开云体育

Update on this scratch build -- the receiver is working well, I use it on 40m regularly.? But the transmitter is unstable. I've returned to it over the last few days to see if I can iron it out.? Recap, this is a scratch (mostly surface mount) build, exciter on one board, Nano/LCD/encoders/switches/si5351 with 2n2222 clock buffers on all clocks on another board, predriver/driver/single RD16HHF1 PA and LPFs on another board.? TDA2003 audio amp replaces the LM386.? Otherwise as per G6LBQ Rev 1.6.??

To cut a long story short, it looks like the first IF amp in the transmitter is unstable (Q5).?

The symptom is that when PTT is pressed, the transmitter produces SSB but it is dirty and full of RF feedback; the signal looks rough on a scope; the power meter kicks up with speech as expected but sits about half power between speech input with all the crud.?

This is what I know so far:

-- if I remove the RF input from the exciter board to the PA board everything after that (predriver, driver, single RD16HHF1 PA) is stable => PA is not the problem?

-- if I inject a few hundred mV of 7MHz signal on the base of the driver I get a healthy 5 or so watts output => PA is working as expected??

-- if I remove the +12v Transmit to Q5, everything after that is stable; touching the crystal filter input or output injects noise which? can be heard and seen on the scope => everything from the second IF amp to the PA is stable and working??

-- if I put +12v T back on Q5 but remove it from the mic amp, there is no change => instability is not coming in thru the mic amp?

Another observation -- when Q5 is powered up and in transmit, it (or something) puts a horrible 0.5v signal at roughly the signal frequency (or it could be the BFO frequency, I can't tell with my cheap scope) on the 12 DC supply line, which makes its way to the si5351, and causes the 2 clocks (VFO, BFO) to be distorted.?

-- if I short out the input to Q5 (to earth) while this stage is powered, the instability stops => Q5 is oscillating .?

The whole thing is securely mounted in an aluminium case.?

There is all the usual 100nF decoupling.? I wound the +12v Tx supply wire to the exciter transmit stages thru an FT37-43, no impact.? I wound the short +12v Tx wire supplying Q5 thru an FT37-43, no impact.

The power supply is a 12v 5amp commercial quality regulated overload protected supply.??
?
At this point I suspected a wiring or component error in the Q5 stage. I have checked all voltages and they are within? about 20% of the pre-driver stage which is similar.? All component values look the same.? I wonder if the output coupling capacitor Cs is 100pF or whether I might have put in a 100nF as these are SMD unmarked.? But I dont think that would be the cause.

if anyone is still reading, does this ring any bells for anyone? Has anyone else had a similar problem?? ?How is the crud finding its way onto the supply lines at such a high level?? Does this stage (Q5) exhibit high gain with some MMBT3904s???

If I cannot tame it at my next bench session, I may remake the Q5 stage on a postage stamp board with all new components and layer it over the top to ensure all components are correct. I may also use a 2N3904 rather than the SM MMBT3904, whose gain may be less predictable, although these SM transistors are behaving in the other transmit and receive stages, other than the first RF stage, which is the MPSH10 as per the design.?

Other suggestions welcome.

regards Paul VK3HN.??

?