QCX-mini failure on new build


Timothy Freeze
 

Well I just finished my 20 meter mini and it doesn't work.  I'm very bummed about it.  I've been through the preliminary checks, tried starting with control board removed and got the same results.


I was very careful and inspected every solder joint as I built it.  The only mistake was I put the pins for the control board on the  main board,  then I had to put the socket on the control board.  That seems fine
and not worth trying to remove them to reverse.

When it starts up, this is what I get:



I haven't downloaded the full trouble shooting PDF but right now I'm really discouraged.

I've going out of town this weekend for a 3 day bike tour in Williamsburg, Va,  so it will have to wait till later.   

I'd appreciate maybe pointing me in the right direction. I'm even thinking of trying another kit and starting over.  Maybe I'll have better luck. and have some spare parts.

Kindest regards,
Tim


Dean Smith
 

Looks like a lack of comms from lcd to Atmega to me.
Check all pins on the lcd connector to Atmega, no open or short circuits.


Dean Smith
 

Looking at ur picture, the second row up on the left hand side, looks dodgy short maybe? Could be just the photo tho.
Any more pictures of ur board could help,as there are lots of sets of eyes on here hi.


Timothy Freeze
 

Thanks for looking. I did put a scope on the 20 MHZ oscillator and no oscillations at all.  It does show about 2 volts dc to ground.  

I ordered some 20 MHZ crystals to try that.

I've inspected all the solder joints with a magnifying camera. The kind with a little screen.  I'll post some photos of the board when I get home Monday.

I cleaned all the flux off with IPAlcohol so I could inspect the solder joints.  Doesn't mean I didn't screw something up, just that  I tried really hard not to.

Any ideas on the 20MHZ oscillator, things to check?

Kindest regards,
Tim


Alan G4ZFQ
 

Any ideas on the 20MHZ oscillator, things to check?
Tim,

If you are sure that connecting the scope did not stop the oscillator then there is not much to go wrong. Just solder joints, the crystal.

But I'm not sure the crystal is not working. It looks as if the LCD has been initialised, but then maybe it stopped?
I always try to hear it on a receiver around 20 to 20.005MHz

73 Alan G4ZFQ


Evan Hand
 

Hi Tim,
Quick thought, did you inspect the pins of the ATMega328p IC2?  Sometimes one of them gets rolled under when inserting into the socket.

Good luck in finding the issue.
73
Evan
AC9TU


Timothy Freeze
 

Yes I checked the AtMega pins, no problems there.

Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Evan Hand <elhandjr@...>
Sent: Friday, June 18, 2021 5:53:26 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Hi Tim,
Quick thought, did you inspect the pins of the ATMega328p IC2?  Sometimes one of them gets rolled under when inserting into the socket.

Good luck in finding the issue.
73
Evan
AC9TU


Timothy Freeze
 

I did get some audio out on the initial start up.

The crystal was checked with a quality scope with a 10 x probe absolutely nothing there except DC.

I'm afraid it might be the crystal or the processor. I've ordered some crystals, really cheap. When I get home I'll remove the crystal and test it.

Kindest regards,
Tim
KO4ELL

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Timothy Freeze <s35pilot@...>
Sent: Friday, June 18, 2021 7:11:04 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Yes I checked the AtMega pins, no problems there.

Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Evan Hand <elhandjr@...>
Sent: Friday, June 18, 2021 5:53:26 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Hi Tim,
Quick thought, did you inspect the pins of the ATMega328p IC2?  Sometimes one of them gets rolled under when inserting into the socket.

Good luck in finding the issue.
73
Evan
AC9TU


Al Holt
 

Good luck on your bike ride!
It'll be good to get out away from the project and get some fresh air.
No doubt that during the ride some new perspectives on the problem will come to mind while you are on the road. 

--Al
WD4AH


Arthur Paton <artpaton@...>
 

Hi Tim

Some things in the positive camp.  Your display is working since you have characters in the display and the full battery indicator.  You might start with checking connections between the display and the display driver.  Also you might recheck the pins under the main processor chip;  It is so easy to fold one under.  Work slowly and carefully.  My guess is you will find the issue.

73
Art N9AEP


Evan Hand
 

Hello Tim,
If you have an older version of the ATMega firmware, you might want to order a new one from QRP-Guys.  That would give you a new processor and the latest firmware.  The newest version is 1.07a.  It is $6 on the web page.

73
Evan
AC9TU


Dean Smith
 

you mean QRP Labs ;)


Evan Hand
 

Hi Fran,

Yes you are correct. It is QRP-Labs. 


Thank you for the correction 
73
Evan
AC9TU


Dean Smith
 

Someone mentioned battery full indicator on Tim's display, i think you'll find it's a blank block.
With fault finding most times, it comes down to assembly faults rather than faulty components.
Providing the ATmega328 hasn't been inserted wrong way round then powered on, this IC should be ok.
Chances of a Crystal failing are quite slim.
The lines that run from the LCD to the ATMEGA are more than likely to suffer, with solder bridges, or open connections (Dry joints).
Remember there are at least 4 connections on each rail from the LCD, time it gets to the microcontroller, then times this by 12 active rails on the LCD.
There is a pretty good chance at least one.... will play up!!
D0-D7 lines will just cause odd characters, if opened or shorted. whereas R/W line,or +5v line could see no display at all.
The four middle pins on the Display are not used. (Best not soldered up)
Some of the lcd lines have another purpose ie., CAT interface. So expect at least two of the I/O lines to wander off Hi.
Chances are it's something really simple, and when you do find it, it will make your day!
Dean (G7EOB)


Timothy Freeze
 

Thanks for all the tips. I am still curious why I'm not seeing the 20 MHz oscillator on pins 9 and 10. I've carefully checked the processor pins and removed and reinserted it.
I'm sure it's likely something I've done, I've not found any bad joints or bridges.

I've ordered a 40  M kit and could use that processor, but I'm afraid I might damage that one.

It's a brand new kit and should be the latest firmware.
I'll check the items suggested when I get home.  Thanks for helping.

Back to my bike ride, 40 miles and half way.

Kindest regards,
Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Dean Smith <bardezbiker@...>
Sent: Friday, June 18, 2021 10:53:35 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Someone mentioned battery full indicator on Tim's display, i think you'll find it's a blank block.
With fault finding most times, it comes down to assembly faults rather than faulty components.
Providing the ATmega328 hasn't been inserted wrong way round then powered on, this IC should be ok.
Chances of a Crystal failing are quite slim.
The lines that run from the LCD to the ATMEGA are more than likely to suffer, with solder bridges, or open connections (Dry joints).
Remember there are at least 4 connections on each rail from the LCD, time it gets to the microcontroller, then times this by 12 active rails on the LCD.
There is a pretty good chance at least one.... will play up!!
D0-D7 lines will just cause odd characters, if opened or shorted. whereas R/W line,or +5v line could see no display at all.
The four middle pins on the Display are not used. (Best not soldered up)
Some of the lcd lines have another purpose ie., CAT interface. So expect at least two of the I/O lines to wander off Hi.
Chances are it's something really simple, and when you do find it, it will make your day!
Dean (G7EOB)


Evan Hand
 

Tim,

An alternate testing method could be to try the processor from the 20meter QCX Mini in the new 40meter QCX.  If the 40meter QCX works with the 20meter processor you know there is something else wrong.  If not, and the 40meter works when you plug in the new processor, most likely the old processor is bad.

This is the second QCX that has had this type of problem that I have read about on the board.  The first one that I know of was a 40meter QCX+.  The processor was bad.

It is always good to have spare parts for testing ;-)

Just a suggestion.
73
Evan
AC9TU


Timothy Freeze
 

Evan,
Thanks for that suggestion. That way I'm not likely to mess up the new processor.

Kindest regards,
Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Evan Hand <elhandjr@...>
Sent: Friday, June 18, 2021 2:18:19 PM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Tim,

An alternate testing method could be to try the processor from the 20meter QCX Mini in the new 40meter QCX.  If the 40meter QCX works with the 20meter processor you know there is something else wrong.  If not, and the 40meter works when you plug in the new processor, most likely the old processor is bad.

This is the second QCX that has had this type of problem that I have read about on the board.  The first one that I know of was a 40meter QCX+.  The processor was bad.

It is always good to have spare parts for testing ;-)

Just a suggestion.
73
Evan
AC9TU


Timothy Freeze
 

Thanks all for the suggestions and help.

I've got the 20 MHZ oscillator check done and it's working - about 1.3 V at 20 MHz.

I've removed the solder from 75% of the build and re-soldered them.  Completely re-soldered the display, IC2, and most other joints.  If it didn't look perfect, I wicked the solder and re-did it. They look good and I'm thinking the microcontroller got corrupted when I first started it.  I did find that my power supply is very noisy (switching type).  So I put the QCX on the shack ps and it's nice clean power.

The first start of the QCX I used the power supply to ramp up the voltage to 12, there was an audible tone from the ear buds. then it quit.  So I'm thinking the noisy power supply or a bad solder joint supply might have screwed up the processor.  I've ordered a replacement from Hans but that will take a while to get.

I started to look into flashing the firmware but the PDF on how to do it assumes starting with a working QCX.   Any advice on how to do this or the best way is appreciated.

Kindest regards,

Tim


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Timothy Freeze <s35pilot@...>
Sent: Saturday, June 19, 2021 8:22 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Evan,
Thanks for that suggestion. That way I'm not likely to mess up the new processor.

Kindest regards,
Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Evan Hand <elhandjr@...>
Sent: Friday, June 18, 2021 2:18:19 PM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Tim,

An alternate testing method could be to try the processor from the 20meter QCX Mini in the new 40meter QCX.  If the 40meter QCX works with the 20meter processor you know there is something else wrong.  If not, and the 40meter works when you plug in the new processor, most likely the old processor is bad.

This is the second QCX that has had this type of problem that I have read about on the board.  The first one that I know of was a 40meter QCX+.  The processor was bad.

It is always good to have spare parts for testing ;-)

Just a suggestion.
73
Evan
AC9TU


howard winwood G4GPF
 

I think you are on the wrong track, a bad power supply or solder joint  CAN NOT screw up the processor, try thinking through the process logically rather than assume.
The processor either works or it doesn’t (highly unlikely that it doesn’t).
Check all components for value and placement - YOU messed up somewhere……...

On 21 Jun 2021, at 22:44, Timothy Freeze <s35pilot@...> wrote:

Thanks all for the suggestions and help.

I've got the 20 MHZ oscillator check done and it's working - about 1.3 V at 20 MHz.

I've removed the solder from 75% of the build and re-soldered them.  Completely re-soldered the display, IC2, and most other joints.  If it didn't look perfect, I wicked the solder and re-did it. They look good and I'm thinking the microcontroller got corrupted when I first started it.  I did find that my power supply is very noisy (switching type).  So I put the QCX on the shack ps and it's nice clean power.

The first start of the QCX I used the power supply to ramp up the voltage to 12, there was an audible tone from the ear buds. then it quit.  So I'm thinking the noisy power supply or a bad solder joint supply might have screwed up the processor.  I've ordered a replacement from Hans but that will take a while to get.

I started to look into flashing the firmware but the PDF on how to do it assumes starting with a working QCX.   Any advice on how to do this or the best way is appreciated.

Kindest regards,

Tim


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Timothy Freeze <s35pilot@...>
Sent: Saturday, June 19, 2021 8:22 AM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Evan, 
Thanks for that suggestion. That way I'm not likely to mess up the new processor.

Kindest regards,
Tim

Sent from my Verizon, Samsung Galaxy smartphone


From: QRPLabs@groups.io <QRPLabs@groups.io> on behalf of Evan Hand <elhandjr@...>
Sent: Friday, June 18, 2021 2:18:19 PM
To: QRPLabs@groups.io <QRPLabs@groups.io>
Subject: Re: [QRPLabs] QCX-mini failure on new build
 
Tim,

An alternate testing method could be to try the processor from the 20meter QCX Mini in the new 40meter QCX.  If the 40meter QCX works with the 20meter processor you know there is something else wrong.  If not, and the 40meter works when you plug in the new processor, most likely the old processor is bad.

This is the second QCX that has had this type of problem that I have read about on the board.  The first one that I know of was a 40meter QCX+.  The processor was bad.

It is always good to have spare parts for testing ;-)

Just a suggestion.
73
Evan
AC9TU


Alan G4ZFQ
 

I started to look into flashing the firmware but the PDF on how to do it assumes starting with a working QCX.

Tim,

No, all it needs is a working processor.
Howard may be right but there have been a few cases where it seems the firmware may have got corrupted.

73 Alan G4ZFQ