pcb-mount air flow sensor

=A0 =A0 =A0...Jim Thompson

Roger that.

-- Cheers, James Arthur

Hot Air is reporting it contains anti-Constitution statements. ...Jim Thompson

We need a president with a MBA. What, you said we tried that once before> Oh yeah, George W. Bush!

Just how did that work out?

=A0 =A0 =A0 =A0 ...Jim Thompson

MBA "studied economics." Last week I heard Geithner and Steny Hoyer both say the economy's slowing, so gov't needs to spend more money. Brilliant.

Not well, but trillions and trillions (and about 7 million jobs) better than now.

Hey, here's an epic fail: GM got $50B in bailouts, $27B illegally looted from bond and shareholders, all the assets of GM, and $20B swindled from a public offering to investors Nov., 2010[1]. ($50B + $27B + $20B) =3D $97B in cash alone. Today it has a closing market value of ... tada, $31B.

So, by turning ($97B in cash + ??? in assets) into $31B--and losing most of the jobs anyhow--BHO "saved it."

Yeah, some basic economic and investing know-how sure would help.

-- Cheers, James Arthur

ng about maybe using two LM71 temperature sensors, one mounted on a copper = pour with a resistor heater. We don't need a lot of accuracy... this would = be mostly a fan fail sensor.

(On/OFF signal), the Flexpoint Bend resistant Sensor can meet your requirem= ent. The sensor is cheap, if you have any interested in the sensor, you can= check the link

A real mechanical sensor would be fun to make--say some optically- sensed windmill, or your flex sensor. In real life I'd worry about cobwebs and dust though.

Just two diodes in series, one on a copper pour, constant-current drive, sense the voltage in the middle, that should be pretty good. Might even be simpler to drive than Phil's thermistors.

A fin or a grid of Cu posts for the sensing diode, sticking up in the air, might be better than a pour--that improves the thermal interface to the air stream, and you could orient it to respond to airflow in the desired direction.

-- Cheers, James Arthur

If the leds are protected from the customer you just need the right permanent marker...

For

=A0 =A0 =A0 =A0 ...Jim Thompson

Sure - he sat on hios hands while a house bubble blew up - generating lots of jobs, for a while - and then burst, removing all those jobs, many of them before he vacated the White House. Great economic management - in James Arthur's rather biased opinion.

in James Arthur's bizarre opinion

With James Arthur's Tea Party co-delusionists in Congress busy screwing the economy in the hope of getting re-elected, it's surprising that it's worth that much

Saved as much as could be saved. No doubt James Arthur thinks that some right-wing nitwit could have done better, but that's a permanent delusion with him.

But nowhere near as much as if Dubbya had had insight in that area and damped down the house prices rises before they blew up into a bubble that had to burst.

The Karl Rove approach to economic management - elect a Republican, watch him wreck the economy, then criticise the Democrats for not getting fast-enough recovery. The kind of strategy that can only succeed if the electorate is as poorly informed as the right-wing nitwit clowns who post here.

-- Bill Sloman,

marker...

You aren't very smart, are you? One look at units from the the two contracts made it obvious that some were brighter than others. As I stated in another message, they were 'intelligent ASCII displays', not just LEDs.

Probably the big chip is the absolute air pressure sensor, uses piezoresistivity of the strained silicon (with a sealed vacuum chamber on one side, and atmospheric pressure on the other).

I tried an LM45 and an 0805 heater resistor on a small copper island on a board, and the temperature change versus air flow is unimpressive. Too much heat is being lost to the board.

A 2N4400 (NPN, TO92) transistor is better, sticking up off the board into the air flow. If we run it at constant current and snoop Vbe, Vbe changes as a function of air flow. At 750 mW dissipation, Vbe changes about 60 mV from zero air to 150 LFPM.

We were thinking that we could run it at constant current and switch Vbe from, say, 2.5 to 12 volts and measure Vbe in both cases. That gives us an indication of Theta-ja that should be pretty independent of individual transistor variations.

I guess we'll put that on the board layout and, if it works, announce the feature.

Bubba lit the fuse when he signed the bill that killed the Glass?Steagall Act, made worse by Fannie Mae buying up unlimited amounts of bad paper, taking the banks off the hook for home loans. The government manufactured the crisis.

W is on record as having warned about the real estate bubble, but nobody in Congress was listening. Too many people were getting "rich" from the appreciation of their real estate.

Of course, the same things were happening in Ireland and Spain and the UK.

Why do people think that rising house prices are a good thing?

Hardly anyone - the public, Congress, economists, certainly not the President - is thinking long-term. If you keep doing stuff that feels good now, but accumulated longterm damage, eventually things get worse.

I looked at a data sheet for the sensor for a thick film mass air flow meter. it had a schematic showing the internal and the principle of how to drive it

looked like it basically had a ~1000R platinum sensor (at a low current) to measure air temperature, and a ~45R platinum sensor at a higher current i the airflow.

the measured output was the power(voltage) need to keep the heater some degrees higher than the air temperature

so something like a 2*pt1000, two resistors and an opamp

-Lasse

^^^^ Vce, I meant. That will toggle power dissipation.

That would work, but we'd have to get the RTDs up off the PCB surface somehow. We have some thinfilm RTDs in stock, little ceramic things with tiny leads, but they look flimsy to mount up off the PCB. We have

An axial, maybe quarter-watt size, resistor with a high TC would work. It could be spaced up off the board.

The TO92 is at least a standard part, and should be rugged mounted a ways off the PCB surface, in the air stream. It costs 4 cents.

We had some other ideas that were mechanically difficult to implement on a PC board. My favorite was a heated metal tube with a temp sensor inside. A little air flow would make a huge difference to the sensor temp. But that's messy mechanically.

We considered an electret microphone and a bandpass amplifier, to pick up low frequency turbulence. Vortex shedding flowmeters work vaguely like that.

Put some nice tall electrolytics beside it, get some venturi / skyscraper effect?

Or if you use some conspicuous heatsinks (in the sense that they are heatsinks, but not being used to sink heat), you might get more directionality.

Tim

No need for a fancy IC; just a couple of cheap diodes (or diode-connected transistors) in a bridge, will do ya. Heater in the middle, the one downwind of the heater is ... warmer.

Or, you could try putting a teensy fan on the PCB, and monitor its current draw. High current, it's bucking the forced airflow. Intermediate current, airflow is small or absent. Low current, airflow is in the fan direction.

s. For

=A0 =A0 =A0 =A0 =A0 ...Jim Thompson

ism.

ed

The government certainly didn't do anything to stop it. The bankers had been campaigning for the Changes in the Glass-Steagall Act for mnay years, and eventually got what they wanted, but Fannie Maue didn't actually buy up any "bad paper" - homes loans made under the Community Reinvestment Act

sustained much the same default rate as regular home loans. It was the sub-prime - ninja - loans from the bottom feeding mortgage lenders that defaulted on a large scale in the aftermath of the sub-prime mortgage crisis.

James Arthur claimed otherwise immediately after the bubble burst, but he was - as usual - seeing what he wanted to see. When the numbers came in, the Community Reinvestment Act loans turned out to have been made to good credit risks who rarely defaulted.

It was the bottom-feeding end of the banking community who manufactured the crisis. Every last one of them should now be doing hard labour in a chain gang, but I don't think that anybody has been either charged or convicted of making loans to people who shouldn't have got them or charged with fraud for selling them on as if they were regular home loans to people who looked likely to pay them back.

It isn't the executive branch's business to warn about problems - their job is to anticipate them. If Dubbya was actually worried, it was his duty to get off his behind and do something to solve the problem before it turned into a crisis.

state.

And were too dumb to realise that the appreciation was going to be transient.

Ireland was a very rapidly developing economy at the time. Property prices collapsed when the world economy stopped growing, and with it the Irish economy, and a whole lot of property that could have been useful to a rapidly growing economy became worthless.

Spain was building a lot to serve the tourist trade, but that's a luxury industry - the global financial crisis meant that a lot people decided not to go to Spain for their vacation, or put off buying a vacation house there; again, the Spanish economy had been growing rapidly - recovering from the after-effects of Franco's right-wing nitwit economic policies and the global financial crisis hit them very hard.

The U.K. did have a house property price bubble, as they'd also had several times in the period when I was working there, and the global financial crisis made sure that it burst. It wasn't anything like as bad a the US version, and the consequences were less dramatic. The Northern Rock Building Society was the famous casualty

but "In 2006 the bank (Northern Rock)had moved into sub-prime lending via a deal with Lehman Brothers. Although the mortgages were sold under Northern Rock's brand through intermediaries, the risk was being underwritten by Lehman Brothers."

The can borrow against the higher value of their houses ...

But James Arthur's idea of thinking long term is to minimise debt at the expense of everything else. Keynes pointed out that killing the economy isn't a a great way to save it. Your Democratic administration is doing what Keynesian pump-priming it can, but the Republican majority in Congress really doesn't want the economy to actually recover until after the presidential election.

As far as I can seen, you current administration is thinking long- term, but - like all politicians - their primary interest is coping with the situation as it is now, and that means dealing with a series of short term problems, many of them created by the Tea Party.

-- Bill Sloman, Nijmegen

Vortex shedding flow meters only work if the air flow past them is fast enough to be turbulent. Then any kind of bluff body sheds a series of vortices in alternating directions at a rate which is linearly dependent on air speed, over a very substantial range of air speeds. The one I worked on in 1974 incorporated what was effectively a capacitance microphone to detect the alternation, but the actual moving part was two very thin stainless steel membranes electron-beam welded onto either side of our rectangular "bluff body" and the capacitance measured was between the membranes and fixed electrodes just behind them (inside the block)in a space filled with thin oil.

I can't remember what sort of shedding frequencies we saw in air - it must have been something in the audio region. The second URL does mention a specific frequency (around 200Hz) and includes enough discussion about Strouhal numbers that you could work out what it would be in your application.

-- Bill Sloman, Nijmegen

ediate

n direction.

Too many moving parts. Vortex shedding is nicer. A thermistor behind the bluff body would see very different air speeds as one vortex detached and the next started up, rotating in the opposite direction.

-- Bill Sloman, Nijmegen