USB-6009 - multichannel recordings?

[gobeirne]

Hello folks,

I'm posting on behalf of a colleague who is asking whether it is possible to do multichannel recordings from a USB-6009 DAQ using LabVIEW 7.0. They don't have any problems with single-channel recordings, but say that while the PDF of the manual hints that multichannel recordings are possible, they can't quite figure it. They've searched the Developer Zone but have drawn a blank. Any tips that I can pass on?

Many thanks,

Greg O'Beirne

[pallen]

I had similar problems with a USB6008.

You can do a multichannel recording relatively easily if you setup a multichannel task in NIMAX.

For my application I created a task that takes 2000 samples at 1K/sec on all four of the 6008's available analog channels.

If you need some guidance on how to create a task in MAX, just ask.

[Hieu Tran]

I had problem to get multichannels with high rate of sample.

Detail: I want to get 16 channels with 200KHz.

When I choose rate of sample: 200KHz, daq tells error.

Please help me how to solve this problem.

Thank you very much

[drjdpowell]

Your device likely only has one ADC, with a multiplexed input, that can do 200kS/sec in total on all channels, or 200/16=12.5 kS/sec on each of 16 channelschannel.

[rkesmodel]

Actually, it will be a little slower than that.  The rule of thumb I have always been given by NI tech support for all NI mutiplexed devices is (Max samle rate/# of channels*0.8).  So, it would be 200/16*0.8=10KS/sec per channel.

 

If you need faster sampling you will either have to buy a faster module, or one that will do simultaneously sampling. For multiplexing at 200K across 16 channels you would need a module with a 200/0.8*16=4MS/sec capability (at least).

 

Roy

[hooovahh]

Actually, it will be a little slower than that.  The rule of thumb I have always been given by NI tech support for all NI mutiplexed devices is (Max samle rate/# of channels*0.8).  So, it would be 200/16*0.8=10KS/sec per channel.

 

If you need faster sampling you will either have to buy a faster module, or one that will do simultaneously sampling. For multiplexing at 200K across 16 channels you would need a module with a 200/0.8*16=4MS/sec capability (at least).

 

Roy

I have never seen this mentioned before.  I have a 4 channel 400Ks/s device and I can read all 4 at once at 100KHz each.  Maybe what the NI tech support was trying to say is that when you are maxing out the speed of the card and switching that fast, you usually end up with a "ghosting" affect where the other channels appear on top of the wrong one because the impedance of the line is being affected by the previously measured channel.  For this reason adding some extra time between channel reads is a good idea, or even reading the same channel twice and throwing away the first reading as suggested in the linked article.

[hooovahh]

I am in possesion of the USB 6008 and 6009, but it appears that these will not allow for simultaneous data acquisition.  I am a student looking for a low cost solution that is portable and can gather data from at least 4 inputs.  Any suggestions?

The 6008 will gather from 8 single ended analog inputs.  Just not at the same picosecond.  I suggest you read this:

http://digital.ni.com/public.nsf/allkb/4F9D107D8B26233B86256F250057C9B3

 

and this:

http://www.ni.com/white-paper/4105/en/

 

In almost all situations I'd say the muxing of the ADC is acceptable.  It only doesn't work when you need to correlate two measurements of something that can change very quickly.  Things like measuring 2 sine waves and wanting to know how they shift in time.