6062E PCMCIA DAQ card overheats

[CTRInsurgent]

Hi

I just got a 6062E card, and I'm having trouble with the analog outputs flaking out after the application runs for a while. The problem appears to be related to the temperature, although the 2 AO channels together are only driving a bit under 2 mA RMS, less than the 5 mA spec in the manual. Are there any steps I can take to keep the card running cooler, aside, obviously, from lowering the analog output voltages?
In the congfiguration I'm using, I'm running 2 AO channels, one of the counters, 2 AI channels, and the analog trigger is enabled with its output being routed to the counter source (with the counter gate coming from one of the AOs).

Thanks,

[JustinBritten]

Greetings,

There are many notebook manufactures who do not provide adequate cooling/heat dissipation for their notebooks. Therefore the heat generated by the notebook (CPU / Powersupply / PCMCIA devices, etc.) can accumulate. In most cases this additional heat will not cause a problem as long as the temperature is within the Operating Range of the DAQCard.

Many manufacturers are aware of this, such as IBM, who provides recommendations for minimizing heat.

http://www-1.ibm.com/support/docview.wss?rs=0&context=HW170&q=heat%2Bpcmcia%2Bthinkpad&uid=psg1DEDN-3CXSPC

I agree that lowering the analog output voltages is just a work-around and probably not what you would like to do. You should be able to run the card within specs and not have to sacrifice
the functionality of the card. However, there isn't much more we can do. Your notebook manufacturer is the best point of contact regarding heat dissipation in the PCMCIA slots.

Regards,
Justin Britten

Applications Engineer
National Instruments

[CTRInsurgent]

Is there any way to read the internal temperature sensor in the card from LabView?

[JustinBritten]

Unfortunately, no. The sensor is only used for calibration purposes internally.

Regards,
Justin

[CTRInsurgent]

Thanks.

Well, I've tested it in a 2nd notebook, and it usually takes longer (several hours typically) but it does eventually give the same behavior, the exact time required apparently depending on environmental conditions. In both cases the notebooks were operating in an air conditioned room in the mid 70s F or less. It also overheats when I'm just running the Test Panel doing analog input or letting the notebook sit idle.

Given that it overheats in two very different notebooks, a new Compaq Evo N800c (P4) and a fairly old IBM ThinkPad i series 1400 (Celeron), with very different PCMCIA slot configurations in a variety of conditions, one can only conclude that either this particular card is defective or the design is thermally
marginal for its intended operating environment.

Now, I don't doubt that the card works fine if you keep it at the specified maximum temperature, which was apparently changed in a revision of the manual from 50 C to 40 C. However, everything I've been able to find about notebook thermal engineering suggests that they normally run with an internal temperature in the mid 40s C. If you think that this is just a bad unit, I'd be happy to try exchanging it for one that works better, but... I think you can see where I'm having doubts about this.

Regards,
Ben Buckner

[LucaP.]

Hi Ben

Try this:
http://digital.ni.com/public.nsf/3efedde4322fef19862567740067f3cc/5222222ff438914b86256b0c000347c6?OpenDocument

Hope it works

Luca

[JustinBritten]

Ben,

The latest revision of the DAQCard-6062E User Manual is dated July 2002. In the July 2002 revision, the operating temperature is listed as 0 - 50 degrees C. This is fairly typical for PCMCIA devices. I performed a search on a wide variety of PCMCIA cards and found that most had operating temperature specifications of 0-50C up to 0-65C.

It looks as if Luca has pointed out a way to measure your card's internal temperature sensor. I was unaware this sesnor could be read on your card without using register programming; however, give it a shot and let me know what temperature your card is running at initially and after a few minutes to hours of operation. Let's see if the temperature in your laptop exceeds the ope
rating temperature of your card.

Regards,
Justin

[CTRInsurgent]

Yes - it does excellently - thanks!

[CTRInsurgent]

That form of the experiment is a little flawed though in that it measures the internal card temperature with the contribution from the card's power dissipation, not the backplane temperature.

So, what I did was to shut the card down (via Safely Remove Hardware), let it equilibrate for about 45 minutes in the laptop, and then power it up, taking the temperature shortly afterward so as to get close to the actual temperature of the reservoir into which the heat is being rejected.

What I found from that experiment was that just after powering up, the card was surprisingly cool, at around 41 C, indicating that the laptop internal temperature is normally around 41 C (perhaps slightly less even) without the heat from
the card, at least under my usual operating conditions. A measurement on the main heatsink with a thermometer gave just under 37 C, which corroborates that result. Once I leave the card running for a few minutes, it climbs up to a nice and toasty 55-56 C. At 55-56 C it does seem to work OK though, and I haven't been able to get it to show the overheat symptoms since I started monitoring the temperature, probably because the AC is going hard today. I'll let you know what I see the next time it starts acting strangely.

Thanks,
Ben Buckner

[JustinBritten]

Ben,

You have measured the temperature of your laptop PCMCIA slot as 41C. This is the temperature with no cards added. When you add a card, specifically the DAQCard, the slot heats up, as expected. Now, it is up to your laptop to provide proper heat dissipation and temperature control. As you said, it heats up to a toasty 55C. It doesn't look like your laptop is providing sufficient cooling.

In the meantime, I have sent an email to the Product Support Engineer for your DAQCard. This is my point of contact in R&D to determine what the operating temperature of the card should be and why the specifications have changed over the past few User Manual revisions.

Regards,
Justin