Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
QMX+ V2 - I've killed it
|
After successfully building and tuning my QMX+ (V2 PCB), after a day's use I've done it again and blown something up! I managed to accidentally 'wiggle' the 12V connector with the radio on while leaning over it and trying to plug in a PC monitor (to something else!) - in other words, I was being careless.
?
The rig was connected to a power speaker, so a few loud crackles from that, then it was dead.?Now when I reconnect to a current limited PSU, pressing the on switch on the top encoder causes the supply voltage to short to ground somewhere. So something's blown that shorts the 12V line - I think.
?
With the 12V supply shorting as soon as I press the 'on' button, nothing powers up at all - it's completely dead.
?
No signs of damage or burnt components found, tracks all look ok and no familiar burnt plastic smell. I've checked all the diodes and components around the integral SMPSUs - all seem ok.?
?
If anyone has any pointers on what to look out for, any info. on what to check next would be appreciated!
?
Many thanks,
?
Steve MW0KST
?
? |
|
At the risk of being suuuuuper obvious, I'd check that 12V connector that wiggled when you leaned over! If it were me, my first thought would be to wonder if one if the solder joints on the 12V connector were all solid to begin with. I don't think they're supposed to wiggle all that much! Michael WD0OM On Thu, Sep 19, 2024, at 19:10, Steve MW0KST wrote:
|
开云体育Hi Steve;Can you slide the board out of the case and then check to see if the short is still present? 73, Steve, N3SB? -------- Original message -------- From: Steve MW0KST <stevedavies77@...> Date: 9/19/24 8:10 PM (GMT-05:00) Subject: [QRPLabs] QMX+ V2 - I've killed it After successfully building and tuning my QMX+ (V2 PCB), after a day's use I've done it again and blown something up! I managed to accidentally 'wiggle' the 12V connector with the radio on while leaning over it and trying to plug in a PC monitor (to something else!) - in other words, I was being careless.
?
The rig was connected to a power speaker, so a few loud crackles from that, then it was dead.?Now when I reconnect to a current limited PSU, pressing the on switch on the top encoder causes the supply voltage to short to ground somewhere. So something's blown that shorts the 12V line - I think.
?
With the 12V supply shorting as soon as I press the 'on' button, nothing powers up at all - it's completely dead.
?
No signs of damage or burnt components found, tracks all look ok and no familiar burnt plastic smell. I've checked all the diodes and components around the integral SMPSUs - all seem ok.?
?
If anyone has any pointers on what to look out for, any info. on what to check next would be appreciated!
?
Many thanks,
?
Steve MW0KST
?
?
|
|
I did exactly the same thing. Running it for a few hours. I tilted the rig up to get a better view of the screen and heard a crackle.? Then it shut down and will not power up.
Checked the dc plug? for shorts and solid connection. No issues found.
My testing so far is Pressing th Power button shows 22 ma current draw, But the rig does not start up.
?
Hope you and I get get it working,? Me too.
?Lee
AA7HC
? |
|
Another possibly "obvious" question -- the screen doesn't come up, but does the terminal app? There's an awful lot of diagnostic available IF you can get to the terminal app. On Thu, Sep 19, 2024, at 20:02, harleyleenuke wrote:
|
|
Since you both heard a 'crackle':
I think about the only thing that could crackle with movement of the QCX in receive mode is a mostly direct short between the input 12V and ground.? This could come from the board flexing a bit and something on the board contacting the chassis (or similar mechanical movement).? I would first visually inspect very carefully with board in chassis to see what could possibly touch where it is not supposed to. Current flow for such a short would go mainly through Q105 and Q103, possibly frying one or both of them.? It would be reasonably easy to check the +12V rail at one of the power jumper blocks, or better at the input and output of Q105 when you try to power on.? If you see a good 12V at all of those spots, the problem is elsewhere - but from what you are saying, it's likely in that area of the circuit. |
|
Thanks again all. I can’t connect to my PC at all. No signs of shorts to the case internally but I can’t rule that out as the cause.?
I’ll order replacements for Q3 and Q5 and see how that goes. Wonder if it’s worth ordering some other components that might also need replacement. Hm… which ones?
I had to replace the 5V board on another QMX recently due to a scratchy 12V supply - the bad connection there was on the external PSU, and the simple board swap fixed it. Not quite so easy with the SMPSUs being integrated into the main board with V2. In that case I was able to connect to the terminal and see that the 5V board was dead.?
Steve MW0KST |
|
It does sound like maybe some mechanical flexing in the PCB could of shorted something out to the chassis.
I would suggest pulling the board out of the case and looking at it from the side to see if you have component leads that are too long and the board flex put them in contact with the chassis, creating a temporary short that killed something.
?
Since we install our own inductors, transformers and transistors the lengths of the leads are left up to us to cut flush enough to not get near the chassis. It is also possible that the flexing broke free a component, solder joint or trace.
?
A make-or-break electrical transient seems unlikely and if that did happen it might harm some of the start-up circuitry, switching supply or regulators.
?
I have an Acer laptop that has that very problem with their poorly designed power connector on the back, I now need to heat gun off the surface mount connector and put a new one on the laptop (PITA).
--
Tisha, AA4HA |
|
Thanks Dan - and all. All good pointers for investigation - so much appreciated.
?
I'm using the V1 board schematics, but as far as I can see the SMPSU board component numbering has been maintained in the V2 PCB, so l'll follow it as far as I can to find where the short is. Hopefully it'll be something that's relatively easily replaceable. I'm 50/50 on the cause being a short to the case under the PCB or on a rapidly fluctuating external supply voltage.
?
I'll posting my findings when I, um, find them!
?
Thanks again,
?
Steve MW0KST |
|
I finally found the fault - it was D108. I removed it, applied power (7V, 200mA limited) and the display came on, CPU booted etc. Quickly powered off and ordered a replacement. Shame I was being lazy and transferred all the toroids into my replacement QMX+ . On the upside, it looks like I'll end up with 2 working QMX+ rigs! I think I prefer the idea of the separate SMPSU modules - it would have been easy and fast to swap them out as a quick initial check...
?
Hope this is of help to someone in future with a similar problem...
?
73, Steve MW0KST
?
?
? |
|
Robert AD6XJ
Thank you Steve (& Lee), I had the same thing happen with my QMX+. It really hurts having it go kaputz with less than two hours "on" time and only checking the power out on each band into a dummy load once, never on the air.
I'll look at the diode you suggested and see if I can fix the thing I just finished building ;-)
Wish me luck! |
|
Apologies all - it was D109 not D108.
?
Anyway, good luck Robert - if it's just D109 it should be a easy repair. I've ordered a bunch of them for um, future use.
?
I was actually convinced my MCU was dead - so a nice surprise when the rig came to life! I haven't tested it properly yet - that will be after I reinstate all the LPFs and BPFs.
?
I forgot to pass my thanks to others - there are a lot of posts re D109 but the one that prompted me to check was from Paul G0TXL - so thanks Paul and everyone here for that matter - it's great to have the group as a resource, and for me essential when fault finding!
?
73, Steve MW0KST |
|
Here's G0TXL's post about D109 failing:?
Here's discussion of how zener diodes fail:?
?
I would assume the cause of D109 failing is a step change to a higher input voltage.
The STM32F did not adjust the PWM_3V3 signal to the SMPS at Q107 soon enough,
resulting in a too high 3.3v rail.
?
The QMX+ assembly manual says D109 is a 3.6V 500mW SOD-123?
One such part is this from Mouser?
?
From 7 lines down on page 3 of the datasheet, we find that the MMSZ5227
zener voltage at 20 mA and a temp of 30C is between 3.42 and 3.78 Volts.
Fig 9 on page 6 of the datasheet shows little variance in zener voltage as
the current goes up, but only shows a max of 20ma, or 0.020*3.6 = 72 mW,
though device dissipation is rated at 500 mW.
I don't see anything in the datasheet showing what will happen at above 20 mA.
?
I find that a bit concerning.??
We are lucky if that 3.3v zener at D109 typically fails as a short instead as an open.
It is protecting the 100 pin STM32F which few of us would be able to replace.
When it fails, it may have several amps going through it at well over 4 Volts.
?
What firmware version do you have?
Hans reports that the latest firmware should significantly reduce SMPS failures:
?
If the latest firmware is having failures, we may want to replace the 5v switcher with
a TO220 linear regulator and power the 78M33 at IC101 from 5v for the 3.3v rail
until a better solution is found.? This increases supply current to a couple hundred mA
on receive, but that should not be an issue unless operating portable on?a small battery.??
?
Jerry, KE7ER
?
?
On Sun, Oct 20, 2024 at 04:30 AM, Steve MW0KST wrote:
there are a lot of posts re D109 but the one that prompted me to check was from Paul G0TXL |
|
Hi all My QMX and QMX+ here go through a huge number of experiments and a lot of abuse, and I don't see SMPS failures.? On Friday Muhsin TA1MHS and another friend and I did some portable operating with my 80-20m QMX, I used an 8x Alkaline AA battery pack, and a short 2.6m telescopic whip with tunable sliding loading coil. There was a coax fault (open) and once we realized, we had to plug the whip directly I to the QMX. There was a faulty intermittent battery connection too which kept rebooting the QMX as it wobbled about, unless you could keep it still... which was hard when the whip was attached directly to the QMX BNC. We also tried a power bank having 12V output which didn't work at first but seemed to later sometimes work somehow.. Despite all these limitations and abuse (intermittent power connection, open load, high SWR) there wasn't any failure of any sort, and I made a few CW QSOs.? In my?opinion the QMX/QMX+ probably aren't invincible but neither are they *that* delicate either. 73 Hans G0UPL |
|
Robert AD6XJ
My D109 is conducting both directions in situ. D106 & D107 conduct one way and briefly the opposite due, presumably, to C107 & C106 respectively (they are cross enumerated) charging.
Zeners still make my head scratch-- they act like "normal" diodes when forward biased and when reversed biased until the rated voltage then they also conduct reverse? ?At the knee?
Any way I only need a part number to get this QMX+ going, not a complete understanding of how the magic midget squirrels inside stack and load the electrons from one end to the other.
That is a VERY tight spot between its two bigger brothers! Smiling,
Robert AD6XJ
? |
|
Robert,
What firmware were you operating with when it failed? Was it the latest firmware?
?
From my previous post in this thread:
The QMX+ assembly manual says D109 is a 3.6V 500mW SOD-123?
One such part is this from Mouser?
?
?
?
So you want a 3.6 Volt zener good for 500 mW dissipation in an SOD-123 package.
I believe that Mouser part is a reasonable choice.
?
Use two soldering irons when extracting that part.
?
Jerry, KE7ER
?
?
?
On Sun, Oct 20, 2024 at 09:02 AM, Robert AD6XJ wrote:
|
Daniel Conklin
Tisha Hayes, AA4HA
to navigate to use esc to dismiss