¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Topics

OCXO oscillator works fine but not recognized by my U3S

 

I have a number of ultimate 3s WSPR beacons up and running. I recently had an overvoltage issue with one of them which has damaged the OCXO, most likely the oven transistors as far as I can see. This unit is still recognized by the software (ver 3.11. However I decided to start afresh. I've finished the build of a new OCXO but not fully enclosed it yet. The oscillator tests out fine and is ok for frequency and level. However the OCXO is not recognized by the software. What is the software looking for as a valid unit? This may help me in further fault finding. I have a 100meh scope and 3gig spectrum analyser for fault finding.?
Dave ZL1TIU
 

Hi

By not recognized do you mean you get an error message ?

73 Jan
 

The MCU talks to the synth via I2C and expects answers in return, if this does not happen it will show an error. As the data lines goes through a level converter on the synth board, these too needs to be working properly for the communication to work.
Scope the SDA/SCL lines on both the MCU and synth side and see if it gets through. If there's 5V clock/data on one side and 3.3V on the other, the synth could be bad. Otherwise there could be a problem with the level converters. If unlucky and there's 5V on the synth side it could be toasted.

On Wed, Aug 23, 2017 at 9:37 AM, David Parker <dwparker@...> wrote:
I have a number of ultimate 3s WSPR beacons up and running. I recently had an overvoltage issue with one of them which has damaged the OCXO, most likely the oven transistors as far as I can see. This unit is still recognized by the software (ver 3.11. However I decided to start afresh. I've finished the build of a new OCXO but not fully enclosed it yet. The oscillator tests out fine and is ok for frequency and level. However the OCXO is not recognized by the software. What is the software looking for as a valid unit? This may help me in further fault finding. I have a 100meh scope and 3gig spectrum analyser for fault finding.?
Dave ZL1TIU


 

This is a quote from another thread (don't know how to post a link to a special message in a thread).

Hans Summers
Jan 8???

Hi Ryan
?
Apologies for my late arrival in this thread.
?
There are two possible errors when the processor starts up and tests the Si5351A module.?
?
The first error is "Si5351 Error" and it means that the processor failed to communicate with the Si5351A on the I2C bus.?
?
The second error is "No Clk" and if you see that, it means that the processor correctly communicated with the Si5351A but it cannot measure any output. So you have no issues with the two I2C serial lines (data and clock). The Clk2 output of the Si5351A should be connected to the processors T1 (TimerCounter1) input. This is how the processor measures the Si5351A output.?
?
So it is a good idea to trace through the Clk2 signal and find out?
a) is it properly present at the Si5351A pin itself and if so,
b) where it fails to reach through to the processor T1 pin.
An oscilloscope would really help but I know you don't have one.
?
In the event there is some failed component, which does happen sometimes (though not often) I always get a replacement sent out for you.?
?
73 Hans G0UPL?
?
 

Ok, thanks for all the helpful comments. There are 5v signals from the mcu going in to the level converters and 3v signals coming out. The 3v signals appear on pins 4 and 5 of the synth. What is the next step in my testing? Oh the error message was 'error:Si5351A' Thanks
 

¿ªÔÆÌåÓý

Hi David,
RE:- 'error:Si5351A'

When Hans did the earlier firmware that supported both the AD9850 and Si5351 the fact that it had successfully identified the synth as a Si5351 meant successful comms via the I2Cbus.
(I don't know if that is still true or if now with the latest firmware is for the Si chip it is just a default response, Only Hans will know the inner workings of the current firmware.)

Back to basic stuff I can suggest.
5V in the correct places.
Voltages correct inc 3.3v +/- say 10%
27MHz Osc running
Don't know if the Si is sufficiently initialised to provide any outputs, but can you see:-
any O/P on CLK0 (Module pin 17)
any O/P on CLK2 (Module pin 19) should be 6.75MHz
(Which I BELIEVE is the final step in the firmware to confirm initialisation of the Si. and is also the calibration reference signal used later?)

Your unit is a recent order u3S with the CQU Ver3 PCB, not an kit that has being unbuilt for a time.
OR a repurposed clock or or U3 kit that was originally supplied with the AD9850 module, which requires modification to use the Si5351.

regards Alan


On 28/08/2017 09:55, David Parker wrote:

Ok, thanks for all the helpful comments. There are 5v signals from the mcu going in to the level converters and 3v signals coming out. The 3v signals appear on pins 4 and 5 of the synth. What is the next step in my testing? Oh the error message was 'error:Si5351A' Thanks

 

Hi David, I have had this issue with one OCXO I was building and before I sealed the last box, I discovered the the SI5351 ERROR:? I assume that I inserted the pins in the holes wrong when I was testing it as it did work though the building process. I determined that I had destroyed the SI5351 (shorted) chip. I ordered new chips from Mouser and I replaced the chip. It fixed the issue. I wasn't sure I could accomplish the soldering feat but it all turned out for the good. Hope you Si5351 is not damaged. The tip off that the Si5351 is damaged is that the regulator IC gets very warm.? Bill KN8DMK.. .

1 - 7 of 7