Recently a fellow ham sent me a QMX rig to fix.??This is to be my 100th QMX that I have fixed - assuming that I accomplish that with this one.???I wanted to share the journey.??I found this one interesting.??This is a Rev 2 QMX with rev 14 firmware.??Wound for 12V.??This is the first part of a story - the diagnosis and how I arrived at it.??I wanted to share because this is more interesting than most and it may help someone else.??More will be posted later.

?

?My usual routine with a new rig is to disassemble, inspect, clean??and inspect again.??I use a 10x stereo microscope for everything with very bright side and overhead lights.??I scrape flux off, use 99% isopropyl alcohol to clean with an extra soft toothbrush.??Then I inspect again.??I can fix almost all rigs this way.??There are always solder blobs, solder balls, cold solder joints in every rig that comes to me.

?

This rig came in very clean.??No excess flux, no bad solder joints, no solder splashes.??It did have a few long pins that ended up shorting to the case, adjacent boards and wreaked a bit of havoc.??I fixed those before proceeding.

?

After applying power at 7V with current limits, it came up only with top row black boxes.??No backlight, nothing on PUTTY.??Looking with a scope I could see the 3.3V rail was good and 3.3V PWM was within range.??But the 5V rail was sitting at about 5.8-5.9 V and the 5V PWM was at about 78% as measured on my scope.??Not Good!??But I knew the CPU was running enough to produce PWM signals for the SMPS boards.??Thats good!

?

I tested the SMPS power boards on my homemade test fixture.??They were fine.??I tested for near shorts or low resistance on the 5V rail.??It looked fine also.

?

I put the SMPS boards back into the rig and plugged in the controls board.??I also installed the jumper to force the QMX into firmware update mode.??That worked.??The QMX came up wit the “disk drive” on my Linux desktop.??I installed rev 16 firmware knowing that Hans had fixed a bug in initializing the EEPROM.??I thought the forced factory reset upon installation of Rev 16 might shed some light.??It did.??Now the rig came up immediately into Diagnostic mode.??It’s the only screen that will come up in PUTTY if something is seriously wrong.??I could see that the 5V rail was at 78% duty cycle as I saw on my scope but the reported voltage to the CPU was only 4.8 volts.??The CPU was trying its hardest to reach 5V but it maxed out at 78% and couldn’t get there.??I measured the ADC_5V signal on the board and my scope read 2.4V.??That is consistent with the diagnostic screen reporting 4.8V since there is a 2x divider in the circuit.

?

ADC_5V only goes from the SMPS board to the CPU.??Something was clamping it at 2.4V.??It had to be a short on the SMPS (no, because I tested those separately), the main board in the traces (possible, but unlikely), or in the CPU (UGH!).??I lifted pin 32 of the cpu and confirmed that indeed the CPU was clamping that line at 2.4 volts.??After lifting that pin, the board showed 2.95V for ADC_5V.??And that CPU pin 32 in the air showed 2.4V.??Proof enough for me.

?

So this means that replacing the CPU is the only option.??This rig has sentimental value to its owner, so he said please proceed.??I have a few bare boards with programmed CPUs that I have just for this kind of occasion.

?

Tomorrow I will proceed with the operation.??I have to accomplish that in the morning when I am at my best and after only 1 cup of coffee.??I’ll post my progress here.

?

73 for now

Jeff

W1NC

I’m sure it’s due to an inadvertent short circuit somewhere along the path of assembly. ?I suspect the extra long pins that weren’t trimmed short enough to the board shorted against the case, ?this is a very tight assembly that needs close attention to lots of details. ?I never blame ESD. ?It’s almost always some small human error. ?

Jeff
W1NC

开云体育

That is an interesting failure mode - deep inside the cpu. After all that meticulous soldering I can understand the sentimental attachment. Good luck and thanks for all you do! 73, Sandy KB3EOF

Success! ?After a one hour operation, the patient lives. ?I was able to remove the bad cpu, remove the good cpu from a donor board, being careful to not mix them up!!!! clean up the pads on the board, clean up the legs of the good cpu and finally solder the new cpu in place. ?Power on and all is good. ?The rig has a few other minor problems that I’ll fix up and ship it back to its home QTH.

my technique for this is to use a bead of flux and low temp solder around the entire cpu. ?I move my iron around the perimeter until the chip is loose. ?I then slide it off the board.

I clean the legs off the good cpu by placing each side of the chip on a flux soaked piece of solder wick. ?I apply heat and all solder is wicked away. ?

Soldering in the new cpu is done with flux (I love flux), thin solder and a very small iron.

73
Jeff
W1NC

开云体育

开云体育

All I can say it’s a good thing I had lifted the leg of pin 32 of the cpu that was NG before I removed it. ?That saved me. ?

Jeff
W1NC

Final installment…

Today I finished up this QMX. ?T501 and T507 didn’t look right to me. ?Fortunately the PC board pads seemed to be in great shape. ?I removed each transformer, rewound then and soldered them in.

? T501 looked as if too much heat had been applied when soldering it in. ?The enamel had melted and I suspected shorts. ?

T507 looked like the primary and secondary could short easily since too much enamel had been removed. ?Maybe I could have put some kind of adhesive to stabilize it. ?But rewinding it seemed the better approach. ?I always use green wire for the 10 turn portion and red wire for the single piece. ?Looks great that way.

Anyway, the big moment of truth. ?Would everything else work? ?After all there was 5.9v on the VCC rail. ?I know the AD converter is especially sensitive. ?I powered it up at 7v and ran through the audio, Rf and image tests. ?All very good. ?The audio output was even very good in the headphones. ?I powered down. ?That worked too.

i then adjusted my power supply to 12v and powered on for TX testing. ?I got 4 to 6 watts out across all the low bands. ?Very good. ?I’m getting better and better with each RWTST transformer that I wind. ?Hi hi.

Then I remembered that Hans released rev 17 last night. ?I didn’t see any show stopper bugs posted. ?So I loaded up rev 17. ?Went through all the tests again. ?Super! ?Love the AGC! ?Thank you Hans.

The final step is putting it in the case, inspecting for clearances and shorts and one final test on the air. ? That went very smooth. ?A very rewarding project.

the post office had closed for the day. ?This will ship out to its owner first thing next week.

By the way, I’m trying to get an OBS system working on Ubuntu with video so I can record some of these adventures. ?That has given me more trouble than I anticipated so far. ?But I’ll get it - soon I hope.

73
Jeff
W1NC