¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Topics

QMX+ lacks a PWR_HOLD signal - Putty works holding down PWR_ON

 

Friends and fellow QRP-labs builder addicts, it is time to ask the group for ideas.

I have read and ¡°liked¡± a number of articles here, and until now I have not been stumped. ?I have a small pile of working QRP-Labs rigs, the original QCX on 40, 2 QDX¡¯s, and an 80-20 QMX. ?That last one was so tiny and cramped I¡¯m amazed it came through unscathed, though I never was satisfied with power out. ?I was about to swear off these and Hans had to come up with the QMX+. ?It was a ball to assemble - I drug it out for about a week. ?I did have a problem with the received SMS-PS #2 board, as the little 330 coil was broken in shipment, dangling by the wires. ?Since wires looked good I read in the forum that folks were successful gluing them. ?So I completed the rig with the inductor glued to look normal.?

On power up the QMX+ didn¡¯t come up, but no smoke (whew!). ?I suspected the "repaired" SMS-PS #2 board, so I pulled the board from my working QMX and it seemed dead too.? I did notice the Flash drive appeared if I held the power down for a while. ?I thought then that my original SMS-PS #2 board with the broken inductor was OK. I tested it quickly in the old QMX, but it would not work at all. ?I didn¡¯t wait around to see what the voltage was, because I thought it might damage the QMX, so I put the good board in, and the old QMX still worked.? Now I¡¯m borrowing it for the QMX+ to finish the build. ?I proceeded to load the firmware while holding PWR_ON, and it came up like any other upgrade. ?I could get into putty that first time, while I was still holding PWR_ON. ?But once I let go I was only able to get it to wake at all if I held down PWR_ON, and before it finishes initializing it goes to Shut Down!

I can not get into putty if I let the rig start normally because as soon as I let off power it dies. ?If I hold power it goes to Shut Down! ?I Can get into putty if I jumper force the firmware load, toss the firmware and erase it, then load an old version, then update it with current version. ?If I keep holding all this time I finally get into putty. I can change the configuration model from the default to QMX+, I can go into diagnostics and because I¡¯m holding the power down the buttons are all red, but the tune encoder will work as a button. I am pretty sure holding power interferes with the other buttons. All the voltages in the diagnostics view are green. ?So it is Almost live.?

I have?
Vin - 8V
Vcc - 5V
Vdd - 3.3V

Current at startup 208ma - It shows LCD power, Initializing Audio, then the next one starts and it goes to Shut Down!

In firmware mode, when I get it into Putty, I am pulling 164ma. ?

So, what it doesn¡¯t get is a PWR_HOLD signal. ?I have put a ¡®scope on that line and watched it go up about 0.8V when PWR_ON is pressed. ?I have continuity between the JP101 pin 1 and pin 53 of IC203. ?I have my good QMX to compare to, so I removed the SMPS boards and checked PWR_HOLD to ground. ?The working QMX shows OPEN, and the QMX+ shows 5 Meg resistance. I cannot tell from the schematics I have if there is any place that signal comes to the surface besides JP101 and the IC. ?I have scraped around the solder for the pins of JP101, cleaned all the flux - I cannot see any kind of solder bridge there. ?I¡¯ve examined the thing wearing all my magnifying head gear, and I¡¯ve used every color filter on my magnifier phone app to get good looks around the JP101 pins. Looks great.

Should the PWR_HOLD line show OPEN to ground, or is QMX+ different?? Any ideas of what I might try next?

Sorry for the long story but I¡¯ve been at this for nearly two weeks now.... there was a lot to tell

Thanks in advance for any ideas,

Jeff - KS7N
 

Hi Jeff,

I¡¯ve seen this exact situation on a QMX that I repaired a few months ago. ?It was the cpu itself that couldn¡¯t assert the signal. ?Somehow one pin of the cpu was bad. ?Very weird. ?I replaced the cpu and all was good. ?I gently lifted the pin of the cpu to verify that.

I'm not saying that is what¡¯s wrong with yours. ?It¡¯s more likely a short somewhere on the pcb trace on the board. You may be able to trace the path from the diagram in the assembly manual. ?But, Its low resolution may make that difficult. ?Also look under a microscope at the cpu pins and the pins of the SMPS connectors to make sure there is no tiny wire piece or solder there.

good luck.

--
73
Jeff Moore
W1NC
 

Thanks for the input Jeff (W1NC) - I think lifting the leg might be the next step. If the voltage behaves the same, I'll need a new IC203..... If the leg actually provides the 3.3V HOLD voltage then a short it the board is likely.? I just need to figure out the path.?? My guess is something in the board is dragging it down, since the 'scope shows it jumps a little as it tries to go high.

73,
Jeff - KS7N
 

While not a true power hold function, it can be achieved by setting wpm to zero and pressing you key/ paddle.

73, Rick K8BMA?
 

Hi Jeff,
?
I wanted to give you an update since your previous post was so helpful.? You may remember this issue from back on June 30.? I've been letting this QMX+ get the better of me.? I did prove that the cpu was bad as you described: pin 53 behaved exactly the same, only lifting? about 0.7 briefly on power up, with the pin floating - and the trace in the board between pin 53 and the SMPS board was continuous and not shorted.? And of course I checked the behavior on a working QMX, so I knew I was on the right path. I ordered a chip from Digi-key, along with some aluminum tape to protect surrounding components from my heat gun.
?
I removed the chip easily enough, and cleaned everything up, but I knew the soldering would quite difficult.? This is about twice the pin density of any SM chip I've done (and that was probably 15 years ago, when my hands were steadier and I had full use of both eyes) so I took some time getting my courage up for the replacement phase.? I did struggle getting the thing installed.? I just use my phone magnifier to see the pads, but I suspected I had somehow damaged the chip.?? I did get a lot of practice though, and I think I have a process that will work on my next try.? The trick was finding a way to hold the chip still.? I think for my next try I am going to pull this chip, build up some solder on the pads, then with the new chip held still over the solder loaded pads, essentially heat each pad one at a time to flow the solder over the pin.?
?
I was debating on ordering one more chip from Digi-key, or admitting defeat, then I saw your post from the 16th.? Now I know for sure it was never going to work. "You will need a cpu that has the boot loader programmed onto it for the replacement."?? I suspect putting in generic CPU chips without this boot loader will never get me anywhere.? I guess I have to ask Hans for a chip to get one with a boot loader?
?
Thanks again for the help.? I am still hoping I can fix it up myself. :)
?
73,
Jeff - KS7N

?
 

Yes. ?You need a QRPLABS encrypted boot loader programmed on the CPUs. ?I have some on hand for the repairs that I do. ?Ask Hans if he has one he would sell to you.
?
when soldering in a new chip, I do this:
1) clean off all pads with flux and solder wick. ?Be careful not to be too aggressive moving on the 100 pads. ?They are delicate.
2) clean off residual flux with 99% IPA
3) coat all pads with a viscous no clean flux
4) place new cpu on pads - getting pin 1 correct!!!
5) fiddle until all the pads are aligned and near perfect with equal space?
6) with a really tiny tipped iron, coat the tip with a very small blob of 60/40 solder
7) press down on top of the cpu with a finger or small instrument so it won¡¯t move. ?Keep the pressure here and check all pins
8) touch the iron to a non GND pin on the cpu anchoring that pin
9) repeat tack soldering a pin on each side until all 4 sides have one pin tack soldered.
note: if you have solder bridges at this point - don¡¯t worry. ?We will get them later
10) coat all pins with fresh flux carefully. ?Dragging a tip across the pins will bend them. ?Not good!
10) now with a larger iron (I use 1.2mm chisel tip) drag a bead of 60/40 solder gently across a row of pins.
11) repeat for the other three sides
12) I now put another bead of flux on top of the soldered sides.
13) with my 1.2 mm tip I gently touch each pad and the edge of each cpu pin. ?The solder will reflow and clean up many bridges.
14) if you have stubborn bridges, use a 1/2 second dab with solder wick and my iron.
now inspect everything under high magnification and bright lights. ?look for fully soldered pins and no bridges (x100). ?Gently poke the side of a pin that looks suspect to make sure it¡¯s soldered down well. ?I use a dental tool for that
?
good luck
--
73
Jeff Moore
W1NC
 

Excellent Jeff! Thanks! Can you do something similar for UC removal?

Tnx es 73,
Randy, KS4L?


On Sun, Aug 18, 2024, 3:08?PM Jeffrey W Moore via <jeffreymoore=[email protected]> wrote:
Yes.? You need a QRPLABS encrypted boot loader programmed on the CPUs.? I have some on hand for the repairs that I do.? Ask Hans if he has one he would sell to you.
?
when soldering in a new chip, I do this:
1) clean off all pads with flux and solder wick.? Be careful not to be too aggressive moving on the 100 pads.? They are delicate.
2) clean off residual flux with 99% IPA
3) coat all pads with a viscous no clean flux
4) place new cpu on pads - getting pin 1 correct!!!
5) fiddle until all the pads are aligned and near perfect with equal space?
6) with a really tiny tipped iron, coat the tip with a very small blob of 60/40 solder
7) press down on top of the cpu with a finger or small instrument so it won¡¯t move.? Keep the pressure here and check all pins
8) touch the iron to a non GND pin on the cpu anchoring that pin
9) repeat tack soldering a pin on each side until all 4 sides have one pin tack soldered.
note: if you have solder bridges at this point - don¡¯t worry.? We will get them later
10) coat all pins with fresh flux carefully.? Dragging a tip across the pins will bend them.? Not good!
10) now with a larger iron (I use 1.2mm chisel tip) drag a bead of 60/40 solder gently across a row of pins.
11) repeat for the other three sides
12) I now put another bead of flux on top of the soldered sides.
13) with my 1.2 mm tip I gently touch each pad and the edge of each cpu pin.? The solder will reflow and clean up many bridges.
14) if you have stubborn bridges, use a 1/2 second dab with solder wick and my iron.
now inspect everything under high magnification and bright lights. ?look for fully soldered pins and no bridges (x100).? Gently poke the side of a pin that looks suspect to make sure it¡¯s soldered down well.? I use a dental tool for that
?
good luck
--
73
Jeff Moore
W1NC
 

I assume you mean IC removal¡­
?
I have tried a bunch of different ways. ?But I¡¯ve settled on the method that uses low temp melt point solder. ?I use the stuff that melts at 80C. ?This is the bar type solder that I use because I use a lot of it. ?Wire type is a little easier to manage in the beginning.
?
?
?
I do it Like this:
?
1) coat every pin of the IC with viscous flux
2) melt a little low temp solder on your iron
3) transfer the blob on your iron onto all the pins on one side of the IC
5) repeat for all other sides of the IC
make sure you get every pin
6) move your iron gently around all pins until the chip lets go. ?Don¡¯t force it. ?If it¡¯s on the side of the board just scrape it off onto your mat. ?If not, use tweezers to pick it up and move it off. ?The solder will stay melted for a few seconds.
7) clean the pad carefully with solder wick and flux
8) clean off all old flux with 99% IPA
Be careful to not damage pads.
Also make sure the low melt solder didn¡¯t get splashed onto the board or surrounding components.
?
i find this far less risky that a hot plate pre warmer and hot air gun.
?
--
73
Jeff Moore
W1NC
 

I find a hot air gun with lots of flux a prior ideal for de-soldering. For soldering too,but you need solder paste. Best to practice with one of those training kits from AliExpress first.

1 - 9 of 9