"John R. Strohm" wrote in message ...
I see an obvious problem in the schematic.

You show two separate transformers labelled T5. One is clearly a trifilar
device, in your two-diode balanced mixer. The other is clearly a bifilar
device, coupling output from the IRF510 final into the output filter. You
probably want to fix that nomenclature, and add data for the RF output
transformer.

Also, I'm curious. Why did you use 2N3866 everywhere, instead of, say,
PN2222A (plastic 2N2222A)?

john and others -

tnx for the corrections. i also forgot to put the half-wave filter's
specs and couple of other subtle typos. i am trying to assemble the
second prototype (the one in the picture) refering to just the
schematic on the web, to check if things are proper. i will post the
corrections soon.

what i did discover about 2n3866 vs. 2n2222 was this: the 2n2222 is
mostly used in low frequency switching applications in india.
therefore, slimy manufacturers change the markings on other lower
frequency transistors and palm them off as 2n2222. the BF194s that i
have used extensively is a close plastic cousin of 2n2222 that is used
as self-oscillating mixer for shortwave radios in india. It does a
reasonable job at HF frequencies if you are not too particular about
performance.

The 2n3866 is used in india by the cable TV industry for boosting the
line signals. We ordinarily get about 70 channels through the cable
(at $3 per month, including HBO/Discovery etc.). The 2n3866 has to
keep intermod to a minimum. therefore, we get fairly good quality
transistors as far as '3866 is concerned.

in anycase, you will see that the post mix (which should have used a
220R as the emitter bias resistor instead of the wrongly shown 560)
needs to be a large signal transistor. I tried with a 2N918 first, as
the post-mix and noted an immediate increase in the background hiss
(compared to the 3866). I tried the BF194/5 and found it was easily
overloading from the computer noise (i am using a PC-based scope). I
tried also a BD139 as the post-mix(a 15 watt hifi audio beast with Ft
of 150 MHz) and found it to be very noisy.

The circuit uses feedback and degeneration quite extensively, so I
urge builders to try different transistors. Most should work, some
better than the others. You can also try paralleling two 2n2222 with
separate emiiter bias (diving the current in half between them) for
the driver feeding the IRF510.

also, i am sort of giving up the idea of making the txcvr multi-band.
the bandswitch is a pain. and the transciever costs about $10 in parts
here in india, so i might as well build one for each band upto 21MHz.

oh, btw, the local spares shops say that the IRF511 has been
discontinued. is it true?