Air flow measurement

[yesway]

Hi!

For a medical (not life supporting) device we need to measure air pressure and air flow. The device must be low cost. We have found a few pressure sensors which have proved suitable (in the 1-2USD range) for the pressure measurements. We are still investigating a simple and reliable method of measuring the air flow. We have read about several methods, but have no experience to the results we can expect from them. It seems nice if we could use differential pressure (using 2 pressure sensors) over an LFE (Laminar Flow Something?), but will this work with the small pressure differences (+-30mm H2O) we are interested in? We have a micro that can do temp. comp. and other math ops. Advice to guide us in the right direction would be greatly appreciated!

 

[cbarn24050]

Hi, a hot wire anemometer is ideal for low speed air flow, and is very easy to do.

 

[yesway]

Okay, should have mentioned that we wish to keep power consumption minimal (battery operation). Perhaps this is still possible in some way? Could you please give more details, like expected power consumption, or if other methods are better under these circumstances?

 

[cbarn24050]

Hi, power consumption depends on the sensor size, if you use a transistor die you can get down to a few milliwatts. You could also cycle it on and off to reduce it further.

 

[DRWeig]

Hi,

Thought I'd pitch in with my low-air-flow experience...

CBarn's thoughts are on the right track -- there isn't a really good alternative to a hot wire when flows can vary way down low. Although your max range (±30 mm H20) is available in a differential pressure device, remember that this differential pressure falls off with the square of the flow. If you want 10 to 1 turndown in flow, your transducer has to be capable of 100 to 1 with some reasonable accuracy. Pretty hard to do reliably, especially if budget dollars are low.

If you need to know direction of flow (as indicated by the bi-directional range you stated), that is another problem to solve. Let us know how it goes!

Best wishes,

Old Dave

 

[yesway]

Hi folks!
Thanks a lot for your advice, maybe I can send you a respirator sometime to thank you Well, we arent really building respirators, but rather accessories. For the pressure we need to know + and -, but for flow we have got a "wish" from customer to handle an air volume which I think correlates to max 1m/s in a quite small plastic tube, say 5mm dia, without the need of directional information (one way vent installed). The medical aspect would make use of a resolution down to 1mm/sec, but I have "only" agreed to try to get 10mm/sec. I dont have much experience in this field, so even that sounds challenging to me, please correct me if it is fairly easy to do... since we are dealing with breathing, we would like to be able to follow each exhale, to finally be able to measure air volume, I would say 10 measurements/sec would do fine, leading me to think that the sensor, e.g. transistor die must be powered up all the time. Do you think it will respond fast enough? We really must not follow the air velocity during the exhale, but we need to get a good average. A delay in response is ok, even several seconds.
At this point we have lots of time, project launches August-September, but I am sure we have some long nights to look forward too :I