The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Spehro Pefhany
Subject: Re: Humidity sensor - peculiar behavior
X-Newsreader: Forte Agent 1.92/32.572
Date: Wed, 08 Jan 2003 00:31:15 GMT
NNTP-Posting-Date: Tue, 07 Jan 2003 19:31:15 EST
On Tue, 07 Jan 2003 03:34:21 GMT, the renowned Steve Turner
>Okay, this should be titled "Eureka, I've found it!!" (I think).
>This has been one bugger of a problem to diagnose. I am posting the
>apparent solution for those who are interested. Sorry for the length,
>but I find this interesting and wanted to be clear.
>I kept coming back the time-dependent, exponential-like decay of
>voltage output which occurred only AFTER the sensor board was powered
>up. This would not be expected with components or PCB material which
>was itself humidity sensing. The effect was highly reminiscent of a
>heating effect. I now believe that this was the cause of the
>anomalous results I was seeing, but it wasn't self-heating of the
>sensor. The working theory is that the sensor was being heated by the
>power dissipation (all of about 30 mW!!!) of the onboard voltage
>regulator (which takes 12v down to 5v at a tad over 4 mA). The
>following four experiments all point to this conclusion:
Are you counting the regulator bias current in this? If not, double
it approximately, as a 78L05 draws about 4mA with no output.
>1) Feeding the board with a smaller voltage (7v, just enough to
>maintain regulation of the onboard 78L05) results in a much smaller
>voltage decay after powerup of the board. Interpretation: smaller
>power dissipation in the regulator = smaller thermal effect on sensor;
>2) Running the sensor board in a cardboard box with a small fan
>blowing air across it continuously results in a reading with some
>minor random variation but NO systematic decay of output after
>powerup. Interpretation: moving air across sensor prevents heating of
>3) Adding a 1k resistor near the sensor, which could be powered on and
>off at will with the sensor in the humidity chamber, had a dramatic
>effect on output. This small (~140 mW) heater had an effect which was
>qualitatively similar but much larger in magnitude than what I had
>been seeing. Interpretation: heating of the sensor microenvironment
>does cause a large negative shift in sensor output;
>4) Bypassing the onboard voltage regulator and feeding the board with
>5v from a bench supply gave an output which decayed only slightly
>after powerup. Interpretation: most heating was due to the voltage
>regulator, as expected. Some small drift could be due to op amp
>heating or self-heating of the sensor. In fact, results obtained with
>this configuration approach the results with the naked sensor (which
>also drifted slightly, apparently due to self-heating).
>For what it's worth, this hypersensitivity to thermal effects may be
>exactly as anticipated; I haven't done the calculations. If the
>amount of moisture in a given sample of air remains the same, heating
>the air will lower the RH.
Yes, if the air isn't stirred. Your salts should act as a buffer to
maintain the RH fixed if it is stirred, right?
The temperature effect on the actual sensor output should be MUCH
less, zero to 10% for a change of 85'C in Ta, according to the
data at: http://www.aag.com.mx/PDF%20Docs/hih3610.pdf
I have not dragged out Liptak and the psychrometric chart to see
what happens when you heat a sample of air by a few degrees, but
perhaps you should.
"it's the network..." "The Journey is the reward"
email@example.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup