An idea for improved Fine Offset wind vane electronics

[rocketman_k]

Looking at my wind direction graphs I find (as others have) that the intermediate wind direction points are not often indicated. Clearly the reed switch approach to measuring the intermediate directions does not work very well.

Thinking about how to improve it I have come up with the following concept for an improved sensor that simulates the FO direction sensor so that it can plug in without any modification to the FO station at all.

The attached diagram shows the concept:
A 4 bit gray code encoder disc senses the direction via 4 phototransistors. I believe Gina has written about using Gray code so I won't cover the reasons for that choice here.

The outputs of the four phototransistors are used to address an EPROM directly. The EPROM is programmed in such a way that its data line can address a pair of Quad Analog switches (4066) directly. The analog switches connect to a resistor ladder using the same values as the existing FO sensor.

I have (for the moment) ignored power consumption issues, I want to build a prototype to prove the concept, then I will optimise the circuit for minimal power.

I'd welcome feedback on the idea.

Image1.gif

[Gina]

An interesting alternative to my 1-wire system and using the FO transmitter Looks quite feasible to me.

Good luck I look forward to seeing your results.

[rocketman_k]

An interesting alternative to my 1-wire system and using the FO transmitter Looks quite feasible to me.

Good luck I look forward to seeing your results.

Thanks for the feedback. I'm sitting here warming my fingers with the soldering iron building the prototype right now. (Spent too long rummaging around in the loft looking for specific value resistors earlier!)

Kelvin

[hans]

a very good idea.
i know nothing about programming eproms.
it should be easier and less soldering than my "old" analog electronics.

for a plan "B" you could consider using a combination of a
NOT gate with a AND gate for the 4 main wind directions

and a AND gate for the in-between directions
the tables on this page should explain it better
http://www.kpsec.freeuk.com/gates.htm#nand

[Gina]

I think an EPROM is a really neat solution to decoding the Gray code into the 16 resistor combinations required for the FO unit. To do it with separate logic gates would be pretty complicated.

[hans]

i totally agree with that,
rocketman's plan is much better,
and also smaller to build,which is important too.
for me,i would be stuck to the programming by wires,
has it's charme as well.
(i really should learn that eprom stuff,saves me lot of wires crossed )

[rocketman_k]

Thanks to Gina and Hans for your comments/encouragement. I had looked at using logic gates but decided the EPROM route was simpler.

I've got a prototype running on the bench now. It's immediately clear that the ON resistance of the analog switches needs to be fully taken into account. I had allowed for "typical" on resistances when calculating the resistor values to use in the circuit but it appears that the actual ON resistance needs to be allowed for, that may mean the use of variable resistors but I hope not! Actually the prototype uses 4016 analog switches, not the 4066 I intended (A question of which part I had in stock) The 4066 has a lower ON resistance so may be a better choice anyway. I note the original FOS sensor uses 1% resistors.

I just learned the hard way that I needed to pulse the anemometer input to get wind direction displayed......no wind = no direction.....completely logical and won't be an issue in real outdoor conditions.

For an experiment I'm going to temporarily replace my circuit with a variable resistor to try to map the behaviour of the FOS, I want to know what range of resistance value is interpreted as each direction. It seems that an inaccurate resistor value results in the direction display on the console remaining blank (no direction arrow)

I'll keep you posted.

Kelvin

[Gina]

For an experiment I'm going to temporarily replace my circuit with a variable resistor to try to map the behaviour of the FOS, I want to know what range of resistance value is interpreted as each direction. It seems that an inaccurate resistor value results in the direction display on the console remaining blank (no direction arrow)

That's very interesting - I'd assumed it would just switch from one value to the next as the resistance changed. That will be useful data to add to our knowledge base. Thank you, Kelvin

[daedalusminos]

As Gina says, this will be very interesting information, I too had assumed FO just had an ADC and interpretted the voltage produced by the resistor combinations as 16 bearing points. Not sure why they'd need to have some 'dead-band' in between.
I'll be using my ucontroller to generate the analogue voltage to enable the hall-effect sensor to interface directly to the FO unit.

[hans]

when i experimented with my vane add-on,these were the values i found,
N 33k
NNE 6.58k
NE 8.22k
ENE 0.894k
E 1k
ESE 0.69k
SE 2.2k
SSE 1.41k
S 3.9k
SSW 3.14k
SW 16k 1
WSW 14.18k
W 120k
WNW 42.1k
NW 64.9k
NNW 21.9k

maybe it will help you a little

[rocketman_k]

when i experimented with my vane add-on,these were the values i found,
maybe it will help you a little

Thanks Hans. Those values are very close to the ones I predicted from looking at the circuit so that is useful
confirmation.
I have just built an oscillator to simulate the anemometer input......that saes me from pushing a button to simulate wind
while experimenting with the direction electronics.

More to follow......
Kelvin

[rocketman_k]

when i experimented with my vane add-on,these were the values i found,

Hans

Did you measure these with the wind vane plugged in to the Transmitter or just open-circuit?

Thasnks
Kelvin

[Gina]

Thanks Hans Now if Kelvin can find the max and min resistance values for each direction we will have the full info.

[hans]

i measured open circuit and were pretty accurate with these values here which are the official ones i found somewhere.
i also remember these values in voltage for 1-wire,maybe usefull:
(on last page)
http://www.hobby-boards.com/catalog/lin ... Manual.pdf
.

[Gina]

Yes, I posted the FO resistor values as read from the cct board. Some of them are very odd values. What I'm interested in is how accurate these actually need to be.

For my 1-wire solution for the wind vane, see my posts in the 1-wire thread. Mine doesn't need a separate power supply - it uses parasitic power obtained from the 1-wire network. I'll be posting the hardware details and photos later.