Humidity HS1101 to %RH (more integer math help) - Electronic Circuits Projects Diagrams Free

blueroomelectronics · 27th September 2007, 03:18 AM

Thanks to all for the LM34 math help (works like a charm)
Now any help with a simple table fro the HS1101 humidity sensor datasheet.
Here's the table
%RH frequency in Hz
0 7351
10 7224
20 7100
30 6976
40 6853
50 6728
60 6468
80 6330
90 6186
100 6033

The real trick will be temperature to HS1101 compensation, but that's beyond the scope of the tests.

Pommie · 27th September 2007, 07:21 AM

I was going to suggest storing the table and using linear interpolation to get inbetween values. However, your data appears to jump in the middle - the values increase by 130 (ish) but between RH 50 - 60 they jump 260. I found a data sheet but the table in that is completely different. Unless your short on program space it would be simpler to store 100 values than try to interpolate.

Mike.

ericgibbs · 27th September 2007, 07:45 AM

hi Bill,

The 70 value is missing.

If you subtract 6033 from your reading and plot the result the slope is almost linear, wouldn't it be easy to calculate the RH% from the Freq rather than a table?.

EDITED: Plot added.

Pommie · 27th September 2007, 07:47 AM

Doh, didn't spot that.

Mike.

ericgibbs · 27th September 2007, 10:19 AM

Bill,
Like Mike I see the datasheet gives different values from the list you have published, this versions not linear

RH% Freq
0
5
10 945
15 870
20 800
25 735
30 670
35 610
40 550
45 495
50 440
55 390
60 340
65 290
70 240
75 190
80 145
85 95
90 50
95 0
100

These values have 6210 subtracted [ ie:95%RH]

eng1 · 27th September 2007, 11:06 AM

Quote:

Originally Posted by blueroomelectronics

Here's the table
%RH frequency in Hz
0 7351
10 7224
20 7100
30 6976
40 6853
50 6728
60 6468
80 6330
90 6186
100 6033

The relation between RH and f is almost linear, as Eric said.

A good approximation I've found is:

f = -13.2*RH + 7358

RH = -0.0753*f + 554

According to the first formula, for example, RH=70 => f=6434.

Knowing that the slope is -.0753 is useful even if you calculate
RH from a small set of data stored into a look-up table.

ericgibbs · 27th September 2007, 12:58 PM

hi Bill,
This table is from our datasheet.
Extrapolated for table top/bottom end values.

EDITED: now a gif.

You can see the right hand column is very close to the actual RH%

Depends how accurate you require the RH%

blueroomelectronics · 27th September 2007, 03:42 PM

Wow thanks again, expect Cricket PCBs in the mail.

Now the schematic and testing are 99.9% done (I'll post the revised schematic on my site), it's on to the layout. Same size as the Inchworm, Firefly & Dragonfly PCBs 107mm x 57mm (Hammond 1591B case)

Design thoughts, the Cricket needs to generate as little of its own heat as possible. Slow running PIC, latching relay, low baud rates, LCD display, parts location (the LM34 is in the lower left of the PCB)
There is a provision for 3xAA Alkaline for running during power fail (or running sans RS485)
So the problem is the main power source.

A 5V switch mode wall wart (they are cheap now) this is the least expensive route and will keep the kit price down

A LM2575 or other low power switching regulator (can handle up to 40V so 24VAC HVAC can be bridged on to it) $$

A "Roman Black" switching 5V supply (not tested)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Latest
Fahrenheit to Celsius (Integer math for a PIC) help wanted.	blueroomelectronics	Math and Physics	7	3rd October 2007 02:58 PM

27th September 2007, 03:18 AM	(permalink)
blueroomelectronics	Humidity HS1101 to %RH (more integer math help) Thanks to all for the LM34 math help (works like a charm) Now any help with a simple table fro the HS1101 humidity sensor datasheet. Here's the table %RH frequency in Hz 0 7351 10 7224 20 7100 30 6976 40 6853 50 6728 60 6468 80 6330 90 6186 100 6033 The real trick will be temperature to HS1101 compensation, but that's beyond the scope of the tests. __________________ Bill Smart Kits build Smart People http://www.blueroomelectronics.com

27th September 2007, 07:21 AM	(permalink)
Pommie	I was going to suggest storing the table and using linear interpolation to get inbetween values. However, your data appears to jump in the middle - the values increase by 130 (ish) but between RH 50 - 60 they jump 260. I found a data sheet but the table in that is completely different. Unless your short on program space it would be simpler to store 100 values than try to interpolate. Mike.

27th September 2007, 07:45 AM	(permalink)
ericgibbs	hi Bill, The 70 value is missing. If you subtract 6033 from your reading and plot the result the slope is almost linear, wouldn't it be easy to calculate the RH% from the Freq rather than a table?. EDITED: Plot added. __________________ Eric "Good enough is Perfect" PIC tutorials: Gramo's: www.digital-diy.net/ Bill's: www.blueroomelectronics.com/ Last edited by ericgibbs; 7th July 2008 at 11:21 AM.

27th September 2007, 07:47 AM	(permalink)
Pommie	Doh, didn't spot that. Mike.

27th September 2007, 10:19 AM	(permalink)
ericgibbs	Bill, Like Mike I see the datasheet gives different values from the list you have published, this versions not linear RH% Freq 0 5 10 945 15 870 20 800 25 735 30 670 35 610 40 550 45 495 50 440 55 390 60 340 65 290 70 240 75 190 80 145 85 95 90 50 95 0 100 These values have 6210 subtracted [ ie:95%RH] __________________ Eric "Good enough is Perfect" PIC tutorials: Gramo's: www.digital-diy.net/ Bill's: www.blueroomelectronics.com/ Last edited by ericgibbs; 7th July 2008 at 11:21 AM.

27th September 2007, 12:58 PM	(permalink)
ericgibbs	hi Bill, This table is from our datasheet. Extrapolated for table top/bottom end values. EDITED: now a gif. You can see the right hand column is very close to the actual RH% Depends how accurate you require the RH% __________________ Eric "Good enough is Perfect" PIC tutorials: Gramo's: www.digital-diy.net/ Bill's: www.blueroomelectronics.com/ Last edited by ericgibbs; 7th July 2008 at 11:21 AM.

27th September 2007, 03:42 PM	(permalink)
blueroomelectronics	Wow thanks again, expect Cricket PCBs in the mail. Now the schematic and testing are 99.9% done (I'll post the revised schematic on my site), it's on to the layout. Same size as the Inchworm, Firefly & Dragonfly PCBs 107mm x 57mm (Hammond 1591B case) Design thoughts, the Cricket needs to generate as little of its own heat as possible. Slow running PIC, latching relay, low baud rates, LCD display, parts location (the LM34 is in the lower left of the PCB) There is a provision for 3xAA Alkaline for running during power fail (or running sans RS485) So the problem is the main power source. A 5V switch mode wall wart (they are cheap now) this is the least expensive route and will keep the kit price down A LM2575 or other low power switching regulator (can handle up to 40V so 24VAC HVAC can be bridged on to it) $$ A "Roman Black" switching 5V supply (not tested) __________________ Bill Smart Kits build Smart People http://www.blueroomelectronics.com