And a piece of reflective pad and some IR diode emitter and diode =20 receiver, and measure frequency. You will know how is it and how dirty it is due to frequency variation. Or, if the fan has magnets, sense them with a hall effect device. En Sun, 28 Aug 2016 11:06:21 -0300, Denny Esterline = =20 escribi=F3: > On Sun, Aug 28, 2016 at 4:36 AM, James Cameron wrote: > >> Hmm, yes, or pair metal tubes intruding into the flow, upstream has a >> DS18B20, downstream a DS18B20 and a heating resistor in some thermal >> potting compound. >> >> > A good idea, in theory. > Playing the devil's advocate here, how do you calibrate around the random > insect infestation, or when they build a nest on your sensors? > Depends on the part of the world and seasons, but in some cases, even a > little bit of heat is going to attract the little buggers. > > You may have better luck measuring the _rate_ of heating and cooling =20 > rather > than static temperature. It would need to be a fairly slow process (10s = =20 > of > minutes?) With no airflow and a specific heat input, you'll get a > predictable rise and fall slope. With airflow, the rise will be slower =20 > and > the fall will be quicker. > If it's possible to turn off the fan every once in a while, (daily, > weekly?) a cycle with known zero airflow could be used to re-calibrate =20 > the > rise:fall proportion. > > With a little effort, it seems possible to build such a sensor in the =20 > wall > of the pipe, rather than having something protrude into the airflow - =20 > less > chance of crud collecting on it. > > > > > The other idea that occurs to me (and I don't like it) is to use a =20 > pressure > sensor. It seems "Easy enough" to arrange a diaphragm across a chamber > ported into the main pipe. With a suitably light/soft membrane and enough > area, you should get measurable deflection with the fan running. (I'm > thinking latex balloon, maybe 4" diameter?) Then do non-contact sensing > with a photo pair on the "clean" side of the membrane. > > Finding a long-lived membrane that can tolerate the environment you're > working in may be a non-trivial task, then you still have to deal with =20 > the > wildlife. > > > -Denny --=20 Usando el cliente de correo de Opera: http://www.opera.com/mail/ --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .