[time-nuts] RE: Micro phase stepper

buehl at superlink.net buehl at superlink.net
Mon Jul 18 18:21:30 EDT 2005

I just finished a design using programmable delay lines similar to the Maxim 
parts (From Data Delay Devices in New Jersey, USA).  Was disappointed in the 
step to step errors.  Using the 1nS per step device, I measured the delta of 
some steps as short as .4 nS.  Average of many steps was fine.  Also found the 
rise and fall times of output of the DDD ICs to be slow and non-linear.  Did 
not have opportunity to test the Maxim parts.  If anyone has used them, we 
would all like to know your experience.

Cost is another factor; more than US$ 100 is common.

I am watching this thread carefully, and will test out the D/A convertor w/ 
comparitor (and other ideas) for the next generation design.  My design uses 
an integrator/ FPGA input threshold for a fixed 35 nS delay, and it appears to 
be very stable and repeatable;  So a real comparitor version should be great.

Tom Buehl

Quoting Tom Clark <w3iwi at toad.net>:

> Ulrich, DF6JB asked 
>      can someone please explain to me how a micro phase stepper works?
>      Is it
>      a pure digital device or does it make use of analog delay line
>      techniques? What is the maximum delay and what is the delay step
>      size?
> Here I describe two different types of phase steppers:
>     1. If you are trying to precisely delay a pulse train, take a look at
>        Maxim's programmable delay lines at
>        [1]http://www.maxim-ic.com/products/timers/delay_lines.cfm. The
>        DS1020 series is available with 8 bits of  binary programmable
>        steps with resolutions of 0.15, 0.25, 0.5 1 or 2 nsec per step and
>        total delays in the 50-500 nsec range  (see
>        [2]http://www.maxim-ic.com/appnotes.cfm/appnote_number/107 ). This
>        will work to delay an arbitrary pulse train  If you look on
>        [3]GPSTIME for my "Timing for VLBI 2005" version and scroll down
>        to slides 21 and then 25-27, you'll see why commercial GPS
>        receivers have a "sawtooth" jitter on their 1PPS outputs, and how
>        it is being removed in a commercial product.
>     2. If you want to phase delay a sine wave, a simple trick is to use a
>        double-balanced mixer to up-convert from a VFO. Pass the signal
>        thru a long coax delay line. Then use the SAME vfo to down-convert
>        back to the original frequency with another mixer. As you tune the
>        VFO, you change the phase delay at the output.
>    Hope that helped -- 73 de Tom, W3IWI
> References
>    1. http://www.maxim-ic.com/products/timers/delay_lines.cfm
>    2. http://www.maxim-ic.com/appnotes.cfm/appnote_number/107
>    3. http://gpstime.com/
> _______________________________________________
> time-nuts mailing list
> time-nuts at febo.com
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts

More information about the time-nuts mailing list