Humidity sensor and AC output

Hi I have a TruSens humidity sensor bought from Rapid and I have discovered that it requires AC power and will not run on DC power which the PICAXE output. I hadn't intended to use the mains or a motor and dynamo so does anyone know how to do it with a PICAXE. I have searched around and haven't found out much.

Someone said in another thread about usin PWM and pulseout could create one but I wondered if anyone had done this before and has any code they could help me with.

Thanks,
George S.

I wonder whether pulsed DC would do, or if it needs reversing polarity (ie an H bridge).

For pulsed DC, a 555 would be best. For proper AC, you would need an H bridge. Both of these suggestions assume it will work with a square wave - not a sin wave.

Andrew

Edit: Looking at the datasheet, it needs fast AC (0.5 to 2kHz), and a sin wave would probably be needed. This may be too fast for a picaxe.

That sucks, what about some sort of ic that can be read by serial or i2c to produce an output, or an external ic to produce this fast sine wave.

EDIT: Our buddy maxim seems to know: http://www.maxim-ic.com/appnotes.cfm/an_pk/2081

I also found this, http://www.redcircuits.com//Page13.htm, I may build these and see what happens.
Then I take it the circuit is:
~1kHz Wave 5V ~=SENSOR=~~|TO INPUT|~~~1kHz Wave 0V

What current does the unit use?

If its only a few milliamps, then you can use the sound command to generate a 1Khz square wave.

Invert the square wave with an inverter IC and tap the power off the inverted signal and the original signal. This will give you a 5V peak to peak AC supply.

Run through a non polarised resistor capacitor circuit to smooth out the square wave into a more or less sine wave.

The PICAXE is perfectly capable of producing the desired frequency using PWM. If a sinewave REALLY IS required, then simply filter with a second order RC network. I've not read the datasheet so have no idea about the exact requirements but you may need to buffer the post RC sinewave with an amp such as an op-amp.
Having given it the required sinewave, how will you read the (presumably) sinusoidal signal it gives off?

Haxby said:

Invert the square wave with an inverter IC and tap the power off the inverted signal and the original signal. This will give you a 5V peak to peak AC supply.

Click to expand...

That's 10v peak to peak.
Voltages are relative. A signal changing between 0v and 5v is AC 5v peak to peak.

George Sephton said:

Hi I have a TruSens humidity sensor bought from Rapid and I have discovered that it requires AC power

Click to expand...

I'm guessing this is the cheapo 50p HCZ-J3A HUMIDITY SENSOR (61-0984) device. If I get an order into Rapid ( not sure when ) I'm going to order a few and have a play. I reckon there may be some way to use it easily and simply even if not driven under recommended conditions and, at that price, limited lifetime may not be a problem, for example, don't run it continuously. As to any accuracy, I have no idea.

With these sort of components it's often a case of busk it and see, wire something up, see what happens, and if it works everyone benefits, if it catches fire, emits a bang, melts a PICAXE or creates a black hole which consumes the universe, mark it up as part of the fun of pushing the frontiers of bodging things

Most of my electronics is done for fun and I get enjoyment from hacking things and nudging at the boundaries of 'maybe it will work' but that's not to everyone's taste. Unless you can find the time to do the experimenting yourself and put up with the trials and tribulations of something which ultimately may not work the best solution is often to get something which is plug and go or is tried and tested and known to work, although that usually is a more expensive route.

Unfortunately there are a lot of things out there which 'could be used' but no one's ever done that, so it's often do it yourself, pay the cost of someone having done the work or hope somene gets round to having discovered how to use it before you need to.

http://www.rapidonline.com/netalogue/specs/61-0984.pdf is the datasheet.

I agree with hippy Im not doing all this for gain, I really dont need to know how humid my room is but it's a lot of fun and it doesn't matter what it costs. Thanks very much about ordering them and yes it is the cheapo one, lol.
Im not sure but maybe it'd work with a square wave as that would be helpful, would it matter too much, the only time ive come across sine waves are in maths so im not sure how its all used in electronics.

But yeah if its a limited time then it would have to be run maybe once a day instead over every 10 seconds like i will with the LDR and temperature sensor. (part of my project)

The datasheet does not give any warning about DC levels. For the cost & risk, I'd experiment with DC and see what happens.
The figures quoted are for 1v rms sinewave. Note the 1V MAX rating, so don't go sticking 5v into it!
Creating a method of reading the AC impedance will be much harder than generating a nice clean 1v rms sinewave.
If you want to try to do it "properly", lookup "precision rectifier" in google or whatever. Not too tricky to make but will require a few op-amps. Then it's a question of building an AC driven potential divider with the sensor as one limb and using the "precision rectifier" to measure the AC volts across it.
It is also VERY temperature dependant so you will need to compensate.
Not very accurate due to the 2% hysterisis giving overall accuracy (AFTER compensation) of +/- 5%.

The moral here is "you get what you pay for" & "read the spec before you buy".
The golden rule is, if there is no datasheet, DON'T BUY.

It is not too difficult to build a sine wave generator from a cheap dual op amp. Easy to find circuits with a Google search. Here's one:
http://www.play-hookey.com/analog/sine_wave_generator.html
An additional op amp could be used to adjust the amplitude to the 1V rms required.
I believe that there are also function generator IC's which would make it even easier.

My thoughts were to build an oscillator so the humidity sensor had AC through it and changed the frequency, or somehow use it as an RC from a couple of PICAXE pins and time the charge/discharge time. Triangle is close to a sine wave if you squint hard enough

Yet another sine wave generator circuit.

I think you can get a reasonable result quite simply. The datasheet defines the output as impedance vs humidity. I don't think the waveform has to be perfect to measure that, and it's not a very accurate device anyway - see the temperature dependency. I would use the picaxe PWM output to generate a squarewave, put that through two or three stages of RC filtering to get it close to a sine wave and apply it to the sensor (one side grounded) through another resistor. Connect the junction of the resistor and sensor to a simple diode/capacitor detector and feed the resulting dc voltage to the picaxe ADC input.

BeanieBots said:

The datasheet does not give any warning about DC levels. For the cost & risk, I'd experiment with DC and see what happens.

Click to expand...

Bottom-right of page 2 in the datasheet:

Caution remarks on operation:
• To avoid direct application of DC voltage on humidity
sensor.

Click to expand...

Ken

Well I have ordered a square wave generator from maxim so if it ever arrives I'll follow one of those circuits to create a sine wave then pass an ac current through the humidity sensor which can then be read by the PICAXE. I'll post my circuit here when I get it working.
Thanks everybody.

@ KMoffett, thanks, I missed that.
@ George Sephton, there's a perfectly good square wave generator in the PICAXE! (PWMout).
I'd also slap a cap in series just in case the PWM duty is not exactly 50%.
Then go with LizzieB's suggestion as the sensor accuracy does not warrant much else.

there's a perfectly good square wave generator in the PICAXE! (PWMout).

Click to expand...

I was aware of that but I've run out of pins on my PICAXE, if I have any I will use PWMOUT instead as it is easier. What would the code be to produce a square wave, because I can then use that to convert it into a sine wave.

One thing to watch out for with PWMOUT is it has a lower limit of frequency which is quite high, but this can be reduced by under-clocking the PICAXE or by poking various internal pre-scalers; more info via Forum Search.

With a lower limit of 0.5Hz, it would also be an option to use high/low statements but the use of a DC blocking cap would be more essential to gaurantee no DC component.

It would also give the option of synchronising the ReadADC command with the "AC" and might be a way of avoiding the need for a detector circuit.
Certainly worth a play.

do
High H_drive
pause N
readadc H_sense
{other code}
Low H_Drive
Pause (N+time for {other code})
loop

From the datasheet:
Impedance range at 60%RH and 25°C (1kHz, 1Vrms) 14.2 – 38.5 kΩ

That's a wide range; if you need an absolute measurement, you'll need a way to calibrate it. If you just need to know if it is more humid today than it was yesterday, then your job is easy.

Most items run happily on pseudo-AC, hence the proliferation of 12>240v inverters using it, cant see an issue using the axe to generate it?

Coyoteboy said:

Most items run happily on pseudo-AC, hence the proliferation of 12>240v inverters using it, cant see an issue using the axe to generate it?

Click to expand...

Do what??
No relevance whatsoever!

This is a capacitive-type sensor. Its capacitance (and thus, its impedance) changes with humidity, and somewhat also with temperature.
Therefore, frequency accuracy is another factor you should consider. I believe the Picaxe's internal clock is far more accurate than any external RC based oscillator.

That is what I would do:

1) Use the Picaxe's PWM output. Also use 1 Khz, since the device is characterized at that frequency.

2) The PWM wizard indicates that these values are outside range. What you could do is use a CD4013 as a divide by 4, which will also provide a perfectly symetrical square wave. The PWM frequency would be now 4 Khz. From the wizard, this equals to pwmout 2 , 249, 500

3) Filter it with a 2 pole RC filter. Buffer the signal with an opamp before feeding it to the sensor.

4) The sensor will form the upper leg of a voltage divider, the bottom leg being a resistor whose value is halfway between the sensor's upper and lower resistance range.

5) Rectifiy the signal with a precision rectifier, and feed it to the PICAXE's ADC.

That is the hardware part, the software follows.

6) Since the device has a lot of temperature dependency, it would be worthwhile to include a temperature sensor and perform some temperature correction.

7) Also the device is highly non-linear. Some sort of curve fitting correction would be nnecessary. I don't know how to do this with the integer math. Maybe other gurus know.

All in all, this is a handful, and looks like a very challenging (read: fun) project. If you need help with the analog portion of the circuits, feel free to ask.

If its a capacitance sensor why not use it as the timing capacitor in a 555 oscillator circuit then use the picaxe to measure that frequency?

Jaguarjoe said:

If its a capacitance sensor why not use it as the timing capacitor in a 555 oscillator circuit then use the picaxe to measure that frequency?

Click to expand...

Doing a Google search, I found that there is an additional RH sensor technology, which I was previously not aware of: resistive humidity sensor.

According to the information in the link below: "...Most resistive sensors use symmetrical AC excitation voltage with no DC bias to prevent polarization of the sensor. The resulting current flow is converted and rectified to a DC voltage signal for additional scaling, amplification, linearization, or A/DRconversion..."

Therefore, this sensor must be a resistive, not a capacitive one.

http://www.sensorsmag.com/sensors/Technology+Tutorials/Sensors/Humidity/Moisture/Choosing-a-Humidity-Sensor-A-Review-of-Three-Techn/ArticleStandard/Article/detail/322590

Jaguarjoe said:

If its a capacitance sensor why not use it as the timing capacitor in a 555 oscillator circuit then use the picaxe to measure that frequency?

Click to expand...

The datasheet clearly stated not apply DC directly to that sensor. The timing capacitor in a 555 oscillator is continuously at a varying DC voltage. Pulsating DC is not AC.

Ken

Doing yet another Google search, found the following circuit for a capacitive type sensor, a Humirel HS1101LF, where the 555 idea could be put to good use... Please note this is different from the one originally discussed in the thread.
This particular sensor is available from DigiKey.

While "do not apply DC" is there for a reason, pulsed DC isn't the same as AC, a square wave isn't a sine wave, the real question is what effect will those have and how large will the effects be ? That can probably only be determined by experimenting.

Considering these sensors are cheap and stocked by Rapid it suggests they mey be a component in widespread use in cheap consumer product, desktop hygrometers and all-in weather-station-come-alarm-clocks. Maybe a place to look is where have these been used and see what the designers have done there. We may be imagining it's a lot more complicated than it actually is.