Testing WVO and biodiesel for water.

[Julian]

Exactly what sensor mentioned in the post, the DHT22 (which is the white one pictured above), the HIH-4000, or the Honeywell HCH-1000-001 pictured in the 5 ^th post?

Very interesting that it's working at all.  I wonder if the results are varying because of a sort of "spot" measurement issue.

What happens if you swirl the bio around in a jar so the sensor see more of the sample?

When the cheap ones arrive, I'll try cutting one open as the plastic "cage" around the sensor may have the same effect.

[Bill]

The sensor I used is this one Honeywell HIH-5031-001. Capacitive sensor with an internal circuit to give voltage readout. Each reading the senor was immersed, gentle swirl to move some liquid over the sensor, then stabilise for 2 minutes and read. Yes it does take a couple of minutes to stabilise on a value. Readings were measured on a DVM to 3 decimal places but only taking readings to two decimal places.
The sensor also required a long immersion in 'dried' bio to get it down to the scale range for this application. Normally this sensor is used for RH measurements, measuring at ~500 ppm levels is probably right on the end of its scale range.

[Julian]

Ah sorry, I must have mis-counted the posts, I didn't see where you detailed it in the post below, my apologies.

It looks, from the data sheet, that the sensor is within a little chamber covered by a gauze, so that might explain a slower response.

[dgs]

I recieved my sensors from china, just waiting for the meter now.

Bill, do you get the bio to a stable temperature before taking the reading

[Bill]

The samples all stood together overnight so that they all came to the same temperature the next morning. It was dull morning when the measurements were made, the temperature remained constant throughout.

[Julian]

As with anything electronic, I've got this far more by luck than judgement ...




Figures refresh every two seconds and manipulation of the readings is easily achieved ... in the photo the humidity figure is halved.  I guess more complex manipulation could be achieved by someone who understands sums better than me.

The Arduino is still downloading the data to the PC via the serial link.

Approximate cost of Arduino Uno, LCD (like the unused one on the left), DHT sensor, pot and resistor probably less than a tenner. May even be cheaper with a different Arduino model.

Big question is will these sensors work in oil and bio and if so how accurately?

Even if they don't, the Arduino should be able to handle input from other sensors.

[Julian]

Better screen layout ...

[dgs]

That looks excellent Julian. Is the 47ppm reading the headspace above the bio or was this just a trial.

[Julian]

No, just a trial in ambient air by the computer.  I'm going to try the cheaper sensors in bio/oil when they arrive.

I've had a google on the internals of these sensors and it might be that the DHT11 and DHT22 are different designs, so I've put some snipes on another 4 DHT22's.

Bit worrying that the sensor says 46% humidity but the internet for our area states 88% after some rain this afternoon.  Can there be that much difference between inside and out side?  I may try taking it outside tomorrow.

I guess it doesn't matter if the sensor is "accurate", it just needs to be repeatable in oil and we can workout our own scales.

[Bill]

 Can there be that much difference between inside and out side?

Yes there can. It is also temperature dependant.

[Julian]

 Can there be that much difference between inside and out side?

Yes there can. It is also temperature dependant.

Ah, good point.

[Julian]

Having got this far, I'm going to throw caution to the wind and risk my £1 sensor in bio before the others arrive!

Bill, just so we singing as closely as possible from the same hymn sheet, exactly how did you dry your bio prior to making up the 200ppm, 500ppm, 800ppm samples?  I'll try and replicate what you've done as closely as possible.

Is using distilled as against tap water going to make any difference?  I know distilled shouldn't be as conductive.  Not sure if any of these sensors work that way.

[oakwoodtv]

Having got this far, I'm going to throw caution to the wind and risk my £1 sensor in bio before the others arrive!

You have to admire a true risk taker.

[Julian]

I know, call me reckless, but sometime I just act with gay abandon!

[Julian]

I converted my little lash up into a portable instrument by strapping on a battery pack and adding some wires!

Took it out into the shed, did the following and got these results using this setup ...




On the way to the shed I left it outside to measure ambient humidity ... 72.6% @ 10.2°C.  Figures from the internet for two towns either side of my location 74% @ 11°C and 68% @ 11°C, so not too shabby!

In the shed I made two samples ... Bio from my final settling tank = "dry" bio, and the same bio with 50% water added and shaken, free water dropping out  = "wet" bio.

Not at all scientific but enough to see if the sensor was going to play ball and shock horror it seemed to.

"Dry" bio
Reading took a while to stabilise but settled at 53.3% @ 19.4°C.  Temp check on a probe 19.8°C

Wash in methanol
Seemed like a good idea to remove the bio ... big mistake, big big mistake! reading shot straight to 99.9%.
Wash furiously in "dry" bio sample and eventually get the reading to fall.

"Dry" bio
New sample from the settling barrel in case extra methanol effected readings.  Stabilised at 57.5% @ 17.5°C.  Temp check on a probe 17.8°C

"Wet" bio
Very quickly went to 99.9% @ 17.1°C.  Temp check on a probe 17.8°C

"Dry" bio
Quick initial drop to circa 80% but then slowed, settling at 58.6% @ 16.8°C.  Temp check on a probe 17.3°C

"Wet" bio
A quick dip just to get the reading high ... up to 88%

"Dry" bio
Settled at 59% @ 16.5°C.  Temp check on a probe 16.8°C.

The "instrument" is currently sitting inside by the PC as I type and is reading a very plausible 47.9% @ 20.1°C.

For something as cheap as this I'd say that's a reasonable result.  Need to do some slightly more scientific tests now.