I spoke too soon !

I spoke too soon with my recent Elecraft K3 repairs (click). The radio happily sat on the desk for a few days behaving itself nicely when wham the fault returned. Drats I hadn’t fixed it !

So the covers came off and when power was applied the fault had cleared again, here we go again *sigh*.

This time I was suspicious of the cabling between the reference oscillator and the synthesiser. So I spent a bit of time pushing, pulling and prodding this cable. At one point I had slowly bent this cable sharply down towards the main PCB in the bottom of the radio when the LED on the Synthesiser board went nuts. By nuts it started flickering on and off, this time it did not flash or go out. When I slowly released the cable the LED would flash slowly and transmit was restored when the radio was power cycled.

Right so were getting somewhere could the cable could be intermittent ? So I duly dug through the spares box and found a spare 150mm long piece of RG-174 with the appropriate Taiko Denki connectors and replaced the short 70mm cable. Since the replacement cable was longer than the original, I’d put a small loop on the end so that it sat clear of it’s surroundings. I left the radio with the covers off for a few hours with the radio on test, pushing and prodding things and it seemed fine.

After a day and half with the covers off I then screwed the covers back on again. Within 2 hours the radio faulted. What on earth was going on ?

My next move was to pull out a test jig I’d made years ago that allowed me to put a decent amount of RF power (50-70W) through the short 70mm cable into a dummy load. With RF being applied I then push, prod and pull the cable and watch an inline power meter for any reflected power. If the cable is intermittent having RF with some power behind it will quickly show problems.

So it was definitely time for a coffee and a re-think.

I spent a bit of time lifting and replacing the lid when I decided to bring ones face down to see what was happening when I spotted something. The speaker is mounted to the top panel of the K3 and as the lid was closed the speaker would interfere with the cable between the reference oscillator and synthesiser; pushing and holding it in a downwards direction.

This then got me thinking, has this interference between the speaker and cable flogged out the Taiko Denki sockets ? A quick check of the circuit shows that sockets J1 and J2 are effectively in parallel, well they will out the same level from both sockets. So I moved the reference oscillator from J1 to J2 and put the radio back on test, this time with the short cable again. I could not get the radio to fault for a further day and a half. All of this silliness had gone away. However unless I can explain what has happened, well I didn’t fix it.

The easiest way to to find out of the connector was flogged was to pull the reference oscillator out of the radio and get it under my microscope. I wish I had a camera attachment for my microscope, but needless to say this was the culprit. I could clearly see that the fingers within this connector had been pushed apart and that contact with the centre pin was intermittent. I can see two ways in which this was being done;

  1. The interference between the speaker and short 70mm cable had over time worked the Taiko Denki plug in the socket, eventually working the fingers open;
  2. That some knuckle dragging rock ape repairman (ie yours truly) inserted the plug incorrectly chasing this issue and opened the fingers

Right now option 2 is the most likely explanation. So for the time being I’ve left the reference oscillator plugged into J2 and have ordered more Taiko Denki sockets and plugs from the interwebs. With luck I’ll find a crimp tool, otherwise I’ll be making my own, Hmm that sounds like a good blog post in the future.

I’m fingers crossed that this fault is now fix and it’s time to find a camera for my microscope.