LM1819 Air Core Meter Driver...

Interesting. Looks like a lot of TANH :>

piglet

Sort of. Gilbert cell multipliers to approximate sine and cosine.

Probably all done now-a-days with a uP. ...Jim Thompson

Den tirsdag den 6. marts 2018 kl. 23.12.02 UTC+1 skrev Jim Thompson:

he LM1819 Meter Driver? Air Core meters have two coils mechanically 90" ap art. Used mainly in automotive applications because of the available rotati on angle compared to DeArsenval movements.

n the datasheet schematic of the chip internals... Not that I'm surprised t hat happened.

would be very easy with a small uP but you'd need a voltage regulator and you'll need a driver unless you can get by with much less voltage swing

Hm. So, Taylor series?

SIN(V) = k*V - (k* V)^3 /6 + ...

COS(V) ~= 1 - (k*V)^2 /2 + ...

That does only take two multipliers... and the range would be limited to k *V in {-2.5, +2.5) range Its a little brute-forceish, though, you'd print the numbers to make the ends fit

There's some benefit from Chebyshev rather than Taylor fitting, but that's an analysis that can wait. It's dinnertime...

Go ahead and use sin from the Taylor series, but derive cos coefficients from minimizing the deviation of the ratio from tan(k*V)

There's ways to do sin\cos slicker... phase-shift oscillators and timed sampling would be more than fast enough to drive a meter needle.

Nope. It's ratiometric, doesn't take a tight reference at all (except as needed to generate the input numeric). ...Jim Thompson

Den onsdag den 7. marts 2018 kl. 04.12.26 UTC+1 skrev Jim Thompson:

n the LM1819 Meter Driver? Air Core meters have two coils mechanically 90" apart. Used mainly in automotive applications because of the available rot ation angle compared to DeArsenval movements.

e on the datasheet schematic of the chip internals... Not that I'm surprise d that happened.

m1819.pdf

ing

sure, but you won't find a uP that runs on 12V

That's why outputs can be taken as PWM ;-) ...Jim Thompson

Den onsdag den 7. marts 2018 kl. 04.44.22 UTC+1 skrev Jim Thompson:

s on the LM1819 Meter Driver? Air Core meters have two coils mechanically

same on the datasheet schematic of the chip internals... Not that I'm surpr ised that happened.

s/lm1819.pdf

swing

but you'll still need a ~5V supply for the uP and 20mA is pushing it

Oh, I know how to do it with modern parts*, I really just want to know how that core works for fun.

• Not just withlookup tables either...

Steve

There are even slicker ways than that--see the AD639 sine/cosine/tangent generator. Barrie Gilbert has a patent that explains how it works.

Cheers

Phil Hobbs

TAN(Theta) = Icoil_1/Icoil_2... it's just a vector machine.

When I get a "round two-it" I thought I'd fix my nice native American decorative outdoor wall thermometer by adapting a car gauge to drive the pointer, rather than the crappy inaccurate bi-metal.

The "core" is just two Gilbert cells. ...Jim Thompson

something like an attiny85 gives you the three PWMs and an ADC in a 8pin DIP add an NTC and bit of code an you are set ...

I want better accuracy than an NTC... probably adapt something like this...

...Jim Thompson

Den fredag den 9. marts 2018 kl. 02.04.51 UTC+1 skrev Jim Thompson:

ow how that core works for fun.

DIP

so calibrate the ntc ...

LM35? Good for +/- 0.25DegC @25DegC, +/- 0.75DegC from -55..150DegC

You can get NTCs guaranteed interchangeable to 0.01C from several vendors. Digikey has gobs of +-0.1C ones for about three bucks in onesies. (Search for "interchangeable thermistor" on your favourite search engine.)

The main issue is thermal leakage via the leads, but that's not nearly as bad as in IC sensors, which have big fat copper leads (400 W/m/K) and plastic packages (0.1 W/m/K). There's a National Semi handbook that talks about that, which you can get at .

Cheers

Phil Hobbs

Provided that it's the temperature of the board that you're interested in. See the link I posted just upthread.

Cheers

Phil Hobbs

And makes a mighty-fine core for a full-featured function generator that fits easily in a brief-case-sized bed of nails tester for PCB's...

See "Function_Generator_OmniComp-GenRad.pdf" on the S.E.D/Schematics Page of my website. ...Jim Thompson

They've been discontinued for a long time, but they're pretty neat. I have five or ten in my drawer still.

I used them in the proto for the first commercial atomic force microscope, the IBM SXM, which was later sold to Veeco and then (iirc) Thermo.

It made a convenient way to phase-shift a sine wave, by taking a sawtooth a nd shifting it up and down. There was a bit of a glitch at the retrace, bu t it was easy to filter out.

I'd never have used it in production, but it was a seriously cool part in 1

Cheers

Phil Hobbs