time-nuts@lists.febo.com

Discussion of precise time and frequency measurement

View all threads

HP53131A Measurement Throughput (was: HP53131A RS232 question)

SS
Scott Stobbe
Sat, Dec 3, 2016 10:02 PM

Attached are a few plots of bits right on the RS232 line, for various gate
times.

First plot, "HP53131A_BAUD.png" serial transmit duration ~10.4 ms for 20
characters, plus one start and one stop bit yields a baud rate of 19.2
kBaud, as expected.

The next plot, "HP53131A_1sTrigDelay_1sGate.png" is just as wonderful as
the many NEMA serial debates, the mean period is ~1150 ms, with +-50 ms of
jitter.

The next plot,  "HP53131A_100msGate.png" shows the measurement rate is no
where near 10 Sa/s, but is approximately 4 Sa/s. The serial transmit
duration is only ~10ms so its not being bandwidth limited on the UART.
Again roughly +- 50 ms of jitter.

Finally plot, "HP53131A_10msGate.png" the measurement rate is only 6.25
Sa/s for a 10 ms gate time. The jitters is slightly better at +- 25 ms.

For time interval measurements the 53131 keeps up to about 17 events/s
before it starts dropping samples.

When I pick up a long enough GPIB cable I will see how much better the
results are over GPIB versus RS232.

Attached are a few plots of bits right on the RS232 line, for various gate times. First plot, "HP53131A_BAUD.png" serial transmit duration ~10.4 ms for 20 characters, plus one start and one stop bit yields a baud rate of 19.2 kBaud, as expected. The next plot, "HP53131A_1sTrigDelay_1sGate.png" is just as wonderful as the many NEMA serial debates, the mean period is ~1150 ms, with +-50 ms of jitter. The next plot, "HP53131A_100msGate.png" shows the measurement rate is no where near 10 Sa/s, but is approximately 4 Sa/s. The serial transmit duration is only ~10ms so its not being bandwidth limited on the UART. Again roughly +- 50 ms of jitter. Finally plot, "HP53131A_10msGate.png" the measurement rate is only 6.25 Sa/s for a 10 ms gate time. The jitters is slightly better at +- 25 ms. For time interval measurements the 53131 keeps up to about 17 events/s before it starts dropping samples. When I pick up a long enough GPIB cable I will see how much better the results are over GPIB versus RS232.