A project that some might be interested in...

F

Fowkc

Senior Member
http://www.fowkc.com/gliding/gmeter.shtml

I'm building a PICAXE based G-meter for a glider in our club (a gliding club, for the slow among you). I've no idea how it will output the G-reading yet. It might be an audible output based on frequency at the request of one of our instructors, but I've not much idea how to attempt that.
 
A

andrewpro

New Member
Well, more specific details would be a place to start. Another thing to get a good response would be not calling people slow in your opening sentence.

--Andy P
 
F

Fowkc

Senior Member
Well I hope people will realise the "slow" thing was a joke...

I wasn't really asking for help on the project, just seeing if anyone is remotely interested.

If it's an audible output, it'll likely be similar to the instrument in gliders called a variometer. Loosely, it tells you how fast you're climbing or descending. Most (if not all) have an audio output so the pilot can keep his eyes on the sky rather than on the instruments (and hence see things he might otherwise hit). If the glider is going down, the output is a constant tone, with frequency dropping as descent rate increases. If the glider is going up, the output is a pulsed tone, with frequency rising as ascent rate increases. Shouldn't be that hard to do, but I haven't really started thinking about it.
 
hippy

hippy

Ex-Staff (retired)
I'll have to let the "slow" comment pass, as I admit to missing the URL as the first line of your post :)

Sounds interesting, but then I find most projects interesting. And as projects invariably cause some random thought to enter my head, although not applicable in this case ( and a $22 USD accelerometer probably can't be beaten ), has anyone ever tried measuring G with a pot as a pendulum; a long stick with a weight on the end ?
 
A

andrewpro

New Member
Sorry, My mistake. Most of the other groups and lists I belong to are rather harsh, to say the least, and these are supposed ot be the "professionals". I just saw it and immediately made the, amittedly wonrg, connection. I'm glad to see that this board hasn't stopped to that level (which is what I was affraid of!).

As for the pots on a stick: I've used them in cars. I had a gimbal setup, with the pots on the axles of the gimbal, with a weighted rod, and used it to track the directions of force throught a pass on a track. The gimbal was damped, to an extent, but I still had ot go through and filter out the overshoot.

I dont see it working in a three dimensional plane, though (no pun intended!!). The banking, climbing, and diving of the aircraft would pretty much make it useless, as far as I can guess anyways.

--Andy P
 
M

Michael 2727

Senior Member
Oh N0 !, Sebastiaan is in OZ.
Makes note to pay up insurance.
Which City, Sebastiaan, and what time do you
leave for work in the morning, and what time
do you knock Off,?
 
F

Fowkc

Senior Member
"How does the pendulum method distinguish between accellerating on a level road and going at a steady speed up hill?"

Essentially it doesn't. The accellerometer I have is based on the changing strain of silicon wafers inside the chip and has the same problem. To overcome it, there's actually two sensors at right-angle to each other. When properly mounted, I can use the reading from one to determine the attitude of the aircraft (I hope).

I think I'm going to build a datalogger with some EEPROMS and take the thing on some aerobatic adventures to see what happens.

 
J

Jeremy Leach

Senior Member
Well, another idea is to use a material called Quantum Tunnelling Composite. This is cheap, varies it's resistance with applied pressure. I've not used it yet myself but am interested.

In theory all you'd need is a weight, a piece of this stuff in a resistor divider into an ADC of a Picaxe08M. Then one output of the PicAxe08m driving a piezo speaker with appropriate noises. Could be very cheap.

Big question is though - how sensitive would it be ?

PS: Great project. I've long been interested in gliders, although the radio-controlled kind. Will be fun calibrating. I'd suggest some sort of rotating turntable with the device strapped on. If you know mas, and speed of rotation then can work out G.

Edited by - Jeremy Leach on 1/11/2006 11:48:07 AM
 
F

Fowkc

Senior Member
Hey! Using QTC is a brilliant idea! I work part-time at Maplin here in the UK, we sell the little pills and the coax-cables of QTC.

I'm heading off to work in half an hour, I'll buy some.
 
Y

ylp88

Senior Member
I think you also need the radius of rotation to work out the centripetal force (and the resultant "centifugal" reaction force), and thus, g.

f = ma
f = (mv^2)/r
ma = (mv^2)/r
g = a = (mv^2)/(mr)

Something like that... mass will be constant, velocity could possibly be measured, radius... hmmm... not sure...

<b><i>ylp88 </b> </i>
 
J

Jeremy Leach

Senior Member
That's it ylp88 ;-) ... which reduces to V^2/R.

(By the way I've got this great little book ISBN 0748700773, &quot;Tables, Data and Formulae for Engineers and Mathematicians&quot; by A.Greer and D.J Hancox. A lot of the time I never remeber these things but this books great as it's got all the useful formulae in one place )
 
J

Jeremy Leach

Senior Member
Actually, thinking more about this, it may be even better to use a sandwich approach ...

+ve
|
|
QTC
|
|---&gt; to ADC input
|
QTC
|
|
GND

And the weight sandwiched between the QTC sheets/pills whatever.

More sensitive than just using a single piece of QTC. Would record deceleration as well as acceleration.

Edited by - Jeremy Leach on 1/11/2006 1:29:44 PM
 
Y

ylp88

Senior Member
This QTC stuff sounds really interesting. Is it available in Australia? It also reminds me of the &quot;Smart Wire&quot; in the PICAXE datsheets. Is this available in Australia?

I have a book we used for the last two years in year 11 and 12 with all sorts of mathematical formuals (it is the things you are allowed into all exams, including the TEE) - &quot;Mathematical Formulae and Statistical Tables Book&quot; ISBN 0 7309 4663 0. We have to refer to separate Chemistry and Physics datasheets for those kinds of stuff, though. Your book sounds interesting - I'll keep and eye out for it!

<b><i>ylp88 </b> </i>
 
F

Fowkc

Senior Member
Got the pills, I don't have time to experiment with them now (knee deep in revision).

I think if I use them it will be somehow in conjunction with my existing accellerometer.

Anyway, glad to see that ideas are abound, no doubt I will be asking for lots of advice futher down the line.
 
S

steirny

Member
Thought I&#8217;d throw in a possible simple calibration idea.
The professional accelerometer chips used in medical equipment and guided missiles etc can be static calibrated on the bench at +/- 1G, simply by flipping them over. They are usually linear over their range, therefore higher G&#8217;s can be extrapolated. I don&#8217;t know how many G&#8217;s you guys can handle before the wings fall off or the pilot blacks out, but the errors wont matter then.
The QTC approach is more than likely non-linear but you could still try the static calibration approach. Determine the +/- 1 G value, then apply double, triple, quadruple your normal working mass for the on bench calibration for +/- 2, 3, 4 G etc. If its nonlinear you could do a few points and calculate a curve, or you could shape the mass to have an inversely nonlinear effect on the material.
If you can see holes in this, Gee feel free to shoot me down.
 
J

Jeremy Leach

Senior Member
Re the rotating disc approach to calibrating it :

Could be a bit tricky to actually read the results as it's rotating. However had some thoughts...

The PicAxe Datalogger kit would be ideal for initial calibration (in fact I use it as an initial test bed for lots of ideas). Strap the whole datalogger to the disc. Make the QTC setup inteface to an ADC input. Start logging, rotate the disc at different speeds at set times. Remove logger and extract results.

For &quot;different speeds at set times&quot; ... ok, sounds like you need to also make a rev counter of some sort.

(However, another simple approach to setting a disc at constant speed I messed about with years ago is by simply marking out black and white sectors and rotating under a normal flourescent light , the 50Hz flicker from the light is effectively like a strobe and when the speed is right the pattern freezes. I just used to use a simple motor and a pot to vary the speed manually. Can have different number of sectors for different speeds. I found it good fun at the time ! )

Finally ....yes, I expect the QTC approach would be non-linear, but with a good set of test readings and a lookup table you'd be fine.

 
F

Fowkc

Senior Member
Calibration shouldn't be too much of an issue. I used the values from the datasheet (Output is Vcc/2 at 0g, with +/- 312mV per g above and below), and used it to make the graph here:
http://www.fowkc.com/gliding/gmeter/GTestGraph.JPG

The values are quoted as +/- 4% in the most extreme cases, which is far more accuacy than I'll actually need.

Oh, and the data on QTC pills is here:
http://www.peratech.co.uk/pills.htm

From a first look, the logarithmic response of QTC makes PICAXE processing difficult, unless I use the co-processor or use a simplification of the relationship.

Doesn't look like you can get the stuff in Oz, at least not from Peratech distributors they have listed.

Edited by - Fowkc on 1/12/2006 11:45:46 AM
 
B

BeanieBots

Moderator
The conduction of a diode is very close to logrithmic. If you use it in the feedback loop of an op-amp, you get a log response. If you use it in the input, you get anti-log.
Unfortunatley, they are also rather temperature sensative but if you don't need great accuracy and the temperature range won't change more than a few degrees, it could save you a lot of coding effort to make the response linear and get the most out of the ADC resolution.
 
J

Jeremy Leach

Senior Member
That's really interesting BeanieBots ..and a good idea. I'll have to do some browsing to see some typical circuits. I wonder if there's a simple way to compensate for temperature effects?
 
J

Jeremy Leach

Senior Member
A bit late ....but regarding the talk of QTC and availability: Apparently you can use the conductive foam that ICs come in. Well that's what my Forrest Mims book describes. Thought I'd mention this if anyone was wanting to try it but couldn't get any.
 
You must log in or register to reply here.
Top