K3 Autotuner malfunction

[Rob KK4R]

I have been having problems with my K3 auto tuner using an antenna that has worked well and been tuning easily for nearly all bands.  My first reaction was to assume the antenna, an 80m horizontal loop, was the problem, but I am now convinced otherwise.  To prove that, I did the following:
 - took the antenna down, refurbished it, and replaced the feed line (not wasted energy);
 - tried other baluns;
 - confirmed that my LDG tuner has no problem tuning; and
 - tuned it manually with the K3 tuner.

The problem behavior is this.  On 40m and 80m, the K3 will not find a decent SWR (<1.7:1) when auto-tuning.  On 80m it has done fine tuning one direction across the band and failed going the other direction over a significant portion of the band.  This is repeatable.  On 40m, the SWR for the antenna system at the rig end is only about 1.7:1 at the low end and 1.4 at the high end.  At the low end, the K3 tuner will not find a match under when the ATU TUNE button is used, except in particular circumstances.  Using the menu system to check tuner values, it ends up with the capacitor connected at the antenna side at a low value and with a low value of series inductance.  Effectively this is near the bypass mode.  Hitting the ATU TUNE button to cause a long tune cycle makes no difference.  Manually, the capacitor can be switched to the transmitter side, and a match near 1:1 is easily set by adjusting L and C.

Today I cleared the memory on 40m and manually tuned at a low frequency in the band.  Then, moving up the band at 2KHz intervals, the auto tuner was able to quickly fine tune the match and a full set of memories was created.  This means that I can work 40m without a problem without hitting the ATU TUNE button, but if the ATU TUNE button is hit after a significant frequency change, the match is lost.

Therefore, the key problem with the auto-tuner behavior is that when ATU TUNE is used in a case where there is no nearby memory, or a large frequency change has been made since the last transmission, the tuning algorithm apparently does not try switching the capacitor from the antenna side to the transmitter side, and the tuner does not thoroughly look for a match.

Is this abnormal behavior for my unit, or is this a quirk of the system?

Thanks,
Rob, KK4R

[Don Wilhelm-4]

Rob,

I would suggest that on 40 meters and 80 meters, your loop is presenting
a very low impedance to the shack end of the feedline.
Tuners of all types do not work well into very low impedance loads.
Add or subtract somewhere between 1/8 and 1/4 wavelength of feedline at
the problem frequencies and see if that helps your problem.  Of course,
that solution possibly could introduce problems on other bands, but it
is worth a try.

73,
Don W3FPR

On 7/3/2015 2:29 PM, Rob KK4R wrote:

> I have been having problems with my K3 auto tuner using an antenna that has
> worked well and been tuning easily for nearly all bands.  My first reaction
> was to assume the antenna, an 80m horizontal loop, was the problem, but I am
> now convinced otherwise.  To prove that, I did the following:
>   - took the antenna down, refurbished it, and replaced the feed line (not
> wasted energy);
>   - tried other baluns;
>   - confirmed that my LDG tuner has no problem tuning; and
>   - tuned it manually with the K3 tuner.
>
> The problem behavior is this.  On 40m and 80m, the K3 will not find a decent
> SWR (<1.7:1) when auto-tuning.  On 80m it has done fine tuning one direction
> across the band and failed going the other direction over a significant
> portion of the band.  This is repeatable.  On 40m, the SWR for the antenna
> system at the rig end is only about 1.7:1 at the low end and 1.4 at the high
> end.  At the low end, the K3 tuner will not find a match under when the ATU
> TUNE button is used, except in particular circumstances.  Using the menu
> system to check tuner values, it ends up with the capacitor connected at the
> antenna side at a low value and with a low value of series inductance.
> Effectively this is near the bypass mode.  Hitting the ATU TUNE button to
> cause a long tune cycle makes no difference.  Manually, the capacitor can be
> switched to the transmitter side, and a match near 1:1 is easily set by
> adjusting L and C.
>
>

______________________________________________________________
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]

[Rob KK4R]

Don,

At first (months ago), I looked at the antenna impedance at the rig and thought that it was a problem.  It was an 80m horizontal loop fed with 450 Ohm line.  It would be problematic to change the feed line length, so I did some calculations and found that 300 Ohm line would provide a better match on most bands.  The line was changed and the loop length trimmed a bit (not enough).  The VSWR was measured across the HF spectrum with a good network analyzer.  The result looked good.  On 80m, SWR ranged between about 7:1 to 3.5:1.  On forty, it was between about 4:1 and less than 2:1.

After reading your message, I checked the impedance and SWR again a couple of ways, and although it isn't as good as from the network analyzer, the data agrees pretty well.  If anything, the SWR looks better now.  The bottom line is that the impedance isn't bad, and I barely need the tuner on 40m at all.  In fact, the tuner might be confused by the low SWR, I suppose.  Using a tuner with a good load breaks the rule, "If it ain't broke, don't fix it!"

The antenna impedance magnitude is between 25 and 80 Ohms in both bands.  I could email you the SWR plot I made when the new feed line was put up, if you are curious.  Too bad I didn't measure S-parameters.

73,
Rob

[Rob KK4R]

After getting some data, and considering what others had to say, the answer seems to be that there is no malfunction.  The behavior that I am seeing is likely because the mismatch with the antenna is marginal.  At least my K3 is not very sensitive to VSWR less than 1.5:1, and it reports a VSWR of 1.0:1 when other devices are measuring something more.  In general it reports a lower VSWR than any analyzer I have tried.

Addressing my "problem" on 40 meters, there are good values in the tuner memories for the whole band, but if I change from the high end of the band to the low end, the VSWR sensed by the K3 is not so bad, and I suspect that the K3 doesn't bother to switch to the values in memory for the new frequency.  I might like to see less than 1.5:1.0, but the K3 isn't bothering with it.

On 80m, my antenna actually does need a bit of tuning.  There has been some trouble auto-tuning to get a good match with the tuner around the part of the band where the tuner capacitor needs to switch from the antenna side to the transmitter side or vice versa.  This is just a theory, but possibly the K3 sensing of the match is more sensitive depending on which side the capacitor is on.  Maybe it makes a mistake sometimes when the SWR is relatively low, like less than 3.0:1.  The capacitor seems to end up on the antenna side when there would be a better match if it switched to the transmitter side.  I may mess around on 80 and see if I can learn anything else or confirm this.

Regardless, my tuner works, and if you use it correctly, there is barely a hiccup.

Rob