Temperature data spikes revisited

[Gina]

as far as i know a reed needs some sort of buffering to counteract the contactbouncing.

Probably done in the firmware in the chip.

it seems to me that the trace where the capacitor,resitor,vane and thermistor are joint is the common "feed" from the chip,then via the sensor back to the chip.

Yes, that's what I think

btw ,also found this,he works with unix and the wh1080,
maybe interresting for you
http://www.lemis.com/grog/programs/WH1080.php

Thanks for that. I'll have a better look at it later.

[jim-easterbrook]

it seems to me that the trace where the capacitor,resitor,vane and thermistor are joint is the common "feed" from the chip,then via the sensor back to the chip.

Yes, that's what I think

I was thinking the same. Maybe a stabilised voltage generated within the chip? (Or a pulsed voltage for battery saving.) A poor connection to this pin would make the voltage source much higher resistance. Could this be the root of the problem?

[Gina]

Here's an improved circuit diagram...

AWS_Wind_Sensor_Cct1.png

[hans]

the thermistor does need some current all the time.
you can call it a base current if you like.
with its resitance changing and so the current dependent on temperature,it is compared with the base current.
the chips converts that current value to a temp value

now a reed or reeds opening and closing,cause a current change and peaks,like the contactpoints of a car ignition .since the vane and temp are connected by a common feed,there has to be some buffering.

i think that is the purpose of the 4.7 capasitor,(not sure though)

[Charlie]

Using an elastic band to stop the vane from moving (or the anemometer) could prove conclusively which is the source. I know they are way up a pole... what if you used the spare parts and the left over wire extension to try this while it's windy? It doesn't need any acuracy for an experiment - taping the extra bits to a post and plugging in with an extended cable would be sufficient.

My observation is that with increasing wind comes increasing spikes - yet my modified wind vane actually gets more stable as the wind increases.

Thanks for chasing this mystery! We are all learning a lot from your efforts.

[Gina]

Thank you Charlie I'm going to do some tests using the spare kit. I can do this indoors. No wind is the normal state, of course, and I can turn the vane where I want. For wind, a small fan will drive the anemometer round. To make up an extension cable (as you suggest) I shall probably raid an old telephone for the RJ11 socket to avoid messing with the wind vane - I'll see...

I'm going to see if I can get the two stations working independently in spite of being on the same radio frequency. We can but hope

[Charlie]

I think I've found some more clues into the circuit. If you go to Microchip's website (the maker of PIC microcontrollers) and have a look at AN512 and AN685, there are axamples of similar circuits.

[Gina]

I think I've found some more clues into the circuit. If you go to Microchip's website (the maker of PIC microcontrollers) and have a look at AN512 and AN685, there are axamples of similar circuits.

Thanks for that The AN512 article confirms how I thought it worked. except I hadn't gone as far as the calibration part - removing variations in timing capacitance by calibrating against a known resistance. Using pulse mode excitation of the thermistor overcomes the self-heating problem of using DC. A current of just 20uA was mentioned for the DC mode. That would make the circuit extremely sensitive to interference. Also using pulse mode and timing the charge (or discharge) provides a simpler method when using PICs, in addition to avoiding self-heating (no current most of the time then a short pulse of relatively high current to make the measurement).

I've been thinking of making up separate digital thermometers for taking other measurements such as ground and soil temperatures. I'm very interested in methods of measuring various aspects of our environment and instrumentation in general. I was lucky to find a radiosond unit floating in the sea many years ago when I went sailing. I studied the ultra simple way it measured height and temperature (mechanically) and used an anemometer like device to turn an encoder which controlled the radio transmitter. Fascinating!

[Gina]

I now have my spare tx unit powered up from my bench supply and running on 3v. No sensors plugged in yet. It reacted correctly to applied power - long flash followed by short flashes when transmitting - every 48 secs. It is NOT interfering with the other, outdoor transmitter. Both consoles are still locked onto the outdoor transmission. Now, if I can get the test console to latch on to the test transmitter...

Later... YES! One console on each transmission. Normal weather station web site showing outdoor conditions (very wet and windy) - testing page running from 2nd console which is now showing the indoor temperature, no wind and no rain. The testing page is now showing the test transmitter data. It took two goes to get the 2nd console to latch onto the right transmission - I guess it depend on which it picks up first.

[Charlie]

I've been thinking of making up separate digital thermometers for taking other measurements such as ground and soil temperatures. I'm very interested in methods of measuring various aspects of our environment and instrumentation in general.

I'd recommend using the DS18B20 1-wire digital thermometer devices. The TO-92 package is nice and tiny and easily potted, they can be parasitic powered so you only need to run 2 conductors to them, and a large number can be put on the same "bus", so you'd only need to run one pair of wires. They are fairly inexpensive, 0.5 C acuracy, and considerably higher resolution. (0.0625 C)
You can build a master and display unit from a PIC micro (or similar) in a couple hours, and include a serial interface to get the data into your computer too.

[Gina]

That's interesting Just a matter of where to buy them A 1-wire system certainly has attractions.

[nking]

Later... YES! One console on each transmission. Normal weather station web site showing outdoor conditions (very wet and windy) - testing page running from 2nd console which is now showing the indoor temperature, no wind and no rain. The testing page is now showing the test transmitter data. It took two goes to get the 2nd console to latch onto the right transmission - I guess it depend on which it picks up first.

That's really useful to know that you can run two WH1081 weather stations completely separately from the same location. I’d wanted to test out my spare station but as I thought it might cause a conflict/data corruption with the existing unit I didn’t chance it.

[Gina]

Yes. I wondered if it would work when seeing that the console has to recognise the signal the "lock on" to it. I wonder if these stations have a unique code in the transmission or whether, once it's found the signal it only looks again 48 seconds later. Either way it's a brownie point to Fine Offset

[mcrossley]

I wonder if these stations have a unique code in the transmission or whether, once it's found the signal it only looks again 48 seconds later.

I think it only looks every 48 seconds (or so), I suspect this is why I have sometimes lost contact with the transmitter, and it takes an hour or so for the window to drift back into synch with the transmitter. Maybe wrong about this though.

[Gina]

Sounds very likely to be the correct synopsis, Mark.