6009 voltage offset problem

[falkpl]

I have set up a program to measur 2 AI channels in differential mode from a USB6009.  When I hook anything up to the leads, I get a 1.4V offset.  When I don I get a small bit of noise (60HZ?)  What am I doing wrong? Is there something I have overlooked.

Paul

[Tom_W_[DE]]

Hi Paul-

It sounds like you're running into the issue described in this KB.  Basically, you need to make sure in your application that you specify the correct terminal configuration; 'default' is the default (strangely enough ) and results in a 1.4V offset if you use RSE connections.

Hopefully this helps-

[falkpl]

Thanks for the KB it didnt quite solve my problem.  I have configured the AI to be a differential AI not the RSE as mentioned in the KB.  Also when the signal is not attached there is no voltage difference but when the signal is wired to the 6009 I see a +1.4V offset.  I dont see this from a battery which measures the approperate ~1.5V expected value.  Very strange and has stumped me.  Since the offset is constant I am factoring it out but this is obviously not the ideal situation.

Paul.

[Jennifer O]

It is curious that you are able to read the battery but not your test signal.  Could you provide more information about the nature of the signal you are trying to measure?  Is it floating or ground referenced?  Have you tried bias resistors?  You can find more information about analog signals and bias resistors here.

Do you see the same offset voltage on each of the channels?  Since the offset is 1.4V I believe the issue is still related to the input circuitry.  If you connect one of your analog output channels to your analog input can you measure the full range of voltages accurately by running a test panel and updating the analog output voltage from 0-5V?  This is easier if you select the test panel from the popup that appears when you first plug it in, and then run a second test panel from Measurement and Automation Explorer.  This way you can have one continuously reading the analog input and also be updating the output at the same time.

Jennifer O.

[falkpl]

I am looking into the input circuitry.  I figure it is something to do with this.  The input comes from a analog PMT module, and some electronics wired by a fellow scientist at work.  I will see if I can get more information.  The PMT is measuring the scatter of an aerosol particle as it passes through a laser beam.  The PMT should be a slight negative voltage and go more negative as the scatter increases when a particle passes through the beam.  The voltage should be from 0 to -5V depending on the gain setting of the PMT's amplifier.  I am a CS guy not an EE so I will look at the bias resistor tutorials.  Thanks.

Paul

[Santaprene]

The USB-6009 analog inputs are surprisingly enough also DC-outputs ! ! !     Unloaded they source  1.4V and can supply 9.5uA into a short.
This makes the device useless for monitoring high impedance sources. The differential inputs are interacting thru the DC-bias resistor input network.

The digital side shares ground with the analog, not always a happy scenario.  The digital outputs are not totem-pole and are connected to 5K pullups so that even an unpowered device is still loading those lines.

Bus powered USB is convenient, but the PC-ground in your measurements may not be a good idea.  A separate DC-power supply, like some plug-in-the-wall devices made in the millions, along with an opto-isolators on the USB-lines, perhaps something like HCPL-M600's, should be an inexpensive upgrade.

These shortcomings of the device should be advertised by NI to prevent people from assuming the USB-6009 et al work like normal data acquistion equipment.

This has been a real waste of time.

[AnalogKid2DigitalMan]

Always go through the specifictions AND read the manual before selecting a device for an application. At least that is what I do. Yes, I agree, the 6008 and 6009 do have their shortcomings that people repeatedly run into. I think sometimes people make uniformed decisions on hardware and just buy the cheapest devices available.

[Santaprene]

"Yes, I agree, the 6008 and 6009 do have their shortcomings that people repeatedly run into"

All the more reason that this "problematic" information should be included the datasheet, not buried in the manual that may only be available when you get the device. People expect data aqusition devices observe conventions unless otherwise noted.  The fact that the unit is "cheap" makes it all the more reason to declare the goods.
At least provide the link to the manual, on  the http://sine.ni.com/nips/cds/view/p/lang/en/nid/14605  page. The manual says nothing about the inputs being "voltage sources", and that the inputs can be differential, implies isolation between channels. This is borderline false advertising and this should be corrected at once.

[Chris_Rake]

Sorry for the confusion on the specifications.  The USB-6008/9 User Manual and Specifications document is available online through the "Resources" tab under the product page.   The Analog Input Circuitry is show on page 11 and shows the 2.5V Ref feeding the resistor divider network used to scale and shift the input voltage.  In addition the input impedance is listed as 144kOhm in both the User Manual as well as the data sheet.  As far as isolation, the USB-6008/9 does not offer isolation, but another NI USB bus-powered device (the USB-6215) does offer isolation.  Also the USB-621x devices have >1GOhm input impedance which is great for high-impedance sources.  The full specs for these devices can be found here.

The USB-6008/9 are great for some applications, but if you need the performance (isolation, speed, input impedance, etc.) the USB-621x is a nice step up.

Message Edited by Chris Rake on 12-06-2006 05:38 PM

[Willy2a]

Is there any point in trying bias resistors with the 6008/6009 in differential mode, or does the unique input circuit already perform the same function? The input circuit already has 40k resistors to ground on each AI pin. I don't see the use of a parallel resistor. Am I missing something?