开云体育

Thanks for the report.
Very informative!

I think the consensus is that replacing L5,L7 should reduce coupling from the power amp?
back into the 45mhz IF amp where it takes another trip through the D1,D2 mixer.
Raj had a specific manufacturer in mind:??/g/BITX20/message/61261
They have to be well shielded, commonly available surface mount inductors did not help at all.



Here's a brief recap regarding what other changes and why between v3 and v5,
as best I recall:

As mike gain is increased, Allison has noted that spurs increase disproportionately
due to distortion, especially in the 45mhz IF amp.? The mmbt3904's there are being
asked to work well beyond their gain-bandwidth spec.? So we have other stuff besides 45mhz
going into the first mixer.
On v5, Farhan added a low pass filter with L31,L32,C205, I'd assume it's there mostly to?
remove the 90mhz 2'd harmonic due to distortion in the 45mhz IF, which mixes
with the VFO to produce a very strong spur at? 2*45mhz - (VFO=DialFreq+45mhz) == 45mhz - DialFreq
where DialFreq is the desired transmit frequency.

He went to higher frequency transistors in the 45mhz IF amps to reduce distortion
and give more gain.

He also changed R26 from 470 ohms to 220 ohms, at the output of the 45mhz IF amp for tx.
I believe the reason for this was to give a higher level 45mhz signal into the first mixer,
helping swamp out the 12mhz crosstalk from CLK0 into CLK2.

R100, R106, R110 where CLK2, CLK1, CLK0 go into the mixers have changed to larger values,
probably to reduce loading on the si5351, and thus crosstalk between the three channels.

The routing for the three relays that switch between the four transmit LPF's has changed to
reduce the transmitted harmonics.?
The 2'd IF has been moved from 12mhz to 11.059mhz to avoid the possibility of nasty audio tones
due to beat notes of the BFO with clocks on the Nano.
These two changes both require minor adjustments to the firmware.

?I see these changes between v3 and v5 down in the PA:??
? ? c81=470pf r83=2.2 R97=R98=220 C261=C262=0
which I assume is mostly to get more power out on the high bands.

Stuff from the M2 relay contact on through the volume control and audio power amp, out to the speaker
has changed considerably across v3,v4,v5.
v3 had the TDA2822 which worked fine, except that parts from one particular clone manufacturer were exploding,
so v4 when to a discrete push-pull audio amp.
But v4 had issues with crossover distortion and low max volume, so v5 punts with the tried and true LM386.
v4 and v5 have a 2n7000 (perhaps its a 2n7002?)? at Q74 added to reduce PTT pops to the speaker.

I've undoubtedly missed a few changes, but that should be most of it..

Jerry

Mike

thanks for posting your changes and measurement.? Nice data that the Axicom relay provided around 10 dB improvement, that is what I need on 40 and 30 meters.? I have Gordon's relay board installed, and Saturday is our club session - I imagine this set of new relays will do at least that good.?

for the higher band, non-harmonic spurious, some form of added 45 MHz filtering should help.? first I tried a xtal filter stage - but it disrupted my 40m CW output (I could restore output only by touching L5 - yes the reason for considering shielded inductors).? I added only an LC filter similar to v5, not sure of inductor Q but maybe around 70 - so it provides more than 10 dB I need in the model -- real life, well we will see - I am not quite as hopeful here.?

in our initial measurement of many ubitx rigs -- there is appreciable variation even though these v4 all came together.?

Curt