Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
KiwiJohn Guru Joined: 01/12/2005 Location: New ZealandPosts: 691
Posted: 06:17am 19 Oct 2006
Copy link to clipboard
Print this post
RifRaf, I have been thinking about how to actually make the coil type sensor. I already have quite a few bits on hand like small bearings and shafts etc plus the cups (which are plastic ballcock floats cut in half). There are even some coils in the box that might be OK.
There are some things I am uncertain of such as what voltage I will get from the coils and how effective the magnetic circuit will be through the shaft of the wind direction piece.
I am thinking I will start by putting four bearings into a tube (for bearing housings) with a piece of shaft right through so that all are in line then making some sort of fibreglass yoke to hold the top and bottom bearing housings and when it sets cutting out the middle section the tube, that way the two shafts should be right in line with each other. I just have to paint the house first!
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344
Posted: 09:44am 19 Oct 2006
Copy link to clipboard
Print this post
Hiya Rifraf,
Yea mate my anemotor gives 1 pulse per rev and the pic program I'm working on counts these pulses and after 1 second moves the total count to a temp register then after 60 counts the result is added to give the total revs per minute. I'm currently on working out the algorithim on showing the result as wind speed but I'm busy with heaps of other jobs first now my 3kw sinewave inverter turned up. As far as that 3422 hall effect my idea was to put say 8 3x3 neo's on a pcd and have the hall sensor mounted so it could read each magnet as it pass's now when the vane turns back the hall effect will sense this and again this could be worked out on a pic. The best idea in my opnion is you can balance the whole assembly and by using bearings the unit will be able to survive very high winds.
Cheers Bryan
RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495
Posted: 09:57am 19 Oct 2006
Copy link to clipboard
Print this post
Do you know even ROUGHLY how many RPM you're going to see?
For my genset controller, I measure both engine RPM, and output frequency (HZ). In order to do this, and since the processor is running at 8MHz and I have a couple of hardware clocks available, I generate an interrupt each time the points closes (optocoupler across the ignition coil), and twice each cycle for the mains (stepdown transformer into an AC optocoupler).
Each interrupt makes the processor jump to an ISR which grabs the current counter value, does the math to see how far it is since the last interrupt (allowing for counter wraps etc) and does a simple reciprocal to get RPM and Hz. Works really well at speeds like 100Hz (50Hz, detecting both half-cycles) and engine RPM (nominally 1883 RPM or 31.4 RPS). The benefit is, the slower things go, the more accurate it is (until you start overflowing the counter of course!)
If you want the code, just holler. It's pretty simple.
RossW
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344
Posted: 10:30am 19 Oct 2006
Copy link to clipboard
Print this post
Hiya Ross,
Eh mate I think you mistook my post I was talking about my wind speed meter not an ammeter and as soon as I can find that code I was working on which is on a backup disk somewhere I get going on it. Anway mate my latest batch of samples finally arrived so I'm on tonight finally making that data disk to go with the chips and I reckon you in for atleast a months of fun projects as I'm throwing in some other chips aswell as the pic's.
Cheers Bryan (and sorry about the delay mate )
RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495
Posted: 08:31pm 19 Oct 2006
Copy link to clipboard
Print this post
No, I definately meant "Anemometer". Counting revolutions in a second and converting to speed (m/s, km/h, whatever) isn't particularly accurate at low (relative) speeds.
The alternative is to measure time between pulses, where you can got the same result much faster.
An example: if you have a shaft rotating at something like 4 times a second (not exactly slow, but not that quick either) - and you count revs in a second, you will see only 4, right up until it goes 25% faster at 5rps.
To get a better resolution, you could measure over 10 seconds, and get measure from 4, 4.1, 4.2, 4.3 etc.
ALTERNATIVELY.... you could run your system clock through a divide-by-64 prescaler to drive a 16-bit counter at 125KHz, and measure where the counter is up to each revolution. At 4 revs/sec, the counter would only be up to 31,250.... but you could measure down to .00012 revs/sec (better than 0.01 RPM), and do it in 1/4 a second!
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
