Can I still get 16 bit resolution with the USB-6210 set to +/-1 volt?

[mhayenga]

The title says it all. I just need to check that, if I change the input voltage range on the USB-2610 multifunction DAQ, will I still get full 16 bit resolution over the +/- 1 volt range? Thank you,

Mark Hayenga

[AnalogKid2DigitalMan]

Yes, and your resolution will be about 31uV (2V/(2^16)).

Here's a link to the manual:

http://www.ni.com/pdf/manuals/371931d.pdf

Happy sampling!

[Neal M]

Hello Mark,

The bits of resolution on your card refers to how many different values your Analog to Digital Converter (ADC) can have.  So a 16 bit ADC has 2^16 (=65536) different values it can output over the given input range.  This means that the smallest change in input voltage that is significant would be equal to your range divided by the number of values your ADC can convert to.

What is actually going on in the guts of the card is that the ADC has a fixed range of -10V to 10V and can detect an ~305µV change (20V/65536).  The range that you specify controls the preamp that scales the data before it reaches the actual ADC.  So for example, our +/-1V signal would be amplified 10x before it reaches the 16 bit ADC and then the value the ADC gives back is scaled back down before we read it.

There are some things we do internally that make the absolute accuracy better, so the calculated resolution will be just slightly better than the actual resolution.  I have copied table 4-1 out of the user manual (AnalogKid2DigitalMan linked above) which shows the actual resolution for the different ranges:

I hope this answers your question satisfactorily Feel free to post back with any further questions.

Message Edited by Neal M on 12-07-2007 10:14 AM

[Mark Hayenga]

Thanks guys!

Question 2: The manual says source impedances should be <1k ohm for best results due to the action of the multiplexer switching channels and the need for its internal capacitor to dischardge through the source.

If I am only using 1 data acqusition channel, will a source impedance of ~75k have negative effects, or will the fact I am not switching between multiple channels negate the need for a low source impedance? How bad will the results be with 75k source Z? I am sampling a 1 to 10kHz signal with the channel (the frequency will be fixed within this range and will be within +/-1V and will be a pure sinusoid) and would like to do it at max sampling rate (250Ksps).

Thanks for any help,

Mark Hayenga

[Neal M]

Hello again Mark,

I found what I believe to be the section in the manual (on page 4-6 of the manual linked a few posts ago).  The reason that it suggests using a low source impedance is due to ghosting of your signals between channels.  Simply put, this means that if you have a large impedance and are reading 2 channels of different voltages, there my be some residual voltage read from the first on the second.  For example, if ch1 is at 10V and ch2 is at -10V and we have a high source impedence (and high sample rate), then we might read -3V on the second channel:



Since you are not switching channels (or large voltages), then you should have no problems with your signal.  From what I can tell from your posts, it appears that you are looking to purchase hardware.  I would strongly recommend talking to your technical representative (www.ni.com/contact).  They are great at finding exactly what you need and what will work best for you.


Message Edited by Neal M on 12-12-2007 05:35 PM