air flow sensor on PCB

Negative temperature coefficient thermistors are about an order of magnitude more sensitive to temperature change than metal film resistance thermometer elements and tend to have specified self-heating constants in still air. A leaded glass-bead thermistor should work.

This might work. Tad expensive, sole source.

torsdag den 6. april 2023 kl. 06.36.04 UTC+2 skrev John Larkin:

a bridge with two RTDs/NTCs, one of them in the air flow the other shielded from the airflow?

Hot wires are quick and simple. As an extra bonus, if you pick the right excitation frequency you can get rid of the poorly-controlled sensor-to-board thermal conduction effect.

What's such a nuisance about it?

-- Cutting the bulb?

-- Soldering the leads?

-- Calibration?

Or something else?

Cheers

Phil Hobbs

this is cute idea, and cheap

RS

and Mr Woodward's simpler circuit, if you just need a binary output (instead of linear analog)

Breaking a light bulb tends to be erratic; we've done that. And a nasty production process. It looks ugly too.

We sure don't want to calibrate an air flow sensor. The flow sense would be a selling feature, a fan+filter check, so we'd want decent accuracy as-built.

A thermistor would do the same job - rather better - and looks neater.

You haven't got a lot of choice.

Then you probably need a turbine meter, or an ultrasonic flow meter that will actually measure just the air speed.

I once worked on a vortex shedding flow meter - it only works for turbulent flows and you don't seem to be looking at that kind of airflow, but it was linear and remarkably precise, if hideousy expensive and somewhat bulky.

That's a nice curve, thanks. I was going to measure that myself. Both your cited curves align with the data sheet value of 200 K/W for a TO92 in still air, and I figure we'll have about 200 fpm air flow.

I did this last night, before (cross-my-heart hope-to-die) I saw your posts.

We could turn on the SSR and heat up the transistor, and then turn it off and measure Vbe as it cools down. The flow measurement is essentially independent of ambient temperature and the parts cost is under a dollar. "SSR" could be a couple of discretes and make it even cheaper.

I guess I'd note to manufacturing to use some long, specified lead length on the PNP.

I'll have a 1 GHz quad-core ARM not working hard, so we can apply lots of software. And the product will have a multiplexed 12-bit ADC already.

But other ideas might be fun too.

Another way is to measure the temperature of air rising from some normally warm part. If the wind blows, the convected airstream goes away and the measured temperature drops.

Bit of a silly idea, but hey.

One can get 1000-ohm platinum RTDs for small dollars:

.

If the issue is just to tell if there is a lot of airflow, the temperature delta does not need to be large, so a simple bridge with AC source and synchronous detector should work.

Cheaper still are RTDs made of nickel, versus platinum. Nickel is OK if the temperature in still air isn't too high.

Joe Gwinn

Not silly, but the hassle is implementation. We once made a flow sensor based on one part heating another part downstream in the air flow, and it wasn't reliable.

My current thinking is to measure the theta of a part sticking up off the board, where theta is temp rise in degrees C per watt. That will drop as air flow increases.

The game is to make that as cheap and simple and as manufacturable as possible.

I was just looking at the EDN article cited above, and noticed the schematic in fig 2.

One could list the several horrors.

Thermistors are ten times more sensitive than metal film resistance thermometers. You don't need to get them all that hot to get a decent sensitivity, and they tend to have specified self-heating coefficients in still air. Interchangeable thermistors are pretty stable, if not all that cheap.

Semiconductor temperature sensors can also work. They tend to be noisier.

We use leaded ceramic slab and 1206 platinum RTDs, and a cute SOT23 nickel RTD, but they would be down flat on the PC board. I think it would be better to get the sensor up in the air stream, so the preferred gadget might be a TO92 transistor.

A 2N4402 costs us 7 cents. An RTD wouldn't break the bank, but simple and cheap are games we play.

Just monitoring temperature in critical locations is usually sufficient instrumentation for most applications.

Fan tach is ~free, but is misapplied if its absence alone inhibits function. It's a warning only, not a detection of imminent system failure.

Two point temperature sensing (or single point plus ambient) can be calibrated to monitor flow, if it's really required.

RL

Too bad fan mfrs don't offer them without magnets or motors, just tach.

A simple sub into a fan slot would give indications of system air movement/pressure.

RL

Wow, that schematic is terrible

An unpowered fan with tach, run at PWM=0, is an air flow turbine meter. But my board can't access the crate fans and is too small to have its own fan used as a flow sensor. But it is an interesting idea.

Once long ago, before fan tachs were popular, on some AC powered fans, I did an electrostatic blade tip sensor to perform essentially the tach function. Turns out that a spinning plastic blade generates a nice e field.