Re: my K3 has some issues-SOLVED- Low Power Amp MOSFETs

[Chet S]

Replacing the final output MOSFETs in my K3's  low power amp appears to have
fixed its transmit power issues. Thank you everyone on the K3 forums for
your shared advice and experiences.

The "3 screw" tests were pointing to a bad output transistor in the LPA. The
idle current increase on the K3 meter when putting the radio into XMIT (not
TUNE) was only 310 ma, where at least 600ma was expected, again pointing to
a FET being open.

To be sure of getting the right vintage of FETs, I ordered a couple of them
(and thermal insulator) from Elecraft. My order was shipped that same day
and arrived via priority mail 2 days later. Great service.

Following the procedure described by others, I replaced both output MOSFETs.
Since one had been inoperative, the other had been carrying the whole load
to produce output so may have been overworked, overstressed and hurting.

Here are my additional comments on the repair process:

1) The Elecraft document "KPA3CONMDKT MOD Kit Manual" has a wonderful
description of how to remove and replace the 100W amplifier to allow access
to the LPA's adjustment potentiometers.

2) Turning the power to zero is a good safeguard when doing the bias
adjustment. I also made sure the radio was in USB or LSB so no power output
would be commanded in any case.

Now with the FETs replaced and the bottom cover back on, the moment of
truth..

Three LPA potentiometers turned down.
Turn first pot, +200ma.
Turn second pot, another 200ma.
Turn third pot, yet another 200ma.
Success!

Put the 100W amp module back in and connected to the Elecraft K3 utility.
All 3 transmitter gain tests passed, and now the power setting is stable on
all bands.

On 40M where the reported SWR into a dummy load before the repair was 1.4
instead of 1.0  was now down to 1.2. Still do not know why but I can
certainly live with that.

So. why did this even happen in the first place?
I can only speculate.
Older generation MOSFETs were difficult to keep in their safe operating
region without snubbers. Did some transient zap it?
Power semiconductor chip bonding is subject to failure after a high number
of thermal cycles, and this has been a contest radio.
In January during the VHF contest, my 6M antennas became badly detuned with
wet snow and ice accumulation. This led to a high swr propagating thru the
transmit chain to the K3 which suffered a total loss of power output on 6M
with a very high swr indicated. After a rest, the K3 recovered, but did that
event set this failure in motion? GOK.

And a closing suggestion- if your K3 is working OK now, save a copy of its
configuration: this may turn out to be quite handy in the future.

TU all and 73,
Chet, N8RA



-----Original Message-----
From: [hidden email] <[hidden email]> On
Behalf Of Chet S
Sent: Tuesday, November 10, 2020 4:02 PM
To: [hidden email]
Subject: [Elecraft] my K3 has some issues

Good afternoon K3 folks,

I am new to this list and honored to be accepted.

Recently my K3 has experienced a couple of issues. I have been unable to get
a response from Elecraft e-mail or telephone service regarding the
situation. It's been almost two weeks now so thought I'd ask here if anyone
has some suggestions for me to try myself.

Here is the main concern:

When getting ready for CQWWDX SSB a few weeks ago, I went through setting
the transmit power on each band to drive an amplifier properly. The power
display seemed a bit sluggish and had a delay in indicating the SWR. Worse
yet, when I got to 40M, I could not crank the power up past about 30 watts;
it would momentarily display a higher number but then fall back to 30w. All
other bands cold be set higher.

First I reloaded all the current firmware in the main and subreceiver. No
change to 40M.

So next I went to the K3 Utility ver 1.20.1.15 to do a transmitter gain
calibration. Starting with the 5W calibration, it stepped thru the first few
bands OK but then stopped (I think on 40M) and displayed "ERR TXG" and the
message "calibration settled at 3.6W, TX gain failed because transmit power
did not reach the expected value".

In anticipation of having to send the radio to Elecraft where it might be
totally reset, I manually started recording all the configuration settings.
When I looked at my filter settings, I was shocked to see that filter 1 in
the subreceiver was set to 1.35. Huh? It should have been 2.80 !! Was my
config info corrupted? I reset that to 2.80 and then for the heck of it
tried again to set the power on 40M. That now  worked "OK"  and the swr and
pwr display during "tune" showed instantly on all bands. Something had
changed.

Next, I decided to rerun the TX calibration gain routine. This time if
successfully completed the 5w and 50w routines. However, with a good dummy
load, 40M displays an swr of 1.4, while the other bands are 1.0 or 1.1, so
something is still a little funny on 40M. What might that be?

I have now been considering doing a total reset of the radio using the EE
INIT function and then the self calibrations. Is this wise?

Any experiences or suggestions appreciated.

73,
Chet, N8RA

______________________________________________________________
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
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]