[time-nuts] Thunderbolt serial port problem UPDATE

bill billj at ieee.org
Thu Apr 14 15:38:58 UTC 2011


On 4/13/2011 5:47 PM, shalimr9 at gmail.com wrote:
> Fluke.l is using my design except for a number of units which were shipped with 0 ohm resistors where the 3 diodes were supposed to go. This powers the CPU with 5 V instead of 3 and most don't last long at that voltage. However, the symptom is a blank display.
>
> Also, I understand some units have an actual 5 V regulator and some expect a regulated 5 V.
>
> I have tested a total of 9 units, 2 from fluke.l one from Dan Karg and the other are mine against 3 Thunderbolts (2 group buys and one of the original red boxes) and they all work. As long as the serial data is enough to bias on and off the 2N2222, it should work.
>
> Check the wiring and check that the serial data goes to the CPU. There is not much that can go wrong if the CPU is working.
>
> It is possible that the serial input to the CPU is fried while the rest of the chip works, but I have never seen that happen. Yet I have fried a number of these chips (I design HV supplies in my day job).
>
> Didier KO4BB
OK I think I will check the signal going to the CPU. If it is there, 
then I will throw the unit away.
If the signal is not at the CPU I will trouble shoot further.

Thanks for the suggestion everybody.

Bill K7NOM


More information about the time-nuts mailing list