开云体育

Hi

?

I built a QMX mid-band, it started up first time and everything seemed pretty normal except for low power output.

?

tx into a dummy load with a 11.9v supply:

?

Dan, you are not the first who has slipped with a probe.? Many of us have done that.

Yes, it's probably just the smps - at least we can hope so.? It's easy to short the 12V to the 5V or 3V rail, which will burn out one or more of the protection diodes like fuses.? And thankfully so, since they normally spare us from more damage to other parts.

And Yes, following that thread on testing the power supplies is a good place to start.? The wiki here also has a procedure to independently test the smps cards.

?

?If you have a lower-voltage power supply to use while doing tests, it will help prevent possible further damage - like 7-9V, preferably with a current limit.? But I think you said you don't have the current limit, so it's ok to use whatever you have, just be careful.

?

I haven't gone back and read that whole thread, but I would start with more measurements with everything still put together, except the bottom cover off.? Identify the pins on the power supply connectors for V+(12V), Vcc and Vdd, and a good ground point to measure against.?? Push the power-on button and hold it in while measuring the values on each of those points.? Don't keep the power button pressed for a long time, since something is likely shorted and will be heating up.? These measurements can help you (and us) know what you are starting with, and where to focus efforts.? While you are at it, bonus for measuring the other signals, like Power_On, Power_hold, Lin_Reg_En.

?

Stan KC7XE

Hi Stan

?

Thanks for the pointers.

?

I got hold of a current-limiting PSU, and did the tests suggested in the wiki here: /g/QRPLabs/wiki/37128

?

on PCB#1:

step 2, I got Vcc = 4.1v (should be 0v)

step 3, I got Vcc = 4.2v (should be 0v), I stopped there

?

on PCB#2

step 1, I got Vdd = 0.67v (should be 3-3.3v). I stopped there

?

Here's the results with a 8v supply, 100mA current limit, holding down the power switch:

?

(on PCB#1)

12v goes to 5.3v?

Vcc goes 0.2v to 2.1v

pwr_on goes 7.4v to 0v

pwr_hold stays at 0v

?

(on PCB#2)

12v goes 0v to 5.2v

Vdd goes 0v to 0.1v

lin_reg_en goes 0v to 0.01v

?

when I do click the power button, current limits at 100mA every time.

?

Hope this is useful

Cheers

Dan

?

?

Are your tests with the power supply boards unplugged?? Not connected to anything?? If so then quite a lot of both boards seem to be bad.? They are both shorted to ground enough that the current limiting is restricting the voltage.? You talk about holding the start button down, that won't be connected if the boards are unplugged.

?

100mA is a bit low for a test with the boards fitted because the processor takes about 200mA on start up.? It might be worth increasing the current limit and see if VDD will get to 3V.? Don't go over 250mA, if that isn't enough there's something failed.

?

Chris, G5CTH

Dan,

?

never ever run a QMX with SMPS boards if the off line tests are not ok!!!

It looks like a short at VDD (or a faulty PCB#2).

Please remove all subboards and check resistance at the mainboard VDD-GND and VCC-GND. Both should be some kOhm. If only a few Ohm one or more element(s) connected to this voltage rail is broken (or the rail is shorted by a solder whisker or so).

?

If resistance is not ok, you have to search for the broken element, not so easy.

?

If resistance is ok you have to repair both SMPS boards or use spare parts from QRP Labs.

?

If all is ok you need for start at least 200 mA, use 250 mA limitation @ 6.5 V.

?

73 Ludwig

Chris, first half of tests was done with smpsu boards removed from main board.?

?

Second part, pressing the power switch, was re-assembled.

Ludwig

?

Thanks for your coment, with sub-boards removed, I have 6.6k ohm gnd to Vcc and 2.3k ohm gnd to Vdd.

?

Do you think this indicates something broken on the main board?

?

Dan

?

Dan,

both values are not bad. I guess your controller STM32 isn't broken.

So like I wrote before:

73 Ludwig

Dan, yes, your SMPS boards are both bad.? It must have been a bad short you made with that probe.

?

If you want to try to repair them yourself:

- On the 3.3V SMPS, it is most likely D109 that is shorted.? You can remove it, and see if the offline tests work.? If the tests work, all you need to do is replace D109.

- On the 5V SMPS, it is more difficult to say.? Something is turning on Q108 when it is supposed to be off - or maybe it is shorted.? You can look at the PWM_5V signal (during offline test step 2 or 3) and see if it is low (near 0V) like it is supposed to be.? If it is high, that is what is causing the problem and you need to find the reason (a short in the soldering or missing R108 or whatever).? You can also look carefully at Q108 and L101, and see if either looks burnt.

?

Or if you want, just order a new pair from qrp-labs for $10.

In either case, when you have functional SMPS cards, install them, and power on only with a current-limited (~250mA) supply at about 7V, in case something else undetected has failed on your board.? Hopefully the processor comes up and you can use the diagnostic page of the terminal interface to verify that things are working OK before giving it full voltage/current and trying transmit diagnostics like LPF and SWR scans.

?

Good luck

Stan KC7XE

Hi Dan,

?

Before investing in the new SMPS boards, I suggest first verifying that there is no problem with the motherboard.? With both SMPS boards removed, measure the resistance to ground of the Vdd and Vcc lines.

?

Attached is a PDF with Vcc and Vdd identified.

?

73

Evan

AC9TU

reminder: this radio booted up and transmitted before I shorted something near the PA.?

?

update: new SMPSU boards arrived, I tested before installation, as per the wiki article, and they seem fine.

?

I have the SMPSU boards and controls board installed into the main board.?

?

With 7.5v, 200mA current-limited supply, I have zero current before pressing power on switch.

?

Pressing the power switch briefly I get these on the back of the PCB:

?

Board 1:

+12v: 6.0v

Vcc: 0v

adc5v: 0v

pwr_hold: 0.01v

Vin drops to 6.0v?

pwm_5v: 0.01v

pwr_on 0v

?

board 2:

Vdd: 0.41

+12: 6.0v

lin_reg_en: 5.0v

pwm_3v3: 0v

?

With all sub-boards removed, I'm seeing 1.3kOhm gnd->Vcc and 2Ohm (two ohms) gnd->Vdd, so those measurements have got worse.....

?

I then took the SMPSUs out and re-tested, they still pass the tests from the wiki.

?

Any ideas where to look?

Cheers

Dan

?

On Fri, Apr 4, 2025 at 06:57 PM, Dan G7NNZ wrote:

2Ohm (two ohms) gnd->Vdd

Dan, different to Mar 23 (2.3 k) but why? Some element connected to Vdd maybe broken, in the worst case the controller. I will you more info later.

One question; do you have access to a good bench power supply (variable voltage and current limit)?

?

73 Ludwig

Hi Ludwig

?

Yes i now have variable voltage and current supply. Resolution of current limit is 100mA.

?

Yes a different result, i was surprised, too. This was measured after trying the new SMPSUs. I haven't knowingly changed anything else?

?

Dan

?

My QMX+ died the same way on first start-up smoke test. Perhaps what I have found below may help. You might also want to kill the PA voltage by lifting a wire on L502. ?I did find the known, early manufacturing solder fault on the 5V board. My PA is now crowbarred with IC204 CLK2 into IC501 is 100% @ 3.3V on pin 4. Not sure yet if IC204 or the main mCPU are faulted, more investigation in a couple days. For now, I have the 12v bias on TX turned off in the advanced settings and enjoying the RX.

Dan, its a bit risky but I did it with success. You may do it on your own risk!!!

Set the supply to 3.0 V and current limit to 200 mA. Remove all subboards and connect the bench power supply direct to Vdd and GND. I guess the voltage will drop. Check all IC connected to Vdd for temperature. If you find one hot or significant warm, his should be the culprit. If no IC is hot increase the current to 300 mA, maybe later 400 mA but not more. But be sure while increasing the current the voltage is not going over 3.0 V.

?

2 Ohm and 0.4 A will produce 320 mW heat, enough to feel it.

?

73 Ludwig

On Fri, Apr 4, 2025 at 09:49 PM, Damon wrote:

with IC204 CLK2 into IC501 is 100% @ 3.3V on pin 4. Not sure yet if IC204 or the main mCPU are faulted,

Damon, I wouldn't enjoy the RX. I wouldn't switch on the QMX in this state. Maybe voltage at IC204 is wrong because of IC503 died together with the PA. Best way is to remove IC503 and check output of IC501 and IC502.

?

And if the main mCPU would be faulted you couldn't switch on the QMX.

?

73 Ludwig

thanks again, Ludwig

?

CPU barely warmed at 200mA, and at 300mA I got a voltage drop to 2.5v and the CPU did warm slightly. I couldn't detect any other components warming.?

?

Is it game over?

Dan

?

?

Dan, a high probability for a broken controller. Do you have a sharp tip for your DMM to probe voltage at a pin of the controller?

Driving 300 mA you may check voltage drop from Vdd connection to the mainbord to each of the five Vdd pins of the STM32. If high current is flowing into the STM32 you should see a drop of much more than 10 mV. Last proof would be to remove the STM32 and check resistance Vdd - GND.

?

73 Ludwig

?