Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

开云体育

Forgot to say that I’m now using a 12v supply?

?

On 14 Nov 2024, at 08:31, James Silvester via groups.io <jamessiilvester@...> wrote:

?
?Good Morning Stan,
I have checked R101 again from the gate of Q101 to the source Of Q102 to prove soldered joints at R101 and readings were correct @ 100k. I then shorted the gate of Q102. I now have a current draw of 0.6amps but only 0.8v at Vdd which both obviously drop to zero when I disconnect PWR_Hold as it needs the 3.3v.

Thanks,

James. MJ7HEE
?

On 14 Nov 2024, at 07:22, Stan Dye via groups.io <standye@...> wrote:

?
On Wed, Nov 13, 2024 at 09:46 PM, Jerry Gaffke wrote:

I don't see an R115 anywhere in the QMX+ schematics
Oops.? I guess we all make mistakes.? James, please disregard all my references to R115.? I was looking at the QMX schematics, which do have an R115 (100K to gnd from LIN_REG_EN).? I presumed this was in the QMX+ also, since most else is the same.? But you are correct, QMX+ does not have one - so the R101 should pull up Q101 gate to 5V until the processor is awake and turns it off.

?

Regarding the positive V on the processor pins when processor VDD is near 0: this design relies on that, otherwise LIN_REG_EN couldn't get pulled up while the processor is off, in order to switch on the power and start the processor.?

?

It seems to me that either R101 or Q101 is broken (or has an installation issue) in this case.? That's why I asked for measurements on Q101.??? I guess I should have asked James to also touch up the solder on Q101's gate, which if floating would also cause the problem.? I note that he measured only 0.74V across R101 last time (but not the previous time) - which should mean the gate of Q101 is on - but it is not working.

?

I don't really think there is a way to accidentally load bad firmware into the device - the files are encrypted and would fail decryption.? But I could believe that a bad file could cause the bootloader to continuously try to decode it, rather than go into boot mode.? And with nearly 0V on the processor, it won't be asserting any of its pin drive levels.? What is harder for me to grasp is why a failure like this could happen in conjunction with loading code.? Maybe it was just coincidence, and a part failed, or a cold-solder joint separated due to some movement.

?

And yes, shorting Q102's gate to gnd should turn on IC101 and provide the 3V VDD rail, allowing the firmware to be installed again.? It won't electrically cause any problems to leave it on permanently, because of the presence of D103 - it will just waste some mA.? If LIN_REG_EN is getting pulled low by something else (even a damaged processor pin), everything should still work.? I guess another failure mechanism could have been a short across R101, causing damage from direct Vin to the processor pin.?? This is definitely still a mystery.

Stan KC7XE

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133450

开云体育

?Good Morning Stan,

I have checked R101 again from the gate of Q101 to the source Of Q102 to prove soldered joints at R101 and readings were correct @ 100k. I then shorted the gate of Q102. I now have a current draw of 0.6amps but only 0.8v at Vdd which both obviously drop to zero when I disconnect PWR_Hold as it needs the 3.3v.

Thanks,

James. MJ7HEE

?

Show quoted text

On 14 Nov 2024, at 07:22, Stan Dye via groups.io <standye@...> wrote:

?
On Wed, Nov 13, 2024 at 09:46 PM, Jerry Gaffke wrote:

I don't see an R115 anywhere in the QMX+ schematics
Oops.? I guess we all make mistakes.? James, please disregard all my references to R115.? I was looking at the QMX schematics, which do have an R115 (100K to gnd from LIN_REG_EN).? I presumed this was in the QMX+ also, since most else is the same.? But you are correct, QMX+ does not have one - so the R101 should pull up Q101 gate to 5V until the processor is awake and turns it off.

?

Regarding the positive V on the processor pins when processor VDD is near 0: this design relies on that, otherwise LIN_REG_EN couldn't get pulled up while the processor is off, in order to switch on the power and start the processor.?

?

It seems to me that either R101 or Q101 is broken (or has an installation issue) in this case.? That's why I asked for measurements on Q101.??? I guess I should have asked James to also touch up the solder on Q101's gate, which if floating would also cause the problem.? I note that he measured only 0.74V across R101 last time (but not the previous time) - which should mean the gate of Q101 is on - but it is not working.

?

I don't really think there is a way to accidentally load bad firmware into the device - the files are encrypted and would fail decryption.? But I could believe that a bad file could cause the bootloader to continuously try to decode it, rather than go into boot mode.? And with nearly 0V on the processor, it won't be asserting any of its pin drive levels.? What is harder for me to grasp is why a failure like this could happen in conjunction with loading code.? Maybe it was just coincidence, and a part failed, or a cold-solder joint separated due to some movement.

?

And yes, shorting Q102's gate to gnd should turn on IC101 and provide the 3V VDD rail, allowing the firmware to be installed again.? It won't electrically cause any problems to leave it on permanently, because of the presence of D103 - it will just waste some mA.? If LIN_REG_EN is getting pulled low by something else (even a damaged processor pin), everything should still work.? I guess another failure mechanism could have been a short across R101, causing damage from direct Vin to the processor pin.?? This is definitely still a mystery.

Stan KC7XE

Re: QMX Mid Band Rx issue

#133449

Thank you for the kind words.? I'm glad I could use my newly-acquired knowledge to help someone.

Stan KC7XE

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133448

It'll stay in DFU as long as the jumper is connected as Paul said.?

?

So power off, connect jumper, power on, flash firmware from PC downloads to QMX drive. Then once it's copied wait a bit before removing the power cable. Remove the jumper, then it should power on normally once you plug the power back in.

?

--

Daniel ZL2DTL/ZL4YOTA?

?

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133447

On Wed, Nov 13, 2024 at 09:46 PM, Jerry Gaffke wrote:

I don't see an R115 anywhere in the QMX+ schematics

Oops.? I guess we all make mistakes.? James, please disregard all my references to R115.? I was looking at the QMX schematics, which do have an R115 (100K to gnd from LIN_REG_EN).? I presumed this was in the QMX+ also, since most else is the same.? But you are correct, QMX+ does not have one - so the R101 should pull up Q101 gate to 5V until the processor is awake and turns it off.

?

Regarding the positive V on the processor pins when processor VDD is near 0: this design relies on that, otherwise LIN_REG_EN couldn't get pulled up while the processor is off, in order to switch on the power and start the processor.?

?

It seems to me that either R101 or Q101 is broken (or has an installation issue) in this case.? That's why I asked for measurements on Q101.??? I guess I should have asked James to also touch up the solder on Q101's gate, which if floating would also cause the problem.? I note that he measured only 0.74V across R101 last time (but not the previous time) - which should mean the gate of Q101 is on - but it is not working.

?

I don't really think there is a way to accidentally load bad firmware into the device - the files are encrypted and would fail decryption.? But I could believe that a bad file could cause the bootloader to continuously try to decode it, rather than go into boot mode.? And with nearly 0V on the processor, it won't be asserting any of its pin drive levels.? What is harder for me to grasp is why a failure like this could happen in conjunction with loading code.? Maybe it was just coincidence, and a part failed, or a cold-solder joint separated due to some movement.

?

And yes, shorting Q102's gate to gnd should turn on IC101 and provide the 3V VDD rail, allowing the firmware to be installed again.? It won't electrically cause any problems to leave it on permanently, because of the presence of D103 - it will just waste some mA.? If LIN_REG_EN is getting pulled low by something else (even a damaged processor pin), everything should still work.? I guess another failure mechanism could have been a short across R101, causing damage from direct Vin to the processor pin.?? This is definitely still a mystery.

Stan KC7XE

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133446

Hello Jerry

?

I assume this means that it would be impossible to accidentally

load any firmware into a QMX+ that does not start up the 3.3v SMPS.

Yes that's correct. If you tried to load QDX firmware in, then it would not work. The QMX bootloader would refuse to run the QDX firmware you tried to load in, and it would remain in bootloader mode, appearing to the PC as a USB Flash drive, until you loaded in proper QMX firmware.?

73 Hans G0UPL

Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

#133445

I am these BMS boards for couple of years now. As Dave pointed out they are excellent balancer/protection boards and he is abosultely right about not having the li-ion charging graph characteristic. For that I use one of these 12.6 V li-ion charge adapters:

?

?

The other option is to use this IC if you wanna build yours:

?

Though in my personal experience the 12.6 V charge adapter works pretty good.

Dear Barb,

Merhaba!

?

I have been successfully using your early design and this BMS since 2021 and ended up having multiple own derivative variants, thanks for you sharing your KiCad files!

?

I have multiple of these for myself and family members as a power bank. Never ever had a problem, that is why I have chosen this for Qmx+ mod as well. As my normal field/car charger I am using the one that came with one of my Baofeng rigs, which appears to be the equivalent of what you just shared from Amazon.

?

But at home, I am charging these battery packs from my 13.8v bench power supply which brings it to 12.4V as well as the Baofeng version.

?

On my next battery pack pcb iteration I am going to increase the track width from .5mm to 1mm by David’s suggestion. There is plenty of room on the pcb anyway. However the QMX+ power rail ?header pin thickness becomes a bottleneck in this case.

?

The other reason I am very picky about the BMS - this one does not introduce noise. Neither when power bank is connected externally, nor as an internal Qmx+ module.

?

Hans did hard work with his own power supply design no minimize the noise, my design goal was at least to maintain it at the current level. And as Americans say “excuse my French, do not pee on his campfire” :).

?

Thanks for sharing the IC. It is a good learning material.

--

73,
Sardar (AC8L, 4K6SA, VA3DUA).

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133444

Thanks,

?

I assume this means that it would be impossible to accidentally

load any firmware into a QMX+ that does not start up the 3.3v SMPS.

?

Jerry

?

On Wed, Nov 13, 2024 at 10:04 PM, Hans Summers wrote:

Show quoted text

Hello Jerry

?

QDX and QMX firmware?are completely different and cannot be interchanged....

?

73 Hans G0UPL

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133443

Hello Jerry

QDX and QMX firmware?are completely different and cannot be interchanged....

73 Hans G0UPL

Show quoted text

On Thu, Nov 14, 2024 at 9:00?AM Jerry Gaffke via <jgaffke=[email protected]> wrote:

Hans,

?

Is it possible to load QDX firmware into the QMX+?

James has a QMX+ for which the 78M33 is not seeing +12V

at power up, I am thinking he may have loaded firmware that

drops LIN_REG_EN without starting the 3.3v SMPS.

?

As a fix, I am thinking he might short Q102's gate to ground,

so the 78M33 is running in parallel with the 3.3v SMPS.

That way he can try loading firmware again.

?

Jerry, KE7ER

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133442

Hans,

?

Is it possible to load QDX firmware into the QMX+?

James has a QMX+ for which the 78M33 is not seeing +12V

at power up, I am thinking he may have loaded firmware that

drops LIN_REG_EN without starting the 3.3v SMPS.

?

As a fix, I am thinking he might short Q102's gate to ground,

so the 78M33 is running in parallel with the 3.3v SMPS.

That way he can try loading firmware again.

?

Jerry, KE7ER

Re: My QMX+ switch on now refuses to hold after trying to add the new firmware file and now not seen as USB on computer

#133441

James:

?

The ESD protection diodes in the STM32F processor do not explain your problem.

It is as Ludwig said in post 133310:

>? I guess there are four possible reason for a low voltage at gate of Q101:

>? 1. some unwanted low resistance connection to GND outside an element (low possibility)

>? 2. broken R101 or not right connected R101

>? 3. broken pin of the controller

>? 4. broken Q101.

?

To which I would add the possibility of wrong firmware in the processor

that pulls LIN_REG_EN low even though the 3.3v SMPS is not working.

?

We're still thinking about what might have happened here, and how best to go forward.

?

Stan:

?

Regarding ESD protection,

?

The question was, can the LIN_REG_EN pin go high enough to reliably turn on Q101 when VDD = 0v.

I would expect the ESD protection diodes to the VDD rail in the STM32F to only allow a fraction of a volt on LIN_REG_EN.

LIN_REG_EN is on the PD bank of STM32F IO pins, the expendable PPS signal is also on the PD bank.

?

I tied a 100k resistor from the PPS signal to the + side of an old 9v battery, the - side of the battery to QMX+ ground.

The QMX+ board had no other power supply connected.

I measured 8.2 volts across the battery, and 5.16 volts from PPS to ground.

Not what I expected!

Reversing the battery I then see -0.418 volts from PPS to ground, which seems about right.

?

Here's the STM32F446 datasheet I am using:

Table 55 on page 115 suggests that all pins of the PD bank should be able to sink or source 5ma.

Many of the IO pins in the other banks show a "Negative Injection" of -0 ma, and a "Positive Injection"

of "NA", probably because they can be used as ADC pins.

Table 10 starting on page 44 indicates that all PD pins are 5 volt tolerant.

?

The STM32F has no 5v power supply connections to the pins.??

So protection of the 5v tolerant pins must be using something like a 5v zener to ground.

Hence my measurement of 5.16 volts on the PPS signal.

?

I have not had time to finish my QMX+ build, though I do have it mostly working now.

I have not tangled with firmware updates at all.

Is it possible to load the wrong firmware?? Perhaps the processor wakes up

and pulls LIN_REG_EN low even though the SMPS is not working?

I though Hans would be storing the new firmware in RAM, and only

writing it to flash if it checks out.? Perhaps Hans has cooked up other firmware

on his website that might be accidentally loaded?

?

I am thinking the best way forward is to have James short the gate of Q102 to ground.

Would allow James to at least try loading the firmware again.

Permanently feeds +12V to the input of the 78M33,

and the 3.3v SMPS will be running in parallel with the 78M33. ?

I would try that mod on my own board and watch the VDD rail with a scope

before recommending it to James.

If that doesn't work he should send it out for repair as it is probably in the processor.

Any thoughts??

?

Stan said:

"The R101 resistor should pull the voltage at the Gate of Q101 up towards 5V (in this case, probably 2.5V because of the dividing action created by R115)."

?

Are we looking at the same schematics?

I don't see an R115 anywhere in the QMX+ schematics, either rev 1 or rev 2.

A text search in both versions of the assembly manual have no hits for R115.

And I see no other resistor tied to LIN_REG_EN.

?

Jerry, KE7ER

?

On Wed, Nov 13, 2024 at 03:36 PM, Stan Dye wrote:

Show quoted text

No, 5V is sufficient for these measurements.

These measurements show that the switching that enables the 3.3V power supply is not functioning correctly.? Until that is fixed, no need for anything else.

The R101 resistor should pull the voltage at the Gate of Q101 up towards 5V (in this case, probably 2.5V because of the dividing action created by R115).? But pulling that gate voltage positive should turn on Q101, which should then turn on Q102, enabling the 3.3V supply.?

But this isn't happening.? You first measured about 5V across R101, now you measure 0.74V - but for some reason that is not turning on Q101.

While Jerry is checking the protection diode issue, I think it would be helpful to do the following:

1) With 5V power and PWR_ON asserted like in your prior measurements, measure the voltages on the Gate, Source and Drain of Q101 with respect to ground.? (Its source terminal should be grounded).

2) Carefully look for a possible short across the R115 resistor, and the source/gate terminals of Q101 - a short there could cause this problem.

3) Carefully touch up the solder joints on both ends of R101.? You will need a fine-point tip on your soldering iron for this.? If R101 has a 'cold' solder joint, it could cause what we are seeing.

Re: QMX crashes on RF filter sweep (and other anomalies)

#133440

After a lot of troubleshooting and some sage advice from Jeff W1NC I was able to track the problem down to a bad PCM1804 chip.? I replaced the chip and I now get audio out!? All the hardware tests are returning nominal results.? So, case closed on this one.? Thanks to everyone for the help!

?

John VA7JBE

Re: Why does QMX* drop PTT state during a short pause in a digi transmission?

#133439

Hello Stan

QDX and?QMX use the same code for digital transmission. So why this happens in QMX not QDX is a mystery, I have it on my list to investigate. This was raised the other day here too.?

73 Hans G0UPL

Show quoted text

On Thu, Nov 14, 2024 at 2:06?AM Stan Dye via <standye=[email protected]> wrote:

Hans,
When transmitting in digital modes, under the recommended CAT control, I see that if there is a short pause in the transmitted tone, that QMX (and QMX+) will drop the PTT state, rather than wait for the CAT command to turn off PTT.? Why does this happen?? Certain modes, like Olivia, have some very short breaks in transmission, which cause the QMX transmit to drop off rather than continue properly.? This requires me to use VOX (yuk) when using that mode.

?

I understand the Rise Threshold / Fall Threshold design - but it seems like it is implemented in QMX as a one-shot, not only disabling the transmit envelope at the end of a set of transmitted symbols, but also autonomously releasing the PTT state.? It seems that it should continue the same rise/fall algorithm for each separated set of symbols until the CAT releases the PTT.?

?

Is there a design reason it is behaving this way?? An auto-release of PTT may be a desirable function in case a CAT command is missed, but I think such a feature should have a programmable hang-time to accommodate protocols such as Olivia.?? There are very short pauses which cause it to drop, but also one as long as 450ms between the TxID and the message body.? I understand the QDX does not have this problem with Olivia, does it use a different algorithm?

?

Thank you,

Stan KC7XE

Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

#133438

I am these BMS boards for couple of years now. As Dave pointed out they are excellent balancer/protection boards and he is abosultely right about not having the li-ion charging graph characteristic. For that I use one of these 12.6 V li-ion charge adapters:

?

The other option is to use this IC if you wanna build yours:

?

Though in my personal experience the 12.6 V charge adapter works pretty good.

?

--

?

73, Barb

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133437

Hi Paul,

I will probably have an answer to that question tomorrow. Thanks for jumping in with your thoughts about this.

73

Jim

VE4BWG

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133436

开云体育

My understanding is that it will stay in DFU mode, until you remove the jumper (and restart?).

73, Paul -- AI7JR

On 11/13/24 19:39, Jim VE4BWG via groups.io wrote:

Hi Dave.

Thank you. I'm gonna save that for tomorrow. But I am puzzled about the jumper that forced the QMX into DFU. Is its job finished or do I leave it connected until (fingers crossed) the QMX+ boots up? Or now that it is in DFU, should the jumper be removed??

Thanks all. 73.

Jim VE4BWG

-- 
Paul -- AI7JR

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133435

Hi Dave.

Thank you. I'm gonna save that for tomorrow. But I am puzzled about the jumper that forced the QMX into DFU. Is its job finished or do I leave it connected until (fingers crossed) the QMX+ boots up? Or now that it is in DFU, should the jumper be removed??

Thanks all. 73.

Jim VE4BWG

Re: QMX Mid Band Rx issue

#133434

Special thanks to Stan Dye. I got a new PCM1804 today and swapped it out. Used ChipQuik to remove old one and solder paste and hot air rework station to install new one. Ensured I used ESD strap while handling and very happy to report that the QMX now works as advertised.?

?

Stan, I learned a ton about signal tracing around the 1804 from you. Important thing of note: Even though I had good 12KHz sine waves on all 4 pins and what looked like data coming out, the 1804 was still the issue. I could not verify that it was a good data stream. I think the real proof is the bad IQ representation in the hardware test.

?

73

Bob

WV7W

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133433

Hi Jim

I might be wrong ( I generally am!) but it looks to me like you are actually trying to copy the file in the QMX+ from the QMX+ to your computer instead of the other way?around. IMHO the file required should be shown as from your local PC download folder, then dragged and dropped into the temporary QMX+ file. Worth double checking maybe?

Cheers de?

Dave VE7AHT

Show quoted text

On Wed, Nov 13, 2024 at 6:36?PM Jim VE4BWG via <willijt=[email protected]> wrote:

Thanks Stan. The forced reboot was successful and the Windows PC now recognizes the QMX+. I downloaded a fresh version of the update, unzipped it and when I attempted to either paste it or move it, I received this alert.

?

Re: Is it bricked? #dummy #firmware #problem #QMXplus #troubleshoot

#133432

Thank you Daniel.?