开云体育

I have a strange problem where often the QMX boots up fine but after a few seconds, the screen goes darkish in terms of contrast where I have to angle it to read characters. At the same time, the serial port seems to drop out. Rebooting or pulling power often fixes it but only for a few more seconds on another power cycle, but also often causes the first row of the display to show bars and no backlight on reboot (assuming this means LCD does not communicate). Unit seemed to be working OK last night, so thinking this could be a dodgy connection somewhere.

Curiously, when I run diagnostics in the terminal, I get the attached image. My supply voltage is 8 NiMH AA batteries measuring about 10v under load, so I don't know how it could see a supply voltage of 16V. I've tried tracing around a bit but looking for suggestions
.

Just an "educated" guess since I don't have one, but microprocessors can't measure higher voltages directly.?
Since you know the voltage is not that high, I suggest you look at the resistors that make up the voltage divider to the Analog to Digital converter.
They are probably not in the correct order to do the ratio that the microprocessor is expecting.

-------
Robert Giuliano
KB8RCO

On Fri, Sep 29, 2023 at 09:40 AM, Rob Giuliano wrote:

Just an "educated" guess since I don't have one, but microprocessors can't measure higher voltages directly. Since you know the voltage is not that high, I suggest you look at the resistors that make up the voltage divider to the Analog to Digital converter. They are probably not in the correct order to do the ratio that the microprocessor is expecting.

If that were the case, I would expect that the same problem happens reliably every time regardless of how many power cycles.

Note that the duty cycles are maxed out.? This would indicate that the voltage measurements are off and the values are in error (a high value should have a low duty cycle).? It also means that the SMPSs are not working correctly.? I would look for an open or short on the SMPS connectors.? It could also be, as Ryuji suggests, that the voltage divider is off.??

Trying to test with power applied for this design can be problematic.? The protection circuits can handle only so much abuse before they, too, give up the magic smoke.? A low-voltage supply that is current limited would be a good idea.? Better is to isolate the fault without power if possible.

73
Evan
AC9TU

Thanks for the suggestinons! I spent some more time probing around and testing things as well as reflowing the pins to the SMPS boards. I've noticed now that the VCC line is at 5.6V, so I believe for whatever reason the zener is clamping it. Can anyone help me with some measured voltages around Q108, Q106, and Q112? I see both V_IN and +12V reading about 10V as expected, so it appears there is some issue at the 5V supply.

It also appears this is somewhat intermittent. I was able to boot the radio up, connect to the terminal, and diagnostics read everything correct, but then a few seconds later the terminal lost connection and I saw the same contrast issue on the display.

Another development that might be telling - I just connected and the diagnostics very quicky upon boot, and it shows the errant measured supply voltage and VCC by by the processor ramp up in voltage over the course of several seconds. It starts out as good, and then voltage creeps up. Is this suggesting a processor issue? I instrumented the "ADC_BATT" line and it holds steady at about 1.7V which is what I calculated as expected value for a 10V supply.