The Elecraft K3 is a very well known HF transceiver with a great reputation and it follows quite a different philosophy from what (mainly japanese) competitors were offering at its time. This particular unit, from my good friend Angel EA4XT started giving “ERR 12V” and the output power dropped to a few watts each time. This got more and more frequent until the point he asked me to take a look at it. It turns out there is a surprisingly simple fix.
A quick search revealed this is a common issue, caused by the high DC power required by the PA unit being directed thru a series of pin strips not really designed for that. I can’t possibly get to know what decisions led Elecraft engineers to choose this design but in many other radios the unswitched 12V power to the PA unit comes from a thick red wire with a beefy connection.
In any case, I needed to confirm that this was actually the case and it did not take a lot of effort. Transmitting high power CW was clearly upsetting the connector and creating a hot spot in one of the pins top left row (the other two were probably not conducting at all):
There is a vertical pcb called KPAIO3 that sits between the main PCB and the PA PCB, and it has the same type of pin-strip connectors on both sides, six of which carry the unswitched 12V, with high current, to the PA. It also has a shunt resistor and one IC to monitor the current going to the amplifier. That rules out bypassing this PCB completely, which was my first thought.
Note the two unused PCB pads with vias on the sides of the shunt resistor because they are very interesting as we will see later:
The six last pins on the PA unit looked very dark, not surprisingly they were not conducting well and were getting hot. The plastic supporting the pins even melted:
The female strip on the KPAIO3 was also damaged:
Main board pins
Even the pins on the main PCB had melted the plastic. Note that those six pins that carry power are golden and not silver:
That is actually the official fix from Elecraft(link), to replace the strips with ones with gold pins that should handle higher current.
Fixing the problem
There is a 20 amp circuit breaker on the back panel with both terminals connected to the PA UNIT (red crimp connectors just outside of the picture!):
The “input” terminal of the breaker can be very easily be moved to the upper unused pad of the KPAIO3 that we talked about before, and we just bypassed the top pins that were causing trouble (top left corner, gray cable):
To bypass the bottom connector, we use a short cable (red in this case) to connect the unswitched 12V on the PCB from the ferrite choke to the bottom pad of the KPAIO3. You can now see very clearly that we used the two pads at the sides of the shunt resistor on the KPAIO3 to bypass the top and bottom connectors:
We assemble everything back and we check we can transmit 100W for as long as we want and the error will not appear:
Someone will come up with a different idea, but for now this looks good to me and this K3 is going back to his owner. Thanks for reading.