USB 6259 Operating Temperature

[TriL]

Hi Folks,

I'm building a data acquisition piece of hardware that may be sitting out in the sun for extended periods of time. My estimate is that the internals of the box can go as high as 55 degrees C.

Unfortunately due to the nature of the hardware, the DAQ must be USB. 

Looking at the environmental specs on the 625x boards:

PCI/PXI/PXIe: 0-55C

PCIe            :  0-50C

USB            :  0-45C

My questions are:

1. Why is it that USB is so much lower than PCI? Is it purely due to the USB controller? If I put a heat-sink on the USB controller can it operate up to PCI level?

2. What affects would operating the USB 6259 at 55C? Is it simply a degradation in AI accuracy?

3. Are there any alternative USB solution that can achieve 55C and still have the same number of AI and sampling rate?

4. I've seen non-NI USB DAQs which are rated up to 70C! Is NI been conservative in their specifications?

Any help you can provide is appreciated.

Tri

[TriL]

I should mention that I'm referring to the NI USB 6259 OEM boards.

Cheers,

Tri

[Corby_B]

Hello Tril!

Thanks for your post and questions about the environmental specs of the USB 6259. In the manual it does say the following:

Environmental specs on the 625x boards:

PCI/PXI/PXIe: 0-55C

PCIe            :  0-50C

USB            :  0-45C

The reason that the PXI/PCI boards are at a higher spec is because the cards are usually in a systems such as a computer or PXI chassis that is internally cooled. This is usually done with fans or other methods. Since the USB DAQ is not cooled by an external method the temperature is a bit lower than our PCI/PXI cards. I would advise, if using above these temperatures that some type of cooling be used.  We do have our C-Series modules that can go up to 55C in operating temperature. What could happen if operating above spec is the system could shut down and stop sending data. WIth 45C equating to about 113 degrees F I would say that the OEM 6259 could be operated in most palces with minimal cooling. 

NI cDAQ-9172

https://www.ni.com/en-us/support/model.cdaq-9172.html 

Let us know if this clears up some of the questions/concerns you may have. Thanks again for your post!

Cheers!

Corby_B

[TriL]

Hi Corby_B,

Thanks for your reply.

Is it correct to assume that if the internal temperature sensor of the board reports a temperature within spec then it should operate within spec? If this is correct then it means that it shouldn't matter whether or not the chassis is internally cooled or not - therefore USB boards should be the same as PCI/PXI boards as it is independent of the environment.

You mentioned that if operating above spec, the system could shut down and stop sending data. Can you please clarify whether this is a planned event triggered by the onboard electronics or a non-deterministic event? If it is triggered by the onboard electronics, what is the absolute temperature that will cause this for an USB-6259 OEM board?

Thank you for suggesting the C series modules, I will investigate this but space is at a premium.  Can you suggest any other boards similar in dimension as the USB-6269 OEM boards that has a higher operating temperature? The minimum requirement is that it has at least 8 AI ports, 8 DIO ports, max sampling (on AI) of 1Mhz and USB interface.

Thanks again for replying and I hope you can assist me a little further.

Cheers,

Tri

[Neil_S.]

Tri,

The temperature spec is for ambient temperature, not the temperature of the board.  The temperature on the board can reach higher than the 55 deg C specification even in PXI or PCI.  This is acceptable.  However, if the ambient temperature exceeds this temperature the boards will become less reliable, will fail more often, and may fall short of their accuracy specifications.  The USB devices have a lower speced temperature because USB devices are typically used in a desktop environment and we chose not to test the boards at higher temperatures.  There is a good possibility they will still operate at higher temperatures, and I don't know of any component differences that would make them more susceptable than PCI or PXI devices.  That being said, we did not test it and we cannot know for sure.  I would recommend fully testing your final application at this temperature before deploying it.  My one final parting recommendation is that you run selfcal on a regular basis.  Large swings in temperature can be devastating to accuracy.  Self-cal will help bring the board back into specifications.

Let us know if you have any additional questions.

Regards,

Neil S.

National Instruments R&D