[time-nuts] TAC/TDC

[Bruce Griffiths]

Josef Kölbl wrote:
> I am looking for a commercial drop-in module/solution of a 
> time-to-analog or even time-to-digital converter module with <= 1 ps 
> resolution (device should eventually go on a printed-circuit board). 
> Critical issues are: linearity, stability, accuracy, jitter and 
> precision. The maximum time interval to be measured shall be <= 25 ns. 
> Do you know of any commercial products? I will require more than 1 unit.
> Thanks.
> Josef
The only single chip TAC/TDC with 1ps resolution and noise that I've 
seen any publications on used a custom implementation of a TAC  that 
used a state of the art IC process.
These were intended for particle physics use.
I'm not sure if they moved beyond the prototype stage.
Moving the entire TAC to a single chip allows a very small ramp 
capacitor to be used as well as reducing interconnection parasitics etc.

Without using a custom IC implementation of a TAC, one can either use a 
similar approach to that used in the Wavecrest DTS2075 with its 
attendant high power dissipation or perhaps use a triggered ringing LC 
circuit plus a fast ADC.

The highest single shot resolution TD chip (TDC-GPX) available from ACAM 
is about 10ps.
Suitable averaging techniques can achieve sub ps resolution provide that 
the chip nonlinearities are corrected for.

Bruce