Sudden failure in early life


Steve in Okinawa
 

Just finished building QCX40 Rev.3 (I never got a serial number), and it
started up and seemed to behave correctly. I got sidelined when the BPF
could not be peaked so I played with the number of turns, re-wound the
whole toroid twice (frustrated at not having enough 0.3mm magnet wire
and not finding any at our poor excuse of a local electronics shop, so I
used slightly larger wire for the 3 small windings). Third time around I
still got the same results, just a steady line of rectangles at level 7.
I had lifted a pad and a trace so jumpered them and rechecked
continuity, and was planning to try it again with a 50-ohm resistor in
the BNC.  I was also trying to figure out why I saw zero voltage at the
junction of R1 and R2, as related to not being able to peak the BPF.

By the way, is the hole numbering on the schematic on p.52 correct? The
connections from the secondaries to IC4 pins 7 and 9 are from toroid
holes 6 and 7, not 5 and 8 as shown.

Then I went to grab some lunch and when I came back there was nothing on
the LCD. The contrast adjusts but no data from the microprocessor. Most
discouraging. I have the usual test instruments but no idea where to
start looking for the problem.

Any help is very welcome.

Steve JS6TMW


Alan G4ZFQ
 

  I was also trying to figure out why I saw zero voltage at the
junction of R1 and R2, as related to not being able to peak the BPF.
Steve,

The RX will not work unless this is correct.

By the way, is the hole numbering on the schematic on p.52 correct?
No-one has commented about this before.

there was nothing on
the LCD. The contrast adjusts but no data from the microprocessor. Most
discouraging. I have the usual test instruments but no idea where to
start looking for the problem.
A start for troubleshooting http://qrp-labs.com/qcx/qcxtrouble.html

73 Alan G4ZFQ


Steve in Okinawa
 

I've removed the input toroid completely. I think I had two wires
reversed, so that would explain some part of the problem. And I have the
correct voltage at the R1-R2 divider, so I probably had grounded it.

BUT - still no data on the LCD. I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
having lunch????

JS6TMW

On 3/16/2018 7:59 PM, Alan G4ZFQ wrote:
  I was also trying to figure out why I saw zero voltage at the
junction of R1 and R2, as related to not being able to peak the BPF.
Steve,

The RX will not work unless this is correct.

By the way, is the hole numbering on the schematic on p.52 correct?
No-one has commented about this before.

there was nothing on
the LCD. The contrast adjusts but no data from the microprocessor. Most
discouraging. I have the usual test instruments but no idea where to
start looking for the problem.
A start for troubleshooting http://qrp-labs.com/qcx/qcxtrouble.html

73 Alan G4ZFQ



Alan G4ZFQ
 

correct voltage at the R1-R2 divider, so I probably had grounded it.
Steve,

Yes, maybe, just keep an eye on it as you progress.

BUT - still no data on the LCD.
Nothing, or blocks, one row two rows??

I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
http://qrp-labs.com/qcx/qcxtrouble.html

Probably unlikely, bit just about possible. More likely joints, shorts.
Do all the checks in the LCD section show no errors?

73 Alan G4ZFQ


jjpurdum
 

Actually, if you have a spare Nano lying around, you would wire the LCD display to a breadboard and compile/load one of the sample programs from the LiquidCrystal library and see if the display works.

Jack, W8TEE


On Friday, March 16, 2018, 11:06:47 AM EDT, Alan G4ZFQ <alan4alan@...> wrote:



> correct voltage at the R1-R2 divider, so I probably had grounded it.
>

Steve,

Yes, maybe, just keep an eye on it as you progress.

> BUT - still no data on the LCD.

Nothing, or blocks, one row two rows??

I checked all the connections. Contrast
> voltage range is correct. The 20 MHz oscillator is going strong. Is it
> just possible that the microprocessor gave up the ghost while I was
  http://qrp-labs.com/qcx/qcxtrouble.html

Probably unlikely, bit just about possible. More likely joints, shorts.
Do all the checks in the LCD section show no errors?

73 Alan G4ZFQ




Steve in Okinawa
 

Thanks Alan,

I'm not saying there could be no bad joints or shorts but I checked carefully as I went along, and it did work right for a few hours.

Nothing, or blocks, one row two rows??
Nothing or two rows of blocks depending on the pot position.

I went down the whole list of voltage checks. Lots of problems, so maybe you can figure out where I should be looking. I'll only list the significant voltage discrepancies:

Clk0 and Clk1 test pins - no squarewave, no DC

Top side of C43 0.95v

Top sides of C44, C45, C46 all 1.31v

IC5 pin 1, pin 2  1.31v,  pin3  0.86v,  pin5  0.87v,  pins 6,7  1.32v

IC6 pins 1,2,3,5,6,7 all 1.37v

IC7 pins 1,2,3,5,6,7 all 1.37v

IC8 pins 1,2,3,5,6,7 all 1.38-1.40v

IC9 pin 3  5.9v,  pin 5  1.38v, pins 6,7  1.41v

IC10 pins 5,6  5.9v

LCD pin4 1.21v, pins 7,8,9,10,11  5v,  pins 12, 13 0v.

I hope this provides plenty of clues.

73,

Steve


On 3/17/2018 12:06 AM, Alan G4ZFQ wrote:
correct voltage at the R1-R2 divider, so I probably had grounded it.

Steve,

Yes, maybe, just keep an eye on it as you progress.

I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
 http://qrp-labs.com/qcx/qcxtrouble.html

Probably unlikely, bit just about possible. More likely joints, shorts.
Do all the checks in the LCD section show no errors?

73 Alan G4ZFQ


V Zecchinelli
 

I had a similar issue with no data on the LCD.  Replacing the ATMega solved the problem.

73 de N1VIN (Vince)

On 3/16/2018 8:30 AM, Steve in Okinawa wrote:
I've removed the input toroid completely. I think I had two wires
reversed, so that would explain some part of the problem. And I have the
correct voltage at the R1-R2 divider, so I probably had grounded it.

BUT - still no data on the LCD. I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
having lunch????

JS6TMW


On 3/16/2018 7:59 PM, Alan G4ZFQ wrote:
  I was also trying to figure out why I saw zero voltage at the
junction of R1 and R2, as related to not being able to peak the BPF.
Steve,

The RX will not work unless this is correct.

By the way, is the hole numbering on the schematic on p.52 correct?
No-one has commented about this before.

there was nothing on
the LCD. The contrast adjusts but no data from the microprocessor. Most
discouraging. I have the usual test instruments but no idea where to
start looking for the problem.
A start for troubleshooting http://qrp-labs.com/qcx/qcxtrouble.html

73 Alan G4ZFQ



Alan G4ZFQ
 

Nothing or two rows of blocks depending on the pot position.
If you read the troubleshooter it gives you instructions on what to do next.
What firmware?

I went down the whole list of voltage checks.
As you are concerned with the micro these are not required at the moment. Although micro voltages might be.

73 Alan G4ZFQ

Lots_ of problems, so
maybe you can figure out where I should be looking. I'll only list the significant voltage discrepancies:
Clk0 and Clk1 test pins - no squarewave, no DC
Top side of C43 0.95v
Top sides of C44, C45, C46 all 1.31v
_IC5 pin 1, pin 2_  1.31v, _pin3_  0.86v, _pin5 _0.87v, _pins 6,7_  1.32v
_IC6 pins 1,2,3,5,6,7_ all 1.37v
_IC7 pins 1,2,3,5,6,7_ all 1.37v
_IC8 pins 1,2,3,5,6,7_ all 1.38-1.40v
_IC9 pin 3_  5.9v, _pin 5_  1.38v, _pins 6,7_  1.41v
_IC10 pins 5,6_  5.9v
_LCD pin4 1.21v_, _pins 7,8,9,10,11_  5v_,_ _pins 12, __13_ 0v.
I hope this provides plenty of clues.
73,
Steve
On 3/17/2018 12:06 AM, Alan G4ZFQ wrote:
correct voltage at the R1-R2 divider, so I probably had grounded it.

Steve,

Yes, maybe, just keep an eye on it as you progress.
I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
http://qrp-labs.com/qcx/qcxtrouble.html

Probably unlikely, bit just about possible. More likely joints, shorts.
Do all the checks in the LCD section show no errors?

73 Alan G4ZFQ


Glen Leinweber
 

Steve,
Among a few other suspicious voltages:
IC9 pin 3: 5.9V !!
This looks bad. Your 5v regulator IC11 should be putting out 5V, not 5.9V
These regulators can only supply voltage, and cannot deal with
something that tries to pull them higher.
Perhaps a current path somewhere between +12v and the 5V rail?
Re-check that the regulator is actually providing +5v at its output.


Steve in Okinawa
 

Glen,

I must have written the numbers down wrong. Checked it again and IC9 pin 3 is 4.5 volts, still wrong but it's not a regulator problem (regulator is putting out 5.1v)


On 3/17/2018 3:42 AM, Glen Leinweber wrote:
Steve,
Among a few other suspicious voltages:
IC9 pin 3: 5.9V !!
This looks bad. Your 5v regulator IC11 should be putting out 5V, not 5.9V
These regulators can only supply voltage, and cannot deal with
something that tries to pull them higher.
Perhaps a current path somewhere between +12v and the 5V rail?
Re-check that the regulator is actually providing +5v at its output.


Steve in Okinawa
 

Good to know I wouldn't be the first lucky one!

73 JS6TMW

On 3/17/2018 2:47 AM, V Zecchinelli wrote:
I had a similar issue with no data on the LCD.  Replacing the ATMega
solved the problem.

73 de N1VIN (Vince)


On 3/16/2018 8:30 AM, Steve in Okinawa wrote:
I've removed the input toroid completely. I think I had two wires
reversed, so that would explain some part of the problem. And I have the
correct voltage at the R1-R2 divider, so I probably had grounded it.

BUT - still no data on the LCD. I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
having lunch????

JS6TMW


On 3/16/2018 7:59 PM, Alan G4ZFQ wrote:
   I was also trying to figure out why I saw zero voltage at the
junction of R1 and R2, as related to not being able to peak the BPF.
Steve,

The RX will not work unless this is correct.

By the way, is the hole numbering on the schematic on p.52 correct?
No-one has commented about this before.

there was nothing on
the LCD. The contrast adjusts but no data from the microprocessor.
Most
discouraging. I have the usual test instruments but no idea where to
start looking for the problem.
A start for troubleshooting http://qrp-labs.com/qcx/qcxtrouble.html

73 Alan G4ZFQ






Steve in Okinawa
 

Hello again Alan,

I have followed the troubleshooter now. Looked for opens and solder
bridges on both sides of the board. The Si5351a soldering looks good.
The problem isn't with the 5 and 12 volt rails, and I also checked pins
7 and 20 of the microprocessor to make sure they were getting 5 volts.

I am not (yet!) familiar enough with the QCX circuitry to know whether I
have a faulty Atmega or if all the discrepant voltages are due to
something else.  I have the latest board Rev 3, so I assume it has the
latest firmware as well.

Any other hints before I ask Hans to have a new Atmega sent out?

73,

Steve JS6TMW

On 3/17/2018 3:24 AM, Alan G4ZFQ wrote:

Nothing or two rows of blocks depending on the pot position.
If you read the troubleshooter it gives you instructions on what to do
next.
What firmware?

I went down the whole list of voltage checks.
As you are concerned with the micro these are not required at the
moment. Although micro voltages might be.

73 Alan G4ZFQ

Lots_ of problems, so
maybe you can figure out where I should be looking. I'll only list
the significant voltage discrepancies:

Clk0 and Clk1 test pins - no squarewave, no DC

Top side of C43 0.95v

Top sides of C44, C45, C46 all 1.31v

_IC5 pin 1, pin 2_  1.31v, _pin3_  0.86v, _pin5 _0.87v, _pins 6,7_ 
1.32v

_IC6 pins 1,2,3,5,6,7_ all 1.37v

_IC7 pins 1,2,3,5,6,7_ all 1.37v

_IC8 pins 1,2,3,5,6,7_ all 1.38-1.40v

_IC9 pin 3_  5.9v, _pin 5_  1.38v, _pins 6,7_  1.41v

_IC10 pins 5,6_  5.9v

_LCD pin4 1.21v_, _pins 7,8,9,10,11_  5v_,_ _pins 12, __13_ 0v.

I hope this provides plenty of clues.

73,

Steve


On 3/17/2018 12:06 AM, Alan G4ZFQ wrote:
correct voltage at the R1-R2 divider, so I probably had grounded it.

Steve,

Yes, maybe, just keep an eye on it as you progress.
I checked all the connections. Contrast
voltage range is correct. The 20 MHz oscillator is going strong. Is it
just possible that the microprocessor gave up the ghost while I was
http://qrp-labs.com/qcx/qcxtrouble.html

Probably unlikely, bit just about possible. More likely joints, shorts.
Do all the checks in the LCD section show no errors?

73 Alan G4ZFQ



Alan G4ZFQ
 

I am not (yet!) familiar enough with the QCX circuitry to know whether I
have a faulty Atmega or if all the discrepant voltages are due to
something else.  I have the latest board Rev 3, so I assume it has the
latest firmware as well.
Steve

I seem to be RTFM for you:-)

Hans' comment about the firmware is confusing.
The latest is said to be 1.00e (is it on a label on the micro?) which is also said to have the bug.
You need Hans to confirm.

73 Alan G4ZFQ

If you read the troubleshooter it gives you instructions on what to do
next.
What firmware?

I went down the whole list of voltage checks.
As you are concerned with the micro these are not required at the
moment. Although micro voltages might be.


Steve in Okinawa
 

Alan - Your remark raises the question of which is the most up-to-date,
the Rev 1.10 assembly manual or the Troubleshooting Guide. I understood
from both that the processor startup or freezing bug was fixed either by
firmware or some parts changes. My Atmega says 1.00 e, and, obviously I
would not be asking all these dumb questions if my radio had kept working.

Would I learn anything useful if I pulled the Atmega and checked all the
voltages I measured again? I also have a good 'scope and can check
waveforms, but I don't see anything about this in the manual or
troubleshooting guide that is relevant until the unit is somewhat
functional.

73,

Steve

On 3/17/2018 3:57 PM, Alan G4ZFQ wrote:
I am not (yet!) familiar enough with the QCX circuitry to know whether I
have a faulty Atmega or if all the discrepant voltages are due to
something else.  I have the latest board Rev 3, so I assume it has the
latest firmware as well.
Steve

I seem to be RTFM for you:-)

Hans' comment about the firmware is confusing.
The latest is said to be 1.00e (is it on a label on the micro?) which
is also said to have the bug.
You need Hans to confirm.

73 Alan G4ZFQ

If you read the troubleshooter it gives you instructions on what to do
next.
What firmware?

I went down the whole list of voltage checks.
As you are concerned with the micro these are not required at the
moment. Although micro voltages might be.




Alan G4ZFQ
 

Your remark raises the question of which is the most up-to-date,
the Rev 1.10 assembly manual or the Troubleshooting Guide.
Steve,

Assembly 1
.
1
0
2
6
-
J
a
n
-
2
0
1
8

Strange Copy/Paste:-)
Troubleshooting, I think completed later but I see no date.
But the firmware question remains.
If the 5V and ground is on the micro, the 20MHz working and all connections to the Si5351 good then it points to the micro. BUT the question of the firmware bug is not answered.

73 Alan G4ZFQ

I understood
from both that the processor startup or freezing bug was fixed either by
firmware or some parts changes. My Atmega says 1.00 e, and, obviously I
would not be asking all these dumb questions if my radio had kept working.
Would I learn anything useful if I pulled the Atmega and checked all the
voltages I measured again? I also have a good 'scope and can check
waveforms, but I don't see anything about this in the manual or
troubleshooting guide that is relevant until the unit is somewhat
functional.
73,
Steve
On 3/17/2018 3:57 PM, Alan G4ZFQ wrote:
I am not (yet!) familiar enough with the QCX circuitry to know whether I
have a faulty Atmega or if all the discrepant voltages are due to
something else.  I have the latest board Rev 3, so I assume it has the
latest firmware as well.
Steve

I seem to be RTFM for you:-)

Hans' comment about the firmware is confusing.
The latest is said to be 1.00e (is it on a label on the micro?) which
is also said to have the bug.
You need Hans to confirm.

73 Alan G4ZFQ


Hans Summers
 

I read the while thread again, I don't think I missed anything. What bug? 

The troubleshooting was written after 1.00e, and after PCB Rev 3 and the lateat manual. But the troubleshooting should apply to everything, all versions (PCB and firmware).

So I'm confused about everything. What comment did I make about what? And what bug? Please enlighten me gents!

73 Hans G0UPL 


On Mar 17, 2018 10:34 AM, "Alan G4ZFQ" <alan4alan@...> wrote:

Your remark raises the question of which is the most up-to-date,
the Rev 1.10 assembly manual or the Troubleshooting Guide.

Steve,

Assembly 1
.
1
0
2
6
-
J
a
n
-
2
0
1
8

Strange Copy/Paste:-)
Troubleshooting, I think completed later but I see no date.
But the firmware question remains.
If the 5V and ground is on the micro, the 20MHz working and all connections to the Si5351 good then it points to the micro. BUT the question of the firmware bug is not answered.

73 Alan G4ZFQ


 I understood
from both that the processor startup or freezing bug was fixed either by
firmware or some parts changes. My Atmega says 1.00 e, and, obviously I
would not be asking all these dumb questions if my radio had kept working.

Would I learn anything useful if I pulled the Atmega and checked all the
voltages I measured again? I also have a good 'scope and can check
waveforms, but I don't see anything about this in the manual or
troubleshooting guide that is relevant until the unit is somewhat
functional.

73,

Steve


On 3/17/2018 3:57 PM, Alan G4ZFQ wrote:
I am not (yet!) familiar enough with the QCX circuitry to know whether I
have a faulty Atmega or if all the discrepant voltages are due to
something else.  I have the latest board Rev 3, so I assume it has the
latest firmware as well.


Steve

I seem to be RTFM for you:-)

Hans' comment about the firmware is confusing.
The latest is said to be 1.00e (is it on a label on the micro?) which
is also said to have the bug.
You need Hans to confirm.

73 Alan G4ZFQ





Alan G4ZFQ
 

I read the while thread again, I don't think I missed anything. What bug?
The troubleshooting was written after 1.00e, and after PCB Rev 3 and the lateat manual. But the troubleshooting should apply to everything, all versions (PCB and firmware).
So I'm confused about everything. What comment did I make about what? And what bug? Please enlighten me gents!
Hans

Troubleshooting, LCD blank

"Check with a DVM that there are no short-circuits between adjacent LCD pins, or between any LCD pin and Gnd or Vcc. Check for continuity between each LCD pin and the corresponding pin on the processor. You can do this with the board upside down, and the LCD module plugged in.

If ALL of that checks out fine, then there is one more possibility for a blank screen, in firmware versions up to and including 1.00e. In these versions the firmware carries out initialization steps which clear the screen. Then it tries to communicate a command to the Si5351A Synthesizer chip IC1. The commands to IC1 are sent over an I2C serial communication link. If for some reason, the Si5351A does not respond to the command to acknowledge it, then the I2C bus will hang up while infinitely waiting for the response of the Si5351A. The screen will remain blank indefinitely and.."

So was this altered? can a bad Si5351 or connections to it still cause the blank display?

73 Alan G4ZFQ


On Mar 17, 2018 10:34 AM, "Alan G4ZFQ" <alan4alan@gmail.com <mailto:alan4alan@gmail.com>> wrote:
Your remark raises the question of which is the most up-to-date,
the Rev 1.10 assembly manual or the Troubleshooting Guide.
Steve,
Assembly 1
.
1
0
2
6
-
J
a
n
-
2
0
1
8
Strange Copy/Paste:-)
Troubleshooting, I think completed later but I see no date.
But the firmware question remains.
If the 5V and ground is on the micro, the 20MHz working and all
connections to the Si5351 good then it points to the micro. BUT the
question of the firmware bug is not answered.
73 Alan G4ZFQ
 I understood
from both that the processor startup or freezing bug was fixed
either by
firmware or some parts changes. My Atmega says 1.00 e, and,
obviously I
would not be asking all these dumb questions if my radio had
kept working.


Hans Summers
 

Hi Alan

That paragraph In the troubleshooting is completely consistent and correct. It says firmware up to and including 1.00e. And 1.00e is the latest version. So everything is consistent and there is no ambiguity. And this behaviour is still present, yes... 

Neither is it a bug. It is just the order of how things work in the firmware currently. Architecturally this is the way the firmware is written. If everything is working properly, then there is never any issue. But if the Si5351A has a fault, then the screen stays blank, whereas it would really be more helpful if it wrote something about the problem on the screen. It would be a useful improvement to the firmware.

I did take a look at it last week and I will figure it out. If it was trivial I would have done it already. It isn't just a matter of swapping the ordering of two initialization function calls. There are detailed reasons why the initialization at power-up has to do things in a specific order. Changing that is possible but it's a little complex.

73 Hans G0UPL 


On Mar 17, 2018 11:06 AM, "Alan G4ZFQ" <alan4alan@...> wrote:
I read the while thread again, I don't think I missed anything. What bug?
The troubleshooting was written after 1.00e, and after PCB Rev 3 and the lateat manual. But the troubleshooting should apply to everything, all versions (PCB and firmware).

So I'm confused about everything. What comment did I make about what? And what bug? Please enlighten me gents!


 Hans

Troubleshooting, LCD blank

"Check with a DVM that there are no short-circuits between adjacent LCD pins, or between any LCD pin and Gnd or Vcc. Check for continuity between each LCD pin and the corresponding pin on the processor. You can do this with the board upside down, and the LCD module plugged in.

If ALL of that checks out fine, then there is one more possibility for a blank screen, in firmware versions up to and including 1.00e. In these versions the firmware carries out initialization steps which clear the screen. Then it tries to communicate a command to the Si5351A Synthesizer chip IC1. The commands to IC1 are sent over an I2C serial communication link. If for some reason, the Si5351A does not respond to the command to acknowledge it, then the I2C bus will hang up while infinitely waiting for the response of the Si5351A. The screen will remain blank indefinitely and.."

So was this altered? can a bad Si5351 or connections to it still cause the blank display?

73 Alan G4ZFQ



On Mar 17, 2018 10:34 AM, "Alan G4ZFQ" <alan4alan@... <mailto:alan4alan@...>> wrote:


        Your remark raises the question of which is the most up-to-date,
        the Rev 1.10 assembly manual or the Troubleshooting Guide.


    Steve,

    Assembly 1
    .
    1
    0
    2
    6
    -
    J
    a
    n
    -
    2
    0
    1
    8

    Strange Copy/Paste:-)
    Troubleshooting, I think completed later but I see no date.
    But the firmware question remains.
    If the 5V and ground is on the micro, the 20MHz working and all
    connections to the Si5351 good then it points to the micro. BUT the
    question of the firmware bug is not answered.

    73 Alan G4ZFQ


      I understood

        from both that the processor startup or freezing bug was fixed
        either by
        firmware or some parts changes. My Atmega says 1.00 e, and,
        obviously I
        would not be asking all these dumb questions if my radio had
        kept working.






Alan G4ZFQ
 

That paragraph In the troubleshooting is completely consistent and correct. It says firmware up to and including 1.00e. And 1.00e is the latest version. So everything is consistent and there is no ambiguity. And this behaviour is still present, yes...
OK Hans,

Sorry for using the B.. word.
Of course am I missing something? If the micro works there will be the initial boot-up then it goes blank if there is a Si5351 problem?

I do not think Steve has seen this.

73 Alan G4ZFQ


Neither is it a bug. It is just the order of how things work in the firmware currently. Architecturally this is the way the firmware is written. If everything is working properly, then there is never any issue. But if the Si5351A has a fault, then the screen stays blank, whereas it would really be more helpful if it wrote something about the problem on the screen. It would be a useful improvement to the firmware.
I did take a look at it last week and I will figure it out. If it was trivial I would have done it already. It isn't just a matter of swapping the ordering of two initialization function calls. There are detailed reasons why the initialization at power-up has to do things in a specific order. Changing that is possible but it's a little complex.
73 Hans G0UPL
http://qrp-labs.com
On Mar 17, 2018 11:06 AM, "Alan G4ZFQ" <alan4alan@gmail.com <mailto:alan4alan@gmail.com>> wrote:
I read the while thread again, I don't think I missed anything.
What bug?
The troubleshooting was written after 1.00e, and after PCB Rev 3
and the lateat manual. But the troubleshooting should apply to
everything, all versions (PCB and firmware).
So I'm confused about everything. What comment did I make about
what? And what bug? Please enlighten me gents!
 Hans
Troubleshooting, LCD blank
"Check with a DVM that there are no short-circuits between adjacent
LCD pins, or between any LCD pin and Gnd or Vcc. Check for
continuity between each LCD pin and the corresponding pin on the
processor. You can do this with the board upside down, and the LCD
module plugged in.
If ALL of that checks out fine, then there is one more possibility
for a blank screen, in firmware versions up to and including 1.00e.
In these versions the firmware carries out initialization steps
which clear the screen. Then it tries to communicate a command to
the Si5351A Synthesizer chip IC1. The commands to IC1 are sent over
an I2C serial communication link. If for some reason, the Si5351A
does not respond to the command to acknowledge it, then the I2C bus
will hang up while infinitely waiting for the response of the
Si5351A. The screen will remain blank indefinitely and.."
So was this altered? can a bad Si5351 or connections to it still
cause the blank display?
73 Alan G4ZFQ


Hans Summers
 

Hi Alan

I will clarify that in the troubleshooting. 

If there is an Si5351A fault, on boot up no text is shown on the display. This is because in the initialization steps, talking to the Si5351A happens BEFORE talking to the LCD. 

The specific reason it was done this way, is to make sure the Si5351A Clk2 output is off... So that the BS170 PA gates will definitely be low, regardless of the lack of definitive Rx/Tx control signal state until the rest of the boot-up is completed.

Note that the specific Si5351A fault which causes this, is a failure to communicate with the processor via the I2C bus. Other Si5351A faults could occur (such as problems with the output signals) but NOT cause the boot-up sequence to halt... because as long as the Si5351A doesn't refuse to communicate, the processor is happy.

It would be nicer to initialize the LCD screen first and write "Si5351A comms error" for example, To help provide diagnostic information to the hapless constructor. 

Steve's problem could therefore logically be due to the Si5351A but I think it is a less likely cause. More likely, some short, or spike (on power supply or SWR perhaps), could have fatally annoyed the processor.

73 Hans G0UPL 


On Mar 17, 2018 11:29 AM, "Alan G4ZFQ" <alan4alan@...> wrote:

That paragraph In the troubleshooting is completely consistent and correct. It says firmware up to and including 1.00e. And 1.00e is the latest version. So everything is consistent and there is no ambiguity. And this behaviour is still present, yes...


OK Hans,

Sorry for using the B.. word.
Of course am I missing something? If the micro works there will be the initial boot-up then it goes blank if there is a Si5351 problem?

I do not think Steve has seen this.

73 Alan G4ZFQ


Neither is it a bug. It is just the order of how things work in the firmware currently. Architecturally this is the way the firmware is written. If everything is working properly, then there is never any issue. But if the Si5351A has a fault, then the screen stays blank, whereas it would really be more helpful if it wrote something about the problem on the screen. It would be a useful improvement to the firmware.

I did take a look at it last week and I will figure it out. If it was trivial I would have done it already. It isn't just a matter of swapping the ordering of two initialization function calls. There are detailed reasons why the initialization at power-up has to do things in a specific order. Changing that is possible but it's a little complex.

73 Hans G0UPL
http://qrp-labs.com


On Mar 17, 2018 11:06 AM, "Alan G4ZFQ" <alan4alan@... <mailto:alan4alan@...>> wrote:

        I read the while thread again, I don't think I missed anything.
        What bug?
        The troubleshooting was written after 1.00e, and after PCB Rev 3
        and the lateat manual. But the troubleshooting should apply to
        everything, all versions (PCB and firmware).

        So I'm confused about everything. What comment did I make about
        what? And what bug? Please enlighten me gents!


      Hans

    Troubleshooting, LCD blank

    "Check with a DVM that there are no short-circuits between adjacent
    LCD pins, or between any LCD pin and Gnd or Vcc. Check for
    continuity between each LCD pin and the corresponding pin on the
    processor. You can do this with the board upside down, and the LCD
    module plugged in.

    If ALL of that checks out fine, then there is one more possibility
    for a blank screen, in firmware versions up to and including 1.00e.
    In these versions the firmware carries out initialization steps
    which clear the screen. Then it tries to communicate a command to
    the Si5351A Synthesizer chip IC1. The commands to IC1 are sent over
    an I2C serial communication link. If for some reason, the Si5351A
    does not respond to the command to acknowledge it, then the I2C bus
    will hang up while infinitely waiting for the response of the
    Si5351A. The screen will remain blank indefinitely and.."

    So was this altered? can a bad Si5351 or connections to it still
    cause the blank display?

    73 Alan G4ZFQ