I want to do a humidity sensing application, using either a G-Cap (General
Eastern Instruments) or Smartec capacitive humidity sensor. There are
diiferent ways to interface these to a micro processor, of which two are
options for me to interface it to a 12c... Pic. One method is to use the
sensor as the capacitive element in a RC network for a 555 oscillator, then
read the resultant frequency and translate it to humidity. Other is to use
an AC bridge (excitation supplied by Pic), with some amplification and then
into ADC onboard Pic.
The 555 option seems to me most cost effective, and easiest, but I'd like to
hear if anybody else did the same type of experiment and what there findings
were?
Tom, while I used Honeywell/HyCal calibrated sensors with a voltage
output in my PIC-based weather station, I've since looked at the G-CAP
family and the related App note using a 555 Timer. I would recommend that
route since it's simple, low-cost, and you just count pulses.
- Tom
At 03:50 PM 4/13/00 +0200, TOM THERON wrote: {Quote hidden}
>Hi all,
>
>I want to do a humidity sensing application, using either a G-Cap (General
>Eastern Instruments) or Smartec capacitive humidity sensor. There are
>diiferent ways to interface these to a micro processor, of which two are
>options for me to interface it to a 12c... Pic. One method is to use the
>sensor as the capacitive element in a RC network for a 555 oscillator, then
>read the resultant frequency and translate it to humidity. Other is to use
>an AC bridge (excitation supplied by Pic), with some amplification and then
>into ADC onboard Pic.
>
>The 555 option seems to me most cost effective, and easiest, but I'd like to
>hear if anybody else did the same type of experiment and what there findings
>were?
>
>Regards
>Tom Theron
------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)
I've used the second method to measure tilt with a capacitive tilt meter and it
worked nicely. Can't say how it would do with the humidity sensor, but the
technique works fine. I take 128 samples and throw away the lowest few bits.
I want to do a humidity sensing application, using either a G-Cap (General
Eastern Instruments) or Smartec capacitive humidity sensor. There are
diiferent ways to interface these to a micro processor, of which two are
options for me to interface it to a 12c... Pic. One method is to use the
sensor as the capacitive element in a RC network for a 555 oscillator, then
read the resultant frequency and translate it to humidity. Other is to use
an AC bridge (excitation supplied by Pic), with some amplification and then
into ADC onboard Pic.
The 555 option seems to me most cost effective, and easiest, but I'd like to
hear if anybody else did the same type of experiment and what there findings
were?
I've used the second method to measure tilt with a capacitive tilt meter and
it
worked nicely. Can't say how it would do with the humidity sensor, but the
technique works fine. I take 128 samples and throw away the lowest few
bits.
Andy
I am also working on a capactive tilt meter, with the same idea as yours -
don't know what sensors you use - AOSI?
Anyway, I think you get better temp stabilization with the bridge, since
with the 555/RC method the R component is influenced by temperature.
However, in my application the device will be used in a temperature
stabilized room, that way it won't matter that much.
I'm not sure which brand (AOSI sounds familiar but I'm not 100%), I just got the
sample from my customer and instructions to "make it work." That it does! You
can see it at
A resistor, especially a 1%, doesn't change much across most temperatures. If
you get the R and C balanced correctly, the tempco's will automatically offset
one another.
I've used the second method to measure tilt with a capacitive tilt meter and
it
worked nicely. Can't say how it would do with the humidity sensor, but the
technique works fine. I take 128 samples and throw away the lowest few
bits.
Andy
I am also working on a capactive tilt meter, with the same idea as yours -
don't know what sensors you use - AOSI?
Anyway, I think you get better temp stabilization with the bridge, since
with the 555/RC method the R component is influenced by temperature.
However, in my application the device will be used in a temperature
stabilized room, that way it won't matter that much.
> I am also working on a capactive tilt meter, with the same idea as yours -
> don't know what sensors you use - AOSI?
You guys should check out the Analog ADXL202 - its a two-axis 2g
accelerometer. One chip, 1/4" square and about 1/8" thick, $20 in
single quantities, and is trivially easy to interface to a PIC (I've
got one interfaced to a F84). You can also get an eval board for $30
that has the chip mounted, you need to add two caps and a resistor,
and you get a standard 5-pin 0.1" header.
It gives you basically 0-90 degree tilt accuracy in both the X and Y
axis. With two of them, you can get full 360 degree tilt accuracy.
Its very cool tilting this board in your hand, and watching a servo
move to a position that correspends with the tilt...
Already evaluated the ADXL202. Works great, but too expensive for the
application.
It still needed a micro on it to provide zero offset, gain, and remote control,
so for double the price we would have gained nothing. The capacitive sensor
also had more data precision in less time.
I'm using an ADXL105 to provide a pitch stabilizer (rate gyro setup) for one of
my model boats. SHould be an interesting thing. The ADXL202 is going to drive
the prototype directly until I get the 105 board done.
> I am also working on a capactive tilt meter, with the same idea as yours -
> don't know what sensors you use - AOSI?
You guys should check out the Analog ADXL202 - its a two-axis 2g
accelerometer. One chip, 1/4" square and about 1/8" thick, $20 in
single quantities, and is trivially easy to interface to a PIC (I've
got one interfaced to a F84). You can also get an eval board for $30
that has the chip mounted, you need to add two caps and a resistor,
and you get a standard 5-pin 0.1" header.
It gives you basically 0-90 degree tilt accuracy in both the X and Y
axis. With two of them, you can get full 360 degree tilt accuracy.
Its very cool tilting this board in your hand, and watching a servo
move to a position that correspends with the tilt...
<x-flowed>One of the electronics magazines that does the design ideas had a neat
solution to AC signal conditioning using a PIC--even included code. I
think it even ended up in PICs app-notes somewhere. They were using it to
condition a tilt sensor from aFrericks (not of Hollywood). I've had a fair
amount of experience along these lines if you want to get in touch off list.
Kelly
>I've used the second method to measure tilt with a capacitive tilt meter and
>it
>worked nicely. Can't say how it would do with the humidity sensor, but the
>technique works fine. I take 128 samples and throw away the lowest few
>bits.
>
>Andy
>
>I am also working on a capactive tilt meter, with the same idea as yours -
>don't know what sensors you use - AOSI?
>
>Anyway, I think you get better temp stabilization with the bridge, since
>with the 555/RC method the R component is influenced by temperature.
>However, in my application the device will be used in a temperature
>stabilized room, that way it won't matter that much.
>
>Tom
ICON.CO.ZA>