Track time / Hold time

[Labalafre]

Hello everybody,

In order to calculate the maximum sample rates of my DAQ cards (6220,
6259) I would to have know the "Hold time" and the "Track time" of the
following formula:

MaxSampleRate = 1/ [HoldTime + (Nc-1)*SettleTime + TrackTime] where Nc
is the number channels being sampled

Can you tell me where I can find these characteristics ? (They are not
in the user manual)

Thank you in advance for your help.

[Mark H]

Labalafre,

I notice from your formula that you may be under the impression that DAQ boards have "sample & hold" circuitry on each input - unfortunately most DAQ boards do not perform true simultaneous sampling.

Most DAQ boards have a maximum sampling rate specification, e.g. the 6220 board's maximimum analogue input sampling rate is 250KS/s. In general, you do not need to calculate the internal track/hold timings of the board yourself, the board will function up to the maximum sampling rate.

If you add your own external "sample and hold" amplifiers, you may well need to calculate custom "channel clock" timing to allow your own S&H circuitry to re-sample/settle between DAQ channel list scans (I've done this myself).

Accuracy is another matter though - if you are sampling multiple channels, then it takes time for the board's amplifier input to settle from the previous voltage level and the next. In order to acheive higher accuracies you have to reduce the scanning rate to give each measurement greater time to settle.

Another trick you could use, if particular signals take longer to settle, is to list a channel twice (or more) in succession in the channel scan list and then only use the last reading for that channel.

The settling time required for a particular accuracy is partly a function of the DAQ board specification, but it also depends on the impedance of your signal source - higher impedance signal sources take longer to settle.

DAQ board specifications are all accessible on NI's web site (like just about everything else) - but to save you a little time, here are links to the specification PDFs of your two boards... 6220 spec's & 6259 spec's.

Hope this helps - if so, please do "Rate this message" - and feel free to ask more questions.

Mark H.

[Labalafre]

Thank you very much Mark. How can I rate your message ?

Labalafre.

Mark H wrote:
> Labalafre,
> I notice from your formula that you may be under the impression that DAQ boards have "sample & hold" circuitry on each input - unfortunately most DAQ boards do not perform true simultaneous sampling.
> Most DAQ boards have a maximum sampling rate specification, e.g. the 6220 board's maximimum analogue input sampling rate is 250KS/s. In general, you do not need to calculate the internal track/hold timings of the board yourself, the board will function up to the maximum sampling rate.
> If you add your own external "sample and hold" amplifiers, you may well need to calculate custom "channel clock" timing to allow your own S&H circuitry to re-sample/settle between DAQ channel list scans (I've done this myself).
> Accuracy is another matter though - if you are sampling multiple channels, then it takes time for the board's amplifier input to settle from the previous voltage level and the next. In order to acheive higher accuracies you have to reduce the scanning rate to give each measurement greater time to settle.
> Another trick you could use, if particular signals take longer to settle, is to list a channel twice (or more) in succession in the channel scan list and then only use the last reading for that channel.
> The settling time required for a particular accuracy is partly a function of the DAQ board specification, but it also depends on the impedance of your signal source - higher impedance signal sources take longer to settle.
> DAQ board specifications are all accessible on NI's web site (like just about everything else) - but to save you a little time, here are links to the specification PDFs of your two boards... <a href="http://www.ni.com/pdf/manuals/371290d.pdf" target="_blank">6220</a>&nbsp;spec's &amp; <a href="http://www.ni.com/pdf/manuals/371291c.pdf" target="_blank">6259</a>&nbsp;spec's.
> Hope this helps&nbsp;- if so, please do "Rate this message" - and feel free to ask more questions.
> Mark H.
> &nbsp;

[Mark H]

---

Labalafre wrote:
Thank you very much Mark. How can I rate your message ?

---

Here's how... http://forums.ni.com/ni/help_faq#howrate (from the "Help" link at the top of the forum/message list).

Rating messages isn't essential, but it is nice to build a good average message rating associated with your profile - and it also helps to highlight those messages in a thread which people have found particularly helpfull (yellow stars appear in the message listings).

You can rate any message you read (except your own).

Not sure why you copied my message back in your reply - but don't worry, I can see that you have only just started to use the forums.

Mark.