawing it for clarity, and will put it up here soon.

You'll notice that there are two 1n5 capacitors in series in the filter. This value calculates at 750pF, but that's not a standard value. I tried both 820p and 680p in the circuit, and neither turned over at anywhere near enough to the right point. It is actually worth cascading two of these filters for a tighter roll-off above 6kHz, but we're pirates and modulation bandwidth really isn't the top of our concerns!

I used the TL074 quad op-amp, and the fourth op-amp was used as part of my series modulator. My series modulator was simply an op-amp driving a Darlington pair of transistors (either a BD-series device, or a BFY51 driving into a 2N3055). I had a clipper on the input of the series modulator to handle any compressor overshoots (and to stop the modulator "bottoming out").

I'll put up the circuit of my basic PLL-based MW / SW rig if anyone's interested. It's not the height of efficiency, but it's cheap and simple to build, and doesn't require any specialised components.

As promised, here's the Vero layout:

Note - the two 1n5 capacitors are soldered in series - the junction of the two doesn't connect to the Vero tracks. You can (of course) replace the two 1n5 capacitors with a single 750pF part if you can get one.