|
I put this here in case others have experienced this and might try this "solution" before taking your radio apart looking for some other problem.? ?On two separate occasions I have been unable to turn on my QMX.? After trial and error I discovered that with power applied to the transceiver, plugging the transceiver into the PC via the type C connector immediately allows it to be turned on and it works after that.? I don't know if this is a software issue where it gets into a state where it is expects a USB connection or some other issue.? I would be interested in knowing if others have experienced this.
?
Brien - KE7WB?
|
Brien, we don't know about the things around your QMX, which power supply or is there a ground connection from this power supply to your computer. In the moment I don't see a relation to the firmware. So maybe a flaky ground connection between the power supply and the QMX bypassed via the computer and USB cable? Or did you move the QMX PCB a little bit when plugging in USB and "repair" a cold solder joint?
?
It would be nice to know how much supply current when you try to switch on without success (and without USB connection). If nearly 0 the QMX is really off and connecting the computer couldn't trigger something from the firmware. If you see something around 100 mA (@ 9...12 V) the controller seems to work but showing no action via the display. And if you see something around 200 mA the controller may stuck somewhere in initialisation.
?
Waiting for some more information.
73 Ludwig
|
Ludwig,
?
Thanks for the reply.? ?I had taken it apart to check other possible causes. I found that when pushing the left knob down to start one of the connections on the power supplies went from 9V to 3.3 volts.? Not sure if it was supposed to and don't remember which connection it was but it seemed like it was trying to turn on but would not.? I checked many things, removed and re-seated the power supply boards, put it all back together and it would still not start.? I then plugged the PC into the USB port, pushed the left knob and it started right up.? When it would not start I was seeing 0 mA of current at 9V on my power supply current display.? Seems similar to the state it is in when there is no software loaded.
?
Brien
|
On Sun, Jan 26, 2025 at 02:04 AM, Brien - KE7WB wrote:
I found that when pushing the left knob down to start one of the connections on the power supplies went from 9V to 3.3 volts.
Brian,
good to know. While QMX is off one pin at only one of the power boards should show the supply voltage and all the other 0 V. When pressing the left encoder this voltage must stay (nearly) stable. Dropping this much (3.3 V) could result from
- a failure between the power barrel (QMX) and the SMPS pin or
- a (nearly) short after this pin causing a to high current and the external power supply is going into over current protection or
- a failure at the external power supply or
- the external power supply is going into over current protection without a to high current.
It's a pity not knowing more about your external power supply. Let's assume you are using a voltage converter (buck converter or so). Most(?) of these devices are using a resistor between GND in and GND out to measure the current. If GND of your power source is also connected to your computer the shield of the USB cable from the computer to the QMX could bypass this resistor. So the current protection couldn't kick in when using a USB cable.
?
When it would not start I was seeing 0 mA of current at 9V on my power supply current display.
0 mA means nothing is working inside the QMX. Without a working controller no USB data could flow (see my post before).
?
Seems similar to the state it is in when there is no software loaded.
No. If you switch on a QMX without loaded firmware you would see a current of around 200 mA.
?
73 Ludwig
?
?
|
Ludwig,
?
Thanks. I normally use a small buck converter between my station supply and my radio.? However, for testing I used my bench lab supply, a KORAD 3005D set a 9v, to exclude the buck converter.? When I took the radio apart I carefully checked the power jack connections and touched up the power board connections.? There is no wiggle or give in the type-C connection. At any rate if it happens again I will need to do more detailed testing.? I used my QMX for several hours outside yesterday for POTA and Winter Field Day with no problem, so for now it is?somewhat of a mystery.??
?
Brien
|
|
|
Brian,
?
Ludwig gives a good analysis of what might be going, though the symptoms are rather confusing.
It could be a bad ground connection from power supply to QMX (but then the drop you saw from 9v to 3.3v seems suspicious).
It could also be that the host computer's USB cable is shorting out the power supply's current sense resistor (but then it is not clear?
why the QMX usually works without the USB cable connected).
?
?
What kind of power supply do you have?? Does it have a current limit setting that might be set too low?
?
You said:? "I found that when pushing the left knob down to start one of the connections on the power supplies went from 9V to 3.3 volts."
Would be good to know if this was the Vin node of the QMX (see the schematic).
This does suggest Ludwig is correct that the power supply is in current limit, but a USB cable shorts out the current sense resistor in the power supply.
But does not explain why QMX often does work (I assume you are not changing the power supply current limit)
?
?
?
?
If you want help on this, you should answer Ludwig's question about what your power supply is.
?
Apparently you have a way to display the 9vdc at your power supply.
?
You also need to measure the current from the power supply to the QMX, either using a display on the power supply if it has one or by inserting a DVM into a break in the 12v wire of the cable configured to read Amps on the DVM.? If you are not sure how to measure current in that cable, ask for local help
as it is possible to mess up your DVM if it is done wrong.
?
I suggest you also measure the voltage across nodes V_IN and GND near the JP101 barrel connector, found at the bottom left of the first page of the QMX schematic.? This could be done by soldering wires to the QMX board at those nodes and then bringing those wires out of the QMX enclosure so they can be attached to a DVM set to measure DC.
?
Then wait till the rig decides not to work.? Report all three readings (two voltages, one of current) when the power-on switch is not pressed, also when the power-on switch is pressed.
?
Jerry, KE7ER
?
?
On Sun, Jan 26, 2025 at 05:16 AM, @Ludwig_DH8WN wrote:
On Sun, Jan 26, 2025 at 02:04 AM, Brien - KE7WB wrote:
I found that when pushing the left knob down to start one of the connections on the power supplies went from 9V to 3.3 volts.
Brian,
good to know. While QMX is off one pin at only one of the power boards should show the supply voltage and all the other 0 V. When pressing the left encoder this voltage must stay (nearly) stable. Dropping this much (3.3 V) could result from
- a failure between the power barrel (QMX) and the SMPS pin or
- a (nearly) short after this pin causing a to high current and the external power supply is going into over current protection or
- a failure at the external power supply or
- the external power supply is going into over current protection without a to high current.
It's a pity not knowing more about your external power supply. Let's assume you are using a voltage converter (buck converter or so). Most(?) of these devices are using a resistor between GND in and GND out to measure the current. If GND of your power source is also connected to your computer the shield of the USB cable from the computer to the QMX could bypass this resistor. So the current protection couldn't kick in when using a USB cable.
?
When it would not start I was seeing 0 mA of current at 9V on my power supply current display.
0 mA means nothing is working inside the QMX. Without a working controller no USB data could flow (see my post before).
?
Seems similar to the state it is in when there is no software loaded.
No. If you switch on a QMX without loaded firmware you would see a current of around 200 mA.
?
73 Ludwig
?
|
Brian,
?
The KORAD 3005D looks like a very nice supply.
It claims to have a linear regulator, can display both voltage and current, and allows adjusting both voltage and current.
My guess is that the supply is isolated from the wall plug safety ground.
?
I do suggest you solder those wires inside the QMX to monitor the voltage at the barrel connector when the rig is mis-behaving.
You could have a bad connection somewhere in the cable or connectors from power supply to QMX.
?
Jerry? KE7ER
|
Jerry,
?
A power fault would have to be within the QMX as two separate power cables were used, one from the buck converter when the problem first occurred and a second one from the lab power supply.? Current limit was set at 1.5A and no current when the left knob was pressed, until after I connected the USB cable.? I did try multiple times to wiggle the power connector while connected and it made no difference so if it is an intermittent in the barrel connector it is pretty fickle.? The USB port does not exhibit any mechanical play.? You are correct, the lab supply negative output is isolated from ground unless of course you jumper ground to negative.? I did not.??
?
Brien
|
Brian,
?
Does sound like you know what you are doing.
You have a very weird set of symptoms.
?
I'm having a hard time imagining how plugging in the USB cable makes it work if there is a solid ground connection from power supply into the QMX and if the power supply is at a 1.5A current limit.
?
The node inside the QMX dropping from 9v to 3v is a mystery, would be good to investigate that further.
My best guess is that you were looking at the V_IN node.
?
I'd operate the rig for awhile outside the case so it can be probed with a voltmeter when it fails.
First see if you have your power supply voltage between QMX GND and V_IN on the QMX.
?
At power up with the power-on knob pushed, the PWR_ON node pulls the gates of the PFET's at Q103 and Q105 down close to ground,
allowing power to flow through them to the node +12V.? Check to see if you have your power supply voltage at +12V.
?
You said:? "Current limit was set at 1.5A and no current when the left knob was pressed, until after I connected the USB cable."
That's just weird if you really have 9v from the power supply going into the QMX, and you can measure 9v between V_IN and GND
on the QMX board.? The STM32F processor on the QMX draws around 0.1 Amps from the 3.3v rail, and you would see some?
current draw from your lab supply when the processor is working.? Assuming this isn't a fault in the ground cabling somehow,
the only way the USB port can affect what happens on the rig is if the processor is working.
?
Jerry, KE7ER
?
?
?
On Sun, Jan 26, 2025 at 11:03 AM, Brien - KE7WB wrote:
Jerry,
?
A power fault would have to be within the QMX as two separate power cables were used, one from the buck converter when the problem first occurred and a second one from the lab power supply.? Current limit was set at 1.5A and no current when the left knob was pressed, until after I connected the USB cable.? I did try multiple times to wiggle the power connector while connected and it made no difference so if it is an intermittent in the barrel connector it is pretty fickle.? The USB port does not exhibit any mechanical play.? You are correct, the lab supply negative output is isolated from ground unless of course you jumper ground to negative.? I did not.??
?
Brien
|
Ludwig,
?
After looking at the schematic I think I can come up with a plausible explanation for my symptoms.
?
When the left knob is pressed it provides PWR_ON, passing the input voltage, in my case 9VDC to the 3.3V converter.? If power PWR_HOLD does not come from the processor then it will not latch the power on.? The point that 9V will show up as 3.3V when the knob is pressed? is output of IC101 linear 3.3V regulator which appears to be there to bootstrap VDD to the processor until it can provide PWM to the 3.3V switched mode power supply.? So I am thinking that this was the point I was measuring when I saw 9V turn into 3.3V and go back to 9V when I let off the knob.? Clearly the 3.3V PWM supply was not turning on.? This rules out an intermittent at the input power connector.??
?
That the unit was not coming up when I ran these tests suggests that there is a possible fault between the processor coming up and providing PWR_HOLD and/or PWM to the 3.3V SMPS (could be caused by an intermittent in VDD from the 3.3V regulator output to the processor).?
?
The part I cannot explain yet is why hooking up the USB resolves this especially assuming it is a mechanical intermittent.? The schematic I am looking at does not show a VDD connection from the USB port, although it is possible that it is not shown.? It does show the USB port connected to pins 70 and 71 of the processor.? These are port pins so not directly connected to VDD unless a weird path through internal pullups, which seems not plausible.
?
So the next time this happens I will be looking at VDD to the processor at L202, PWR_ON, PWM_3V3 to see if I can narrow it down.??
?
Brien - KE7WB
?
?
?
?
|
After studying this a little further I am thinking LIN_REG_EN is not making it to the 3.3V supply. The the linear regulator could show the 3.3V when the knob is pressed but if it is not being passed through to the VDD output then the processor would not be coming up.? ?Hooking up the USB could provide parasitic voltage through internal pullups, enough to turn Q110 on, but this is an unproven theory.? It does not explain why the behavior goes away though.? Another path to look at.
?
Brien - KE7WB
|
On Sun, Jan 26, 2025 at 03:22 PM, Brien - KE7WB wrote:
why hooking up the USB resolves this especially assuming it is a mechanical intermittent.? The schematic I am looking at does not show a VDD connection from the USB port,
But the USB connector DOES provide a ground return, that should be a clue.?
What happens if you provide an explicit ground from the power supply negative to the radio chassis, like the RF connector or the PCB ground?
73, Don N2VGU
|
Brien,
?
Worth looking at LIN_REG_EN with a voltmeter.
However,?R101 should pull up LIN_REG_EN just fine, turning on Q102 to get "+12V" to U101 (the 78M33).
At that point the processor starts running and will drive LIN_REG_EN high till ADC_3V3 shows that the SMPS is working,
then the processor drives LIN_REG_EN low..
Until the processor is running, LIN_REG_EN out of the processor should be high impedance.
?
You said: "Hooking up the USB could provide parasitic voltage through internal pullups, enough to turn Q110 on,"
Q110 has nothing to do with it.? Q100 loads up the 3V3 SMPS with 3.3v/47=70ma while it is starting up.
This avoids a sudden increase in current when LIN_REG_EN goes low, which might drag down the VDD rail momentarily.
LIN_REG_EN goes low to start drawing power for the processor from the SMPS instead of the 78M33.
?
First step is to verify you have power at the +12V node when pushing the knob and rig fails to start up.
Also verify that LIN_REG_EN is high when the knob is pushed.
At that point the VDD rail should be getting 3.3V from the 78M33.
The processor then comes to life and starts pumping PWM_3V3 at about 100 khz to bring up the 3.3v SMPS.
Once ADC_3V3 looks good to the processor, it lowers LIN_REG_EN and power is supplied by the SMPS instead of the 78M33.
?
I suppose it's possible that R101 has a cold solder joint, and so LIN_REG_EN is not going high unless you have
some internal leakage from the USB port, but I sincerely doubt that's the case.
Here's an old post that describes how the LIN_REG_EN pin effectively has an internal zener from that pin to ground, not the more typical ESD protection diode (or parasitic diode) to the Vdd rail: ? /g/QRPLabs/message/133441That old thread describes a similar power-on problem, except in this case the processor turned out to be dead in the water
with the LIN_EN_REG pin shorted to ground.? I doubt that has anything to do with your problem.
?
Jerry, KE7ER
?
?
On Sun, Jan 26, 2025 at 12:38 PM, Brien - KE7WB wrote:
After studying this a little further I am thinking LIN_REG_EN is not making it to the 3.3V supply. The the linear regulator could show the 3.3V when the knob is pressed but if it is not being passed through to the VDD output then the processor would not be coming up.? ?Hooking up the USB could provide parasitic voltage through internal pullups, enough to turn Q110 on, but this is an unproven theory.? It does not explain why the behavior goes away though.? Another path to look at.
?
Brien - KE7WB
|
Jerry,
?
Thanks.? Now that I understand it a little better, I will take a look at this the next time it fails or the next time I have it apart.??
?
Brien -KE7WB
|
On Sun, Jan 26, 2025 at 09:22 PM, Brien - KE7WB wrote:
The point that 9V will show up as 3.3V when the knob is pressed? is output of IC101 linear 3.3V regulator which appears to be there to bootstrap VDD to the processor
Sorry, no. Your 9 V shold be switched to the signal "+12V" feeding also IC101. To produce around 3 V at VDD the input voltage for IC101 must be at least around 5.5 V. You will find this reading the QRP Labs Wiki doc mentioned in my post early today.
?
Let's assume there is a higher resistance in series to your power supply. With left encoder not depressed the current is 0 and no voltage drop across the resistance. When pressing the encoder the QMX would pull current and you have a voltage drop across the resistance. Input current of the QMX goes not linear along the voltage. Up to approx. 3...4 V you have nearly no current. Further increasing the input voltage will increase the current significantly and so the voltage drop across the resistance.
The the linear regulator could show the 3.3V when the knob is pressed
No, not when signal "+12V" is only at 3.3 V. The problem seems to be a large voltage drop before "+12V and IC101
?
73, Ludwig
?
|
Don could be right, the USB cable providing a needed ground return was my first thought.
But Brian has moved to what looks like a properly isolated lab supply, and the problem persists.
The reported symptoms don't make sense to me.
?
Brien is correct, the USB port only has ground connections and the differential data pair into IO pins of the processor.
No VDD wire, the host computer cannot supply power to the QMX.
?
Jerry, KE7ER
?
On Sun, Jan 26, 2025 at 01:04 PM, Donald S Brant Jr wrote:
On Sun, Jan 26, 2025 at 03:22 PM, Brien - KE7WB wrote:
why hooking up the USB resolves this especially assuming it is a mechanical intermittent.? The schematic I am looking at does not show a VDD connection from the USB port,
But the USB connector DOES provide a ground return, that should be a clue.?
What happens if you provide an explicit ground from the power supply negative to the radio chassis, like the RF connector or the PCB ground?
73, Don N2VGU
|
Minor correction:
?
In? ?"Q110 has nothing to do with it.? Q100 loads up the 3V3 SMPS with 3.3v/47=70ma while it is starting up."
Should read "Q110" instead of "Q100"
?
On Sun, Jan 26, 2025 at 01:29 PM, Jerry Gaffke wrote:
Q110 has nothing to do with it.? Q100 loads up the 3V3 SMPS with 3.3v/47=70ma while it is starting up.
|
Brien might have this figured out, though his wording is not completely clear.
?
When Brien says:? "The point that 9V will show up as 3.3V when the knob is pressed? is output of IC101 linear 3.3V regulator which appears to be there to bootstrap VDD to the processor"
He may mean that when the knob is pressed, the output of IC101 is 3.3v, since IC101 is regulating the 9v from V_IN down to 3.3v on VDD.
It is possible that Brien does not mean that the measured voltage on schematic nodes V_IN or +12V suddenly changes from 9V to 3.3v when the knob is pressed.
?
In an earlier post, Brien had said "?I found that when pushing the left knob down to start one of the connections on the power supplies went from 9V to 3.3 volts.? Not sure if it was supposed to and don't remember which connection it was but it seemed like it was trying to turn on but would not."
This is probably what Ludwig remembers when he now reads that "9V will show up as 3.3V when the knob is pressed".
However, it might be that early on when Brien was considerably more confused, he measured 9v on one of the SMPS pins, then pressed the knob and measured 3.3v, perhaps on some other SMPS pin.? After all he does say he doesn't remember which connection he was measuring.
?
Adding to the confusion, Brien said:? "I normally use a small buck converter between my station supply and my radio.? However, for testing I used my bench lab supply, a KORAD 3005D set a 9v, to exclude the buck converter."
Perhaps some of the results he has reported were from before he switched to using the KORAD supply.
?
To move forward, I think Brien needs to identify with certainty where to measure voltages on the LIN_REG_EN, V_IN, +12V, and VDD nodes
(with the other probe on GND).? Use the schematic to do this, also the part placement drawings in the assembly manual,
and perhaps look up datasheets for each part on the web to determine which pin is which.
?
Then when the rig misbehaves, report all four of those voltages, and also the voltage and current shown on the KORAD display.
Give all 6 of these numbers when the knob is not pressed, and also when the knob is pressed.
?
Jerry, KE7ER
?
?
?
On Sun, Jan 26, 2025 at 01:37 PM, @Ludwig_DH8WN wrote:
On Sun, Jan 26, 2025 at 09:22 PM, Brien - KE7WB wrote:
The point that 9V will show up as 3.3V when the knob is pressed? is output of IC101 linear 3.3V regulator which appears to be there to bootstrap VDD to the processor
Sorry, no. Your 9 V shold be switched to the signal "+12V" feeding also IC101. To produce around 3 V at VDD the input voltage for IC101 must be at least around 5.5 V. You will find this reading the QRP Labs Wiki doc mentioned in my post early today.
?
Let's assume there is a higher resistance in series to your power supply. With left encoder not depressed the current is 0 and no voltage drop across the resistance. When pressing the encoder the QMX would pull current and you have a voltage drop across the resistance. Input current of the QMX goes not linear along the voltage. Up to approx. 3...4 V you have nearly no current. Further increasing the input voltage will increase the current significantly and so the voltage drop across the resistance.
The the linear regulator could show the 3.3V when the knob is pressed
No, not when signal "+12V" is only at 3.3 V. The problem seems to be a large voltage drop before "+12V and IC101
?
73, Ludwig
|
On Sun, Jan 26, 2025 at 11:54 PM, Jerry Gaffke wrote:
To move forward, I think Brien needs to identify with certainty where to measure voltages on the LIN_REG_EN, V_IN, +12V, and VDD nodes
(with the other probe on GND).?
?
I have found it helpful to solder fine wires, color-coded, to the various nodes and bring them out where they are easy to access.? Makes voltage readings easy and less chance of an "oops!" slip shorting something out.? I tape them to the bench with the ends readily accessible for connection but spread out so they can't short to others.
I use this, it has 8 colors and works great: ?
?
73, Don N2VGU
|
Brien - KE7WB
Donald S Brant Jr