[time-nuts] Measurement and correction of non-linearity of a TDC

[Erik Kaashoek]

Is there someone on this list that has experience with the measurement 
and correction of non-linearity of a TDC?
It appears the TDC's I'm using have at least two types of non linearity.
One seems to be related to the pulling by the clock used to create the 
stop pulse.
The other is a small repeating non-linearity, every 54 steps of the the TDC
The attached graph shows the error in steps per step being measured. 
There is still some noise in the error measurement as measuring the 
linearity error proved to be rather difficult.
The larger oscillation around step 1250 with amplitude of 10 steps could 
be pulling by the clock.  The repeating saw-tooth error has a period of 
54 steps and an amplitude of 4 steps. Each TDC step is about 41 ps wide. 
Total width of the graph is about 100 ns or about 2400 steps.
The same non-linearity appears on all 6 TDC samples that where measured.
Does this make sense?
Erik.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: TDC_error.PNG
Type: image/png
Size: 32254 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20240609/2a42863c/attachment.png>