[Elecraft] K3 #1851 is (mostly) alive!

[Frank R. Oppedijk]

Hi all!

Another K3 has been built! K3 #1851 was ordered on 17 June, shipped 
on 30 September and was delivered and subsequently built today, 3 
October. It's a K3/100, with KAT3, KBPF3, and KXV3. It has the stock 
2,7 kHz, FM and 400 Hz filters. The KRX3 and accompanying filters are 
on back-order (expected in about two months).

All calibrations and tests passed, but while trying the rig out on 40 
meters, a nasty problem turned up. Transmissing at full power let 
some of the smoke out. I understand the desired operating practice is 
to keep the smoke inside at all times. After removing the top cover 
and doing some more tests at reduced power (30 to 50W), I found out 
that one toroid on the KAT3 board is getting very hot. It is the blue 
toroid, possibly labeled K16 (can't quite see) which has 3 windings 
on it. I'm guessing this toroid is part of the circuit sensing the 
reflected power.

The problem occurs only when transmitting on 40 meters into my 
antenna (a 2x24 meters doublet), using the KAT3 in AUTO mode. The 
MFJ-949E external ATU sits in-line, connected with about two feet of 
coax, but is switched to bypass mode. Interestingly, the MFJ does 
show a very high SWR (1:6 or worse), while the K3 shows a perfect 1:1 match!

The problem with the heating occurs only on 40 meters; on other bands 
it stays cool. Still, on these other bands, the external tuner shows 
high readings too. When transmitting into the MFJs dummy load, both 
the K3 and the MFJ agree that the SWR is perfect.

Does anyone know what could cause this? FWIW, I am still using the 
factory-supplied firmware (MCU 2.38 and DSP 1.90).

Thanks in advance,

Frank PA4N