|
Long story short, I managed to damage my fresh built QMX+ by trusting a USB-C PD power supply too much... The symptoms are that I'm still able to power it up and the display is normal before SMPS kicks in. When the controller starts to use SMPS, the LCD goes all blocks so I guess the 5V rail is damaged. Visual inspection shows 74ACT08 is damaged as well. The terminal application is still working (so 3.3V is fine?), but if I try to access the hardware diagnosis function, QMX+ crashes and the USB connection is lost. The steady power consumption after SMPS bootstrap is about 22W so something is probably shorted.
?
I have an unbuilt QMX+ board that I can take components from. Really appreciate anyone could help me narrow down the components to check/repair. I was trying to use QMX+ SSB to work some WPX SSB contest stations... Really wish I could still make it... ?Many thanks!
?
73 de N6HAN Zhenxing Han
|
On Sat, Mar 29, 2025 at 08:19 AM, Zhenxing Han N6HAN wrote:
Long story short, I managed to damage my fresh built QMX+ by trusting a USB-C PD power supply too much...
If you use an external trigger module with a PD power bank, it is easy to accidentally let it switch to the 20V mode and break your QMX.? I know this from my sad experience.? But now I only use a fixed PD trigger cable, and it always works fine.
The symptoms are that I'm still able to power it up and the display is normal before SMPS kicks in. When the controller starts to use SMPS, the LCD goes all blocks so I guess the 5V rail is damaged.
I'm not sure your analysis is correct.? How are you determining when the SMPS kicks in?? The 5V supply is only an SMPS supply, and it is what powers the display LCD and backlight.? So if the LCD ever comes on, it is because the 5V SMPS is already going.? But it could still be damaged in some way
Visual inspection shows 74ACT08 is damaged as well.
So this could be from an over-voltage on your 5V rail.? Did the fault happen while transmitting, or receiving?? If transmitting, something that blows your finals could also blow this device.
The terminal application is still working (so 3.3V is fine?), but if I try to access the hardware diagnosis function, QMX+ crashes and the USB connection is lost.
Hmmm... this should hold a clue.? Ludwig, have you seen this?
The steady power consumption after SMPS bootstrap is about 22W so something is probably shorted.
Yes, that is way too high for any mode of QMX+ operation.? So something definitely is shorted.? Your receive mode power consumption should be about 1W.? Do you have a way to limit the current into QMX?? For further debugging, until you find the fault and get a successful power-on at normal current draw, you should limit the current to 250mA to avoid damaging additional circuits.? Also it is best to reduce the voltage to about 7V.? Since things are kinda-sorta working a bit, your 3V and 5V protection diodes seem to have saved you from extensive damage, and are not likely damaged themselves.
I have an unbuilt QMX+ board that I can take components from. Really appreciate anyone could help me narrow down the components to check/repair. I was trying to use QMX+ SSB to work some WPX SSB contest stations... Really wish I could still make it... ?Many thanks!
It is really hard to say what all is damaged.? It looks like you didn't hit it with 20V like I did to mine,? since yours isn't totally broken.? I blew Q108, D109, IC403 and the TVS input protection diode, D110.? But it seems your D108, D109 and D100 are probably fine, since all voltages still power on.? Maybe you are lucky and only IC503 is blown - and it could be shorting the 5V supply down after the receiver/transmitter are configured.?
?
So my best shot-in-the-dark guess at an action plan is to remove IC503, and see if the short is gone, preferably by measuring the resistance between Vcc and gnd, which should be a few k-ohms.? Then powering up with a current-limited supply to see if it comes up, and things start working again in receive mode.? Then replace IC503 with a new one, and power-up with a current-limited supply, check the diagnostics, except LPF sweep and SWR sweep (which will return poor results until you increase the current limit).? Then if all looks good, increase the current limit and try an SWR sweep, LPF sweep and diagnostic mode Tx. ? You may find that your finals are also smoked and need replaced.
?
Good luck.
Stan KC7XE
|
On Sat, Mar 29, 2025 at 04:19 PM, Zhenxing Han N6HAN wrote:
The steady power consumption after SMPS bootstrap is about 22W so something is probably shorted.
This would mean 1.8 A @12 V! To much and dangerous for the QMX.
?
the display is normal before SMPS kicks in
What do you mean by "normal"? Before SMPS is working the display should be blank and dark.
When the controller starts to use SMPS, the LCD goes all blocks so I guess the 5V rail is damaged.
How do you know the controller is working and starts the SMPS in this state?
When the controller starts to use SMPS, the LCD goes all blocks so I guess the 5V rail is damaged.
Strange. In some failure situation going to Diagnostics may reduce input current. Do you know your supply voltage every time?
Visual inspection shows 74ACT08 is damaged as well.
When seeing a damage something very bad was happening.
?
trusting a USB-C PD power supply too much...?
Could you tell a bit more about? Is it important it was USB-C? How did you feed your QMX by the USB-C PD power supply and how did you feed it later?
?
73 Ludwig
?
|
On Sat, Mar 29, 2025 at 06:12 PM, @Ludwig_DH8WN wrote:
?
When the controller starts to use SMPS, the LCD goes all blocks so I guess the 5V rail is damaged.
Strange. In some failure situation going to Diagnostics may reduce input current. Do you know your supply voltage every time?
?
Sorry, failure. It should be:
but if I try to access the hardware diagnosis function, QMX+ crashes and the USB connection is lost.
Strange. In some failure situation going to Diagnostics may reduce input current. Do you know your supply voltage every time?
?
73 Ludwig
|
Hi Stan, thank you so much for your analysis!
?
I used a fixed 12V PD trigger, so it's probably not the 20V issue. I'll take a closer look later.
?
I'm not sure your analysis is correct. ?How are you determining when the SMPS kicks in? ?The 5V supply is only an SMPS supply, and it is what powers the display LCD and backlight. ?So if the LCD ever comes on, it is because the 5V SMPS is already going. ?But it could still be damaged in some way
You're absolutely right! I realized this as well. It looks like when the controller is still on the linear regulator, the 5V SMPS and LCD control works fine (at least partially).
?
So my best shot-in-the-dark guess at an action plan is to remove IC503, and see if the short is gone, preferably by measuring the resistance between Vcc and gnd, which should be a few k-ohms. ?Then powering up with a current-limited supply to see if it comes up, and things start working again in receive mode.
I'm removing IC503 now. Without power, the resistance between Vdd/GND is 2.6k so I guess the short is either on the Vcc rail, or only happens when the 3.3V SMPS is running.
?
Thanks again!
?
73 de N6HAN Zhenxing Han
|
Hi Ludwig,
?
Thanks for your reply!
?
What do you mean by "normal"? Before SMPS is working the display should be blank and dark.
Sorry my bad. I thought 5V had a linear regulator as well. What I observed was that after long-pressing the power button, the LCD would turn on normally and show some text (this is probably when the controller is still running on the 3.3V linear regulator). After sth like "initializing audio", all the characters would become solid squares, but the backlight was still on (this is probably when the controller has shut down the 3.3V linear regulator).
?
How do you know the controller is working and starts the SMPS in this state?
Because the USB connection and terminal app is working.
?
Could you tell a bit more about? Is it important it was USB-C? How did you feed your QMX by the USB-C PD power supply and how did you feed it later?
Sure. It's probably not important it was USB-C. I was trying to power my QMX+ with a fixed 12V PD trigger connected to an Anker 36W USB-C power brick. I was able to turn it on partially - the LCD was up for maybe 1 sec, then it suddenly shut down. I tried 2 or 3 times then I realized this particular power brick might have some issues. (BTW I've been using the 67W one with 9V and 12V PD trigger all the time without any issues). I just double checked, the 36W one does generate the correct voltage, so I think QMX+ never saw 20V from it. But I guess when QMX+ was shutting down the 3.3V linear regulator, it probably triggered some protection on the 36W power brick which would cause voltage spikes somewhere and that fried IC503 somehow.
?
73 de N6HAN
|
After removing IC503, I'm able to power it on normally. Terminal app shows both SMPS modules are working fine. I'll order a few ACT08s and test the TX/RX when they arrive. And I'm going to build the low-voltage drop linear regulator for QMX.
?
Really wish I could work WPX SSB and share with the group my experiences, but not in the mood anymore...
?
Thanks Stan and Ludwig again for your help!
?
?
|
On Sat, Mar 29, 2025 at 11:35 AM, Zhenxing Han N6HAN wrote:
Anker 36W USB-C power brick
Han, can you share with us the exact model of Anker 36W power bank you have?? I don't see one labeled that way on their website.? Several of Anker's power banks do not support 12V PD (since it is not part of the current PD spec - I think it was in an older version of the spec). But several do support 12V PD, even the newer ones.? When one that doesn't support 12V PD sees a 12V PD trigger, they may show 11.x volts derived from another one of the protocols - but that voltage could change as current changes, since it is not in PD mode.?
?
I have one Anker powerbank that supports the 12V PD trigger, and one that does not.? They both support the 9V, 15V and 20V PD triggers, which are in the PD spec.? If you look at the detailed specs for each of their power banks on the website, it shows which trigger voltages they support in USB-C PD mode, and which voltage ranges in other modes.? If it doesn't explicitly say 12V in the PD mode spec, it is not supported.
|
There is a current limit for the VCC rail. Duty max is doing this. So I don't believe the high current was flowing from VCC via IC503 to GND. Maybe something is wrong around the BS170. I see two possibility:
- The broken IC503 switched one pair of BS170 on the whole time. The high current was flowing via these BS170.
- One or more BS170 is/ are broken. Or you have a solder bridge from drain to gate. The PA voltage is switched on the most of the time (also in RX mode). So a high voltage around 12 V was shown to IC503, damaged IC503 and after this a high current was flowing via the PA failure into IC503, let explode the poor IC and made it to a simple piece of "wire" to GND. After removing IC503 this path to GND would be open now.
I would check the region around the BS170. You may also check resistance D-G, G-S and D-S.
?
73 Ludwig
|
On Sat, Mar 29, 2025 at 11:45 AM, Zhenxing Han N6HAN wrote:
After removing IC503, I'm able to power it on normally. Terminal app shows both SMPS modules are working fine.
Excellent!? So glad that it is working without other damage.
And I understand completely the drag it is when you wanted to have fun in WPX.? I have been stymied for other (non-QMX) reasons.
?
FYI, I have used exclusively PD power banks and PD trigger cables without any issues - but I did have to make sure the power bank unit properly supports 12V for the 12V rig (I have one QMX wired for 9V, and use 9V PD for it - but one of my power banks won't do the 12V, so I use it only for the 9V QMX).
?
Best luck to you going forward, and when the new parts come, let us know how it goes...
Stan KC7XE
|
Hi Stan, sure. It's a charger (sorry I forgot the word "charger" in the original post), not a power bank. The specs on the bottom are:
?
SWITCHING ADAPTER
PN: A2935 Model ASGaN83w-P40W20
Input 100-240V~50/60Hz 1.0A
Output: 5.0V=3.0A 15.0W
9.0V=3.0A 27.0W / 12.0V=3.0A 36.0W /
15.0V=2.66A 39.9W / 20.V=2.0A 40.0W
?
Seems to be a 40W charger. I think it comes with my 3-in-1 MagGo charger (). I checked it a few times. I have no issue getting 9V or 12V from it with PD triggers.
?
73 N6HAN
|
On Sat, Mar 29, 2025 at 02:28 PM, @Ludwig_DH8WN wrote:
There is a current limit for the VCC rail. Duty max is doing this. So I don't believe the high current was flowing from VCC via IC503 to GND. Maybe something is wrong around the BS170
Good catch, Ludwig. I guess I was writing when you posted. You are always more careful than I am in diagnosing things.
Han, you should order some BS170, too, if you don't already have some, and Ludwig is spot-on in recommending you look around the BS170s for additional issues.
|
Hi Ludwig and Stan,
?
I'd definitely check BS170 as well. Once removed from the PCB, IC503's Vcc to gnd seems to be open, so if it was causing short, it would have to be done via BS170 I guess?
?
I'd provide some more context. I finished building this QMX+ a few days ago. It had good TX/RX. Power output on 20/10 was a bit more than 4W. Compared to my previous QMX HB build, its MDS is 10dB better and its SSB tx from mic doesn't have the constant 2kHz tone in the background like my QMX HB does. I was very satisfied with its performance. I believe the BS170s had been working fine before. This morning I was setting up the shack for WPX and I wanted to power it with a dedicated USB-C charger with 12V PD trigger. I tried a few times to turn it on from that 40W Anker charger without success (LCD went off after showing "Initializing audio..."). A few minutes later, when powered from the known good charger/battery, it started to have the symptoms I described in the original post. So TX was never attempted with that suspicious 40W charger. I'll check BS170s later today. I do have lots of new BS170s - I ordered them last year with PCM1804 and 3253. Maybe I should order SS14, D108, D109 as well just in case..
?
PS I didn't install the BFP/LPFs for the 160/80/60 bands. Not sure if that's gonna cause problem.
?
73 N6HAN
|
Oh Stan,? you embarrass me.
|
Not installing the BPFs/LPFs shouldn't cause a problem.? You should disable TX for those bands in the band config menu, otherwise it is risky.? Accidental TX when the band is selected or running any of the Hardware test tools or SSB Calibration on those bands may cause failures, since they won't be terminated correctly.? So I hope those bands weren't included in your SSB Calibration; in its beta condition who knows if it is honoring that TX disable setting.
?
I purchased a couple of very inexpensive (Ali express) "USB&DC Digital Tester"s (mine are Atorch brand) that I always connect inline with a new power source before and while initially using it with my QMX.? It shows voltage and current, and also the PD and QC protocol triggering status when on a USB port.? It has both USB and 5.5/2.1 digital in/out ports.? So I can be confident of the voltage before I plug it in, and watch the current while it starts up.? I highly recommend something like this.? It sounds like the combination of your trigger cable and that one charger were not doing what was expected.
?
Stan KC7XE
?
|
Accidental TX when the band is selected or running any of the Hardware test tools or SSB Calibration on those bands may cause failures, since they won't be terminated correctly. ?So I hope those bands weren't included in your SSB Calibration; in its beta condition who knows if it is honoring that TX disable setting.
Thanks. I didn't consider the termination. I did disable TX on those bands. But I'm going to install them to avoid any risk now.
?
I highly recommend something like this. ?It sounds like the combination of your trigger cable and that one charger were not doing what was expected.
I have a similar one (identical functionality, but different shape). That charger indeed generates the voltage what the trigger cable asks to, so I really don't know what happened.
?
73 N6HAN
|
The broken IC503 switched one pair of BS170 on the whole time. The high current was flowing via these BS170.
Ludwig, you were right. I measured the resistance between the terminals. All 4 BS170s were damaged. I guess the broken IC503's outputs were floating which in turn caused high current flowing between the drain and the source? But I (think) didn't attempt any TX so the drain voltage should be close to 0 all the time? No? Anyway, all 4 BS170s replaced.
The PA voltage is switched on the most of the time (also in RX mode).
I don't quite get this part. Transmitter Voltage is at 0.5V in RX mode. Or you mean sth else?
?
73 N6HAN
|
When you enter the diagnostics screen, the Tx voltage is turned off, so it can be tested and we can see it turn off and on.
As soon as you exit the diagnostics screen, it turns back on.
If you read Hans' description, they found that certain things work better in receive if they leave the tx voltage on.? But in receive, the TX signal into IC503 is off, so the BS170s never get turned on, never conduct current.? At least if IC503 isn't broken...? So either:
a) IC503 was broken first from an over-voltage on the 5V, with floating or shorted-high outputs that caused the BS170s to fail, or
b) as I think Ludwig suggests, the BS170s had suffered some damage, and finally broke, sending a 12V over-voltage into IC503 outputs, frying it.
And I think Ludwig's version is more likely, since I have no idea how you could get a severe over-voltage on 5V without also damaging the 5V smps protection diode - it would clamp the over-voltage to ~5.6V, except for maybe a fast-and-furious spike that was faster than the clamping speed of the diode.
Or maybe it was something else that we haven't thought of...
?
Stan KC7XE.
?
|
Hi Stan and Ludwig, thanks for the help! I think now I know what happened.
?
b) as I think Ludwig suggests, the BS170s had suffered some damage, and finally broke, sending a 12V over-voltage into IC503 outputs, frying it.
I believe this was indeed what happened:
?
1. I did attempt SSB calibration from the on-device menu a few times although I stopped them promptly. Even though, BS170s were somewhat damaged due to the fact that I didn't install LPF for 160M. The SSB calibration initiated from the on-device menu always starts from the lowest band and it doesn't honor the TX disable setting (I validated it on my QMX HB).
?
2. The 40W Anker charger is innocent. It failed to fully power on QMX+ probably because at that time IC503 had been damaged already which in turn caused a short in the 12V rail via BS170s. The 40W charger was just protecting itself from over-current.
?
3. Switching to other power sources made things worse because it allowed ever higher current to pass through... I guess the reason why running the diagnosis terminal app would crash the controller is because it would disable the PA voltage which would cause an abrupt stop of the high current in the 12V rail which would in turn cause a high induced voltage (probably due to L502?) to trigger the protection in 3.3V SMPS somehow?
?
This morning I tested my QMX HB with that 40W Anker and it was completely normal. After that, guess what, I made another stupid mistake...
But in receive, the TX signal into IC503 is off, so the BS170s never get turned on, never conduct current. ?At least if IC503 isn't broken
Then I tried to power on my QMX+ with that 40W Anker to see if it would work. It did power on fully, but this time I connected the USB-C meter in-line and I immediately noticed the input power was 14W in RX. What happened? ?Well I replaced all the BS170s yesterday, but IC503 was not there so the gates were floating. After the initialization, the firmware would turn on the PA voltage even for RX which fried half of the BS170s in seconds...
?
Sigh, I've learned so many lessons from my mistakes this weekend... One suggestion to Hans, maybe a self recovery 2A fuse could be added to the 12V rail?
?
Thanks again to Stan and Ludwig!
?
73 N6HAN
?
|
It's really nice to discover what really happened.? Great job in patience and persistence to figure it out, and not just blame the kit - it's almost always something we do that causes the failures.
?
In this case, a fuse certainly would have helped.?
And one change Hans should indeed consider is to disable calibration if TX is disabled on that band.
?
Stan KC7XE
|
