[time-nuts] Happy Holidays

[Bruce Griffiths]

Mark Sims wrote:
> Another possible thing to try is one of the Miller integrator chips that Tektronix used in their scopes and time base modules.  I seem to remember that at least one could yield 0.01% linearity.  The 5Bxx and 7Bxx modules for the 5000 and 7000 series scopes are available in vast quantities,  often for less than $10 (plus about that in shipping).  There is also an Ebay seller that has one of the chips in lots of 10...  search for "Tektronix miller".   I don't know what unit those chips were from.  They are in a TO5 type can,  not the usual DIP package.  To use a particular chip  you would need to do some reverse engineering from the appropriate module service manual.  The connections are usually fairly simple.
>   
Mark

As far as I can tell from the 5B40 and 5B42 manuals its just a triggered
ramp generator with no provision for holding the ramp level in response
to an external signal.
Consequently it would only be useful in a TAC (time to amplitude
converter) if a sample and hold were used to sample the ramp at the
appropriate instant.
In principle one could use an ADC with a built in sample and hold that
has a sufficiently low acquisition time and low sampling jitter.
For modern sampling ADCs with no input buffer the RC sampling capacitor
charging time constant ensures that the voltage on the sampling cap is
merely delayed with respect to the input ramp voltage at the sampling
instant (provided that the ramp has been slewing for sufficient time for
the sampling capacitor ramp start transients to settle).

Bruce