Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

How to read thermocouple output with NI USB-6009

I'm also running into this issue, with the exact 330-340mV lower limit. Originaly I had assumed that I had wired up the chip incorrectly only to discover that the DAQ (also USB-6008) was the culprit and not my shoddy soldering. I've now tested my setup using a bipolar power supply and DMM and everything works as advertised. Running the device off of the DAQ power supply results in the above described behavior where the output only varies with temperature if it is outputting something above ~330mV.

 

Some things I've noticed that may (hopefully) help resolve the issue:

 

  • Powering the chip with an external supply (using an Agilent E3646A) and measuring the output with the DAQ works (slightly biased high, but much better with no lower limit). In this case the DAQ's negative terminal is tied to the power supplie's negative terminal.
  • Powering the chip with the DAQ's supply or one of the DAQ's analog outputs does not work, giving a lowest reading of ~330mV.
  • When using the external power supply, if the PS's negative terminal is tied to the DAQ's ground, the 330mV reappears.


I really dont have any good idea of what is going on, but it seems that there is somethign about the DAQ's ground that does not play well with this device.

 

Any more ideas as to what might be causing this issue?

Message Edited by AKA me on 04-08-2010 11:14 PM
0 Kudos
Message 11 of 21
(4,059 Views)

Another interesting discovery...

 

The computer which I am using to connect to this device also has an NI PCI-6731 board installed. Unfortunately this device has no analog input, requiring me to use the USB-6008 DAQ to read in the AD595's output. However, this DAQ is also equipped with a 5V power supply for it's digital ports.

 

I attempted to power my circuit with this second DAQ's power supply and, interestingly, the same problem occured (minimum reading of ~330mV). This seems to imply that perhaps this problem is not necessarily specific to the USB DAQ, but rather that there is a problem with some/all of these DAQ power supplies, or perhaps simply with the computer's ground.

 

I will continue to update as I test further.

0 Kudos
Message 12 of 21
(4,053 Views)

I've tested the system using an older computer powersupply which I now use as a standalone powersupply. Here is what I have found:

  • 5V 20A line, output measured with USB-6008: Same broken behavior (~330mV minimum)
  • 5V 20A line, output measured with Extech Multimeter: works properly
  • -5V 0.3A line (connected + to minus to give +5V), output measured with USB-6008: ~960mV measured
  • -5V 0.3A line (connected + to minus to give +5V), output measured with Extech Multimeter: works properl


The implication seems to be that the combination of PC power supply, the USB-6008, and the AD595 IC causes the problem somehow. However, I believe that the standalone PC power supply was connected to the same circuit as the DAQ's computer power supply. I will try to repeat these tests with the power supply on a different circuit.

 

Then again the Agilent E3646A PS which did work with the USB DAQ was connected to the same circuit... but on the other side of a surge protector.

0 Kudos
Message 13 of 21
(4,052 Views)

Get yourself a clean powersupply!! 

Switched powersupplys need a lot of additional filtering to provide a clean voltage for a sensitive analog circuit. Hook it up to a good (fast, high BW, analog?)) scope and have a look how clean your 5V are!!

PC powersupplys are a great source of noise!

 

Try the 12V line (with a long cable to place the supply far away) with a 10R , >100µF , 7805, 100nF or use a battery.

 

Welcome in the wonderful world of MAXWELL 😛   and reality 😉

The breadboard is the point where theory meet reality 🙂  

 

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


0 Kudos
Message 14 of 21
(4,035 Views)
Oops forgot to reply to this thread.  Well I seemed to have fixed the issue by providing impedance matching using a voltage follower configuration.  Once I did that, it worked with any power supply.  May want to try this out and see if you get the same results 🙂
0 Kudos
Message 15 of 21
(4,023 Views)

Henrik:  While I agree that these powersupplies are probably not incredibly clean, I'm tempted to believe that noise is not the main problem here. My thinking being that if a dirty PS were the culprit, the device would not provide proper output when measured by any device. However in this case I am able to obtain proper behavior using the same power supply by simply measuring the output with a handheld multimeter or a Keithley DMM instead of the USB-6008. This is not to say that I dont think that a dirty PS will have other effects, but I'm starting to think that it is not the primary cause of this problem.

 

mdzz:  I'm glad to hear you've found a solution. If you have the time, would you mind posting a wiring diagram or schematic of what you ahve done? It would be greatly appreciated.

0 Kudos
Message 16 of 21
(4,019 Views)

Sounds like the pull-up/pull-down resistors (look at the manual for the USB-6008) may have been too much of a load for the TC amplifier chip. Perhaps its output stage only sources current? I don't recall.

 

Anyway, glad you were able to fix it. That was going to be the next thing that I was going to suggest, actually.

0 Kudos
Message 17 of 21
(4,006 Views)

Going back to the datasheet for the AD595 provides an answer.

 

First, here's the input circuit of the USB-6008:

 USB-6008-Input.png

 

That is, there is a 127K resistor to a point that is biased to 1.4V (to allow for the ±10V input range given only a 5V power supply).

 

Now, looking at the datasheet for the AD595, under "Output Current" we see "±5mA" but there is a footnote:

 
Current Sink Capability in single supply configuration is limited to current drawn to ground through a 50 kΩ resistor at output voltages below 2.5 V.

 

Since we're definitely working below 2.5V (unless you're measuring more than 250ºC, which I don't think you are), our model of the overall circuit has to include a 50KΩ resistor in series with the input.

 

Doing a "back of napkin" calculation (ignoring the resistance of the divider), we see that the additional 50K resistance contributes an additional voltage of:

 

((1.4V−250mV)/(50kΩ +127kΩ ))*50kΩ = 325mV

 

At a 0V output from the AD595, the additional offset is about 395mV.

 

Test this yourself on the USB-6008/6009 like this:

 

  • Connect a 50K resistor between your +input channel and ground
  • Connect the -input channel directly to ground
  • Measure the offset voltage. What do you read?

 

Mystery solved.

 

Moral of the story:

 

Always read the datasheet and the manual first!

0 Kudos
Message 18 of 21
(3,995 Views)
Another way to fix this (besides the rail-to-rail op amp as a voltage follower) might be to add a 50K resistor in series with the other USB-6009 input. You will probably have to adjust this a bit to null out the offset voltage (I'd do it at 0 ℃).
0 Kudos
Message 19 of 21
(3,994 Views)

As I said, the prior figures were ignoring the restor divider's contribution, and so aren't completely accurate (they tend to overstate the error by about 10%).

 

Taking the entire circuit into account, an input resistance difference of 50KΩ should contribute about 360mV of error voltage at 0V, and about 295mV of error at 250mV input.

 

Another moral of the story:

 

Always keep the input resistance equal on the USB-6008/6009!

Message Edited by ned_konz on 04-10-2010 10:56 AM
0 Kudos
Message 20 of 21
(3,992 Views)