Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

Measuring temperature rise up to 800 degrees Celsius with USB-6009

Hi,

I have a USB-6009 DAQ and Signal Express software and would like to measure a rise in temperature. Would it be possible to do so?

I have an instance were I need to measure the heat transfer through a series of sample plates coated in differing thickness of intumescent paint.

I will be applying around 800 degrees Celsius to the coated side and would like to measure the temperature difference and rise time on the other side of the sample.

 

I have a series of K type thermocouples which I will be using, should I use somesort a cold juction compensation?

 

Many thanks for any advice.

 

Andrew

0 Kudos
Message 1 of 17
(5,197 Views)

You can make a measurement but there will be some limitiations. On the +/-1 V range the absolute accuracy is 1.53 mV = ~ 37 degrees.  The system noise is 0.5 mVrms = ~ 12 degrees.  Resolution is 61 uV = ~1.5 degrees. You could have errors of 50 degrees or greater on a single reading at 800 C. By averaging multiple samples (to reduce noise) you could probably detect changes on the order of 2-3 degrees.

 

I suggest that you consider using an AD595 thermocouple amplifier and cold junction compensation. It produces 10 mV/degreeC over the entire Type K range. The data sheet has formulas to compensate for the thermocouple non-linearity.  They are readily available from major electronics distributors.

 

Lynn

Message 2 of 17
(5,183 Views)

And if you don't want to build it, there are many industrial signal conditioner that will convert the type K µV (0-1000V)  to a 0-10V ...

Drawback is, that 2 usually will cost about the same than your DAQ.

one example

http://uk.rs-online.com/web/p/signal-conditioning/5330435/

 

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 3 of 17
(5,167 Views)

The problem in using the AD595, is the limited temperature range. I'll probably end up using a multimeter.

0 Kudos
Message 4 of 17
(5,116 Views)

The datasheet for the AD595 shows that with a >= 15 V power supply it can go to +1250 C. With a 5 V supply it is limited to about +300 C. The IC itself of course has a limited operating range of -55 to +125 C bu tthe thermocouple can cover its entire range.

 

Lynn

0 Kudos
Message 5 of 17
(5,110 Views)

Thanks very much for your advice chaps..

I now have a voltage reading into signal express, the reading is a little high. I'm getting around 330mV at a room teperature of 22 degress. Acording to the AD595 data sheet it should be 250mV. Should I just adjust with an ofset, or do I need somthing else in the circuit?

 

Many thanks

Andrew

0 Kudos
Message 6 of 17
(5,025 Views)

Ok, While I'm asking, can I apply a transfer function in signal express? It seams I could do so in Labview (which I don't have), but signal express make no refrence to transfer functions.

 

Andrew

0 Kudos
Message 7 of 17
(5,018 Views)

Andrew,

 

I am not sure what to recommend about that room temperature error. 

 

Are you operating the AD595 with a 5 V power supply or a higher voltage which will allow you to read your higher temperatures? The reason I ask is that the data sheet appears to specify the accuracy with a 5 V power source and does not indicate in any way how things might change as the supply voltage increases. Clearly the device power dissipation will be higher resulting in more self heating. I looked at the Applications Notes on the AD website and they do not address this issue either.

 

Make sure that your connections to the thermocouple follow the recommended practices. I doubt that is the issue at room temperature, but get it right before you do anything else.

 

Can you test at a sampling rate of a few hundred Hz or 1 kHz? This would allow you to see if power line frequency interference is present. If you are sampling at low rates, a 50 or 60 Hz interference could look like an offset.

 

Before adjusting anything run some calibration tests over a range of temperatures. You need to determine whether the error is an offset or a gain error. You cannot distinguish between the two from a single point measurement.

 

I do not know anything about SIgnal Express.

 

Lynn

0 Kudos
Message 8 of 17
(5,009 Views)

I'm powering the ad595 with 15v. 

When appling 5v or a range of voltages upto 15v, the room temperatue is measuring about 325mv through the USB 6009, however when I measure the AD595 output with DMM is about 250mv which is about right.   

The USB 6009 must be adding some (75mv) voltage to the AD595 output signal. Do you think that there could be a impeadance missmatch?

 

 

I will compare a range of measurements when I get some new therocouples. I've managed melt the insulation around the tip

 

Thanks

Andrew 

0 Kudos
Message 9 of 17
(4,970 Views)

Andrew,

 

After taking a close look at the datasheet for the AD595 and the specifications for the USB-6009, I may have at least a partial explanation or two.

 

The AD595 datasheet specifies (footnote 4 on the Specifications table) that:

 

"4Current Sink Capability in single supply configuration is limited to current drawn to ground through a 50 kW resistor at output voltages below 2.5 V."

 

The USB-6009 has a wierd input circuit as shown on page 16 of the User Guide and Specifications:

 

USB_6009 input.png

 

I have not done any detailed calculations but I would not be surprised that the combination produces the results you see. At temperatures above 250 degrees this should not produce an error. At 140 degrees the error should be zero.

 

The second issue is the absolute accuracy of the USB-6009. For a single ended configuration at +/-10 V the typical accuracy is 14.7 mV but the maximum over temperature is 138 mV (page 25).  While I think it is highly unlikely that the error would be as large as 75 mV at room temperature, the specification does not rule it out.

 

I think the circuit configuration is the most likely culprit. The fix, if you need one, would be to put a unity gain buffer with low offset voltage (so you do not introduce more error) between the AD595 and the AI connection.

 

Lynn

0 Kudos
Message 10 of 17
(4,955 Views)