34465A Sampling Rate.

[tsj6181]

Hi,

I have a issues regarding the 34465A standard equipment sampling rate.

I need to continuous measure the four wire resistance using 1.25ms for 5 minutes using GPIB control by scpi commands.

However, I have tryout using the C# for 2 minutes continuous measure the resistance by using 1.25ms, it is able to get 50300++ sampling rates only.

2minutes is 120 seconds divided by 0.00125 which is 96000. The array in C# should return me 96000 variables but however it only have 50300++ sampling rates only.

For 1 minutes, it is working normal, getting 48000 points. But for more than 2  minutes, it is not able to get the actual sampling rates.

The setting that I set for 34465A DMM is NPLC 0.02, source timer, sample count 96000, sample timer 0.00125, four wire resistance 1000 ohms, trigger delay 0.05, trigger count 1 and trigger source IMM.

So I would like to ask is anyone encounter this issues and how to solve it ? or this is the 34465A DMM limitation?

[bson]

Isn't NPLC 0.02 a max 3000kHz sample rate at 60Hz, or 2500Hz at 50Hz?  Whereas the 1.25ms trigger time is a 8000Hz rate.  It seems like you need to use a smaller sample duration.

You can also try capturing to a USB storage device (or internal memory); if that works as expected you know the problem isn't with the sampling itself.

[Dr. Frank]

Hello,
I could not fully understand, how you have set up your 34465A.
There are several possibilities, how to shuffle the data, w/ or w/o buffering, LIFO, FIFO, how is the triggering done, which data format (ASCII, binary), DISP OFF, etc.
Are you using OFFSET COMP also?
Also, which FW version is used, hopefully 3.0 or higher? This enables the DIG option, maybe it's needed for your setting, as 4W Ohm requires two measurements.
Do you have the MEM option installed? It was available for free sometimes, and probably required for such high data rates, or data amount of your 240,000 samples.
 
The DMM has several speed limitations concerning bus, and data processing, see specification and manual (very badly described, and much slower than the predecessors 34410/411)
What I have seen, that any fast data transfers use the memory, i.e. if you don't have the MEM option (2M samples), the buffer will fill to 50k samples, and then the transfer slows down.
I strongly assume, that this is your limitation you see beyond these 48000 samples.

The state machine of the DMM (virtual one) is not able to make fast direct data transfers, w/o using the buffer. It's limited to 250/sec, if I remember correctly.
Using GPIB is slower than LAN transfer.

Also, I don't know if reading into an array works correctly with these SCPI commands.

Maybe you check first if sampling into memory is fully possible internally, w/o computer control.

When I did sampling with my 3458A over GPIB and with DMA, 30 years ago, there was a 64kB limitation for array transfer either in the GPIB card hardware / software, or in the DMA access, or in the AT PC architecture (64kB sectors), I don't remember exactly any more.
Maybe this is another limit incorporated with your GPIB interface.

Frank

[LaurentR]

You may want to look at the User Manual, search for "Resistance (4-wire) Default Delays".
You'll see that 4W measurements are slower and based on the range.

Note that there are measurement options (like Auto Zero, although that specific one is not available in 4W mode) that also slow down the sampling rate.

[tsj6181]

Sorry for the NPLC typo, the value I set is 0.002 instead of 0.02.

But I think 0.02 is enough also. 1 second for 0.00125 is 800 sample rates which NPLC 0.02 a max 3kHz sample rate at 60Hz. It still below 3000 sample rates data.

Thanks.

[tsj6181]

Yes, you are correct. This is the 34465A DMM limitation.

I have go through the data sheet. My DMM is standard option which maximum getting is 50K sample rates data and already full of memory inside.

Only with MEM options, it is able to get 2M sample rates data.