Agilent 34970A sampling rate

[tcsaba101]

Based on the forum discussion I have set up my ATE.
here: https://www.eevblog.com/forum/testgear/how-to-build-an-ate/msg479564/#msg479564

Yesterday I have made my first "serious" data logging: a Car battery current measurement in the car from opening the door, starting, charge, etc, .

The setup:
34970A
34901A 20ch multiplexer
E2050A LAN GPIB converter
Hp notebook + crossover cable
Benchlink software for 34970A
300A shunt


Officially the 34901A can work at 60samples/sec.
That gives less then 100ms for 5 channel and less then 20ms for 1 channel.


First I have made 5 channel test programmed to 500ms sampling that gone flawless.
Then programmed 100ms sapling rate, but it has made at 300ms sampling (precisely 287ms).
Finally I made 1 channel sampling when no relay open/close time needed, one channel connected permanently but didn't gone under 42ms the samling rate. (see pictures on two cars)

Originally I wanted to ask why not get the specified rate?
But immediately I have get in my mind why I couldn't go faster: because the specification is for 4,5 digit readings and I had 5,5 digit resolution programmed.

Still the single channel 42 ms is too much.
1 PLC integration is 20ms, a bit more sampling time would be acceptable.
I have programmed a second channel setup in the Benchlink program a calculated channel, that one calculates the current from the one measured shunt voltage, maybe that calculated channels takes the second 20ms?

Is there info existing what is the expected sampling rate for 5,5 digits and 6,5 digit readings with 34901A or 34902A?

[tszaboo]

Csaba, do you have auto zero on? That aproximately doubles the measurement time, as it will make a zero offset measurement before sampling the real signals.

[tcsaba101]

Csaba, do you have auto zero on? That aproximately doubles the measurement time, as it will make a zero offset measurement before sampling the real signals.

That looks to be a good idea.

After reading the regarding users manual section:

Definitely the Autozero is on, because it is the default state, and cannot be disabled from the front panel.
When the integration time is set to less than 1 PLC, it is disabled. This is the only way to control from the front panel.

It has to disabled from the remote interface, and also possible to make a remote commanded "one time autozero cycle".
The Autozero params are stored in a volatile memory, so after every power up it has to be done.

Next time I will try using disabled Autozero!

Thanks.