[K2] Spurious emission in SSB

[G3WIE]

I built K2 s/n 253 many years ago, and it's been updated with the A-B kit and
all other Elecraft updates. It currently has the KSB2, noise blanker and
160m/2nd Rx options. The ATU is not currently fitted to make access easier,
though it has been.

I'm planning to use the K2 with transverters on 28MHz for the microwave
bands. Now that the Elecraft transverter interface is obsolete I plan to use
the radio with ~1W output and fit attenuators as required.

Looking at the Tx output on my spectrum analyser, I see a spurious tone
900kHz below the tx frequency which is only about 35dB less than the SSB
peak output at 1W; that frequency difference does not change with dial
frequency. The tone does not change amplitude with power setting over the
full range, with speech amplitude, nor with the carrier balance adjustment.
It's also present on 21MHz with the same interval.   It is not present in
Tune in SSB mode, or in CW mode at all.

I'd be grateful for ideas on where to start looking to find the source of
this as I can't reasonably use the K2 in this state..

73

Chris G3WIE



--
Sent from: http://elecraft.365791.n2.nabble.com/
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Don Wilhelm]

Chris,

I suspect that artifact is being introduced by the Speech Processor on
the KSB2 board.

Do you know any locals who have a K2?  The quickest way to verify my
suspicion is to swap out the KSB2 board.

BTW - TUNE switches to CW even if in SSB mode.

73,
Don W3FPR

On 8/30/2020 1:07 PM, G3WIE wrote:

> I built K2 s/n 253 many years ago, and it's been updated with the A-B kit and
> all other Elecraft updates. It currently has the KSB2, noise blanker and
> 160m/2nd Rx options. The ATU is not currently fitted to make access easier,
> though it has been.
>
> I'm planning to use the K2 with transverters on 28MHz for the microwave
> bands. Now that the Elecraft transverter interface is obsolete I plan to use
> the radio with ~1W output and fit attenuators as required.
>
> Looking at the Tx output on my spectrum analyser, I see a spurious tone
> 900kHz below the tx frequency which is only about 35dB less than the SSB
> peak output at 1W; that frequency difference does not change with dial
> frequency. The tone does not change amplitude with power setting over the
> full range, with speech amplitude, nor with the carrier balance adjustment.
> It's also present on 21MHz with the same interval.   It is not present in
> Tune in SSB mode, or in CW mode at all.
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[Elecraft mailing list]

Chris

Is the frequency separation exactly 900 kHz?  If a bit more, perhaps the oscillator in the KSB2 MCU (U1/Z1) at (nominally) 4 MHz is coupling into the IF (nominally 4.915 MHz) path on the same board, on the output side of the crystal filter).  Problems with grounding, bypass capacitors and long leads might do this.  I have a vague recollection that the instructions for early KSB2 board assembly did not involve grounding the filter crystal cases and later instructions did.  That shouldn't directly affect that coupling but it might generally help provide a generally more solid ground.

There are other 4 MHz oscillators in the K2 that might couple in, too, but since the problem only exists on SSB, it seems this one would be a more likely culprit.

Could re-aligning the 10 metre bandpass filter (K2 RF Board) help?

Generic ITU Radio Regulations require -43 dB dBc or better spurious levels at 1 watt output, so one hopes it could meet that. Of course if you are running a couple of hundred watts for 23 cm EME, you are required to do better!  Of course, you could get 3 dB better signal-to-spurious level just by going up to 2 watts.

I use a K2 (with an internally mounted 2 m transverter) in my 10 GHz setup....I suppose I ought to check for that issue, if I can borrow a spectrum analyzer.  But then...who would actually hear me, or have a problem with, a mW level signal at 10.3672 GHz and a beamwidth of 3 degrees?

73,
Steve VE3SMA
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[G3WIE]

Thanks to those who replied to my post. I've now reached the point where the
level of the spurious tone is low enough. Here's what I did.

Following Don's suggestion I isolated the KSSB2 speech processor from the
balanced modulator by removing C20 and grounding the free end of R2 via a
100nF capacitor. Unfortunately the spurious was unchanged; a problem in the
speech processor could have relatively easy to fix.

I followed up on Steve's suggestion with some careful measurements on the
spacing between spurious tone and K2 carrier in LSB and USB compared with
the difference in BFO frequency for the two. This confirmed that the 4MHz
oscillator in the PIC U1 was coupling into the KSB2 output. Some years ago I
had made a modification to the KSB2 between D2 and Q2  to reduce the peak
SSB amplitude so it matched the Tune level. I reverted to the published
circuit and the level of the spurious dropped to about -45dB ref the SSB
output which suggested that this signal path is sensitive to pickup. Next I
moved C31, C40 and C43 from the top to the bottom of the PCB and constructed
a 1/2" high screen connected to their Gnd pads to shield the PIC, the
resonator and U2 from the o/p circuits (D1 ... Q2). It made a marginal
improvement, not really worthwhile but I shan't undo it

Next I increased the output to 2 watts. The spurious tone's amplitude is not
affected by the power setting so this should win me 3dB. I also lifted the
base of Q2 off the PCB and joined it to D2 and R8 with a piece of wire 1/16"
off the top of the PCB. That way the PCB track that connects to Q2's base is
no longer used in case it's picking up the 4MHz. Finally I retuned the 28MHz
BPF higher in frequency to try and move the spurious tone towards the filter
stopband.

The spurious is now 50dB below the SSB output and I can't see how to make
any further improvement. There are other spurious outputs at that level so
there's little to be gained in any case. I don't believe I have found the
main mechanism for the coupling but as they say "sometimes good is good
enough"

73

Chris G3WIE




--
Sent from: http://elecraft.365791.n2.nabble.com/
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]