[Laser] led txr

[I]

Improvements to G8CYW transverter and led transceiver.






I have for some time now been aware of the behaviour of my designs with FM versus SSB. On FM, the S-meter goes to full scale on strong signals, but on SSB it stops short, S8 or S9 on the FT 817. I know we are dealing with two different circuits in the FT 817, but I thought we are also getting some limiting on SSB.


I have now completed my investigation now, solved it and got some more performance out of this gear.


Basically, there is too much gain in the receive head and too much attenuation in the receive path of the transverter to make up for it. This does not affect FM too much, but makes the mixer in the transverter limit. This means there is too much noise coming in from either the receive head alone, or the transceiver head, depending on what you have built. As a result I have modified my transverter, a receive only head, and my transceiver, all with worthwhile results. You need to modify both transverter and receive/transceive head to get the benefit.


Find the receive attenuator in the transverter,consisting of 56, 6.8k, and 680 ohm resistors. Remove the 56 ohm resistor, do not replace it with anything else. Swap the other two resistors round so now you have the 680 ohm in the signal path. There is a 47pF capacitor connected in the signal path at this point, replace it with a 1nF capacitor. This mod reduces the output attenuation by some 30dB at least. Finally, where the head rx signal comes to the transverter board lift the signal connection and install a miniature 220 ohm pot, the input signal goes to one end of the resistance track, the other end to earth and the slider to the transverter input pin. This just sits on my board ok. If you power up the transverter board with no head, there still is no noise noticeable on 3.6MHz.


The receive/transceive head mod is simple, just remove the last stage of amplification, yes, less really is more! Lift the 2.2k resistor on the output of the first op-amp (you need this op-amp, it buffers the cascode amplifier)and connect a 100nF capacitor directly from pin 6 of the first op-amp to the signal pin on the 3.5mm jack socket. This removes 24dB of gain that the second op-amp was contributing. You can remove all the components from the last stage as it is now redundant, or just unplug the op-amp if you used a socket.


All you have to do now is connect head, transverter and rig, switch on and with the head in TOTAL darkness, adjust the new pot so the noise (still obvious when you connect it in to the system, all the noise comes from the front end) does not drive the S-meter.A little less than halfway up from 0V did it for my rig. Then check with your beacon with the rig on SSB and you should find the S-meter goes nearly to its fullest extent with the beacon close to the head. I find now that the noise masked weaker signals and I can now receive my beacon reflecting around the room better than before, so with a more balanced gain profile throughout the signal path, I have, (a) got a better receiver, and (b) got the S-meter to go nearly to max on a strong signal, I think I am missing the last block on the display so there might still be some limiting on extreme signals.


The beacon, for those that do not know, is simply a 555 timer astable running at 20kHz feeding an ordinary 5mm led, increase the led resistor to reduce the optical output, I usually use resistors in the range 1k to 10k.


As an intiguing side benefit, one of the correspondents mentioned that the transverter also functions as a VLF transverter, if a long wire antenna is connected to the rx head signal input, rf from virtually 0Hz upwards is mixed up to 3.58MHz and upwards. (The LED output would enable transmission on the same rf frequency but not to be recommended unless you want to go on 136kHz, where you would need to amplify the signal). I tried the receive option with some relevant results to us opto-commers that I will deal with first. Some of us have been troubled by an unwanted response on around 3.605MHz, I previously incorrectly diagnosed this as being from my local MW "Magic Radio" station on 1152kHz (aerials located 400m away from my home QTH) since when investigating a strong response I could hear the programme content when swtitched to AM. It turns out that this was cross-modulation, partly generated by having the mixer limiting. Using the transverter with an aerial connected I have found that there is a strong signal actually radiating on around 25kHz, so that is what it is.
I can hear the time signal transmissions on 60khz, the LF radar at 100khz, all the German time signals near the 136kHz amateur band, all the long wave broadcast band, and I have been up as far as 350kHz into the nautical beacon band and I can hear all the local signals (NT on 350kHz for me), what fun! For the last two Christmasses at least, the Swedish museum has fired up the Alexanderson alternator on 18kHz from Grimeton I think, I will be trying to hear that soon. Also, some German amateurs have been testing on around 9kHz, something else that would be possible to receive here. Anyone interested in persuing this could possibly replace the 3.58 MHz crystal with a 4 MHz one to make the maths easier.



I will send modified circuit diagrams to Tim in the hope that he can replace those shown on his project pages.


Stuart, G8CYW