Solar sensor with OWFS

[nachbar]

I have seen this question several times in the forum, without a clear answer.

I just received a Humidity/Temp/Solar sensor. I have been using a 1-wire network with three
HobbyBoards hubs for the better part of a year, using owfs on a Linux server, with temploggerd
tracking my temperatures, humidity, and barometric pressure. No problem with any of that

However, I cannot seem to read the solar reading from the Humidity/Temp/Solar sensor. I'm not sure
if I know which owfs entry relates to that, but vis always reports exactly zero.

Any hints on how to read this? Should I be seeing the reading in "vis"? Or where?

BTW, I have two temp/light sensors from ibuttonlink.com. On that device,
I have no problem reading light (it is connected to VAD on that device), but the temperature seems
to be read from the DS2438, which self-heats, and thus is high compared to the rest of my sensors.

Thanks!

 

[ericvic]

I can't tell you exactly how to get the readings in OWFS but I can tell you that the solar reading come from the VSENS input of the DS2438. I hope that helps.

Eric

 

[nachbar]

I can't tell you exactly how to get the readings in OWFS but I can tell you that the solar reading come from the VSENS input of the DS2438. I hope that helps.

Eric

Thanks for your quick reply. The hobbyboards.com site does have the schematic, so I had seen the electrical connections. However, I have not been able to get a reading from OWFS that varies with the light that the unit is exposed to. I was hoping, since OWFS is so prevalent, that someone who had figured out how to get the solar reading to work with OWFS would be able to tell me how.

 

[jrshenk]

nachbar,

You should see it as "vis" which has a range of -.250 to .250 V. With the solar sensor you should see "vis" vary from 0 to .250 V. So if you figure a max solar value of about 1000 W/m^2 you can multiply "vis" by 4000 and have a rough idea of solar radiation in W/m^2. I am still playing around trying to figure out how to calibrate it better, but that's how I currently have mine set up.

Jeff

 

[nachbar]

nachbar,

You should see it as "vis" which has a range of -.250 to .250 V. With the solar sensor you should see "vis" vary from 0 to .250 V. So if you figure a max solar value of about 1000 W/m^2 you can multiply "vis" by 4000 and have a rough idea of solar radiation in W/m^2. I am still playing around trying to figure out how to calibrate it better, but that's how I currently have mine set up.

Jeff

Thanks. Perhaps I just don't have it exposed to enough light, which is why I am getting a zero. I will have to try it outdoors in the Arizona sun when I get a chance and see what happens. In the meantime, I will use it for temp and humidity.

 

[marten]

nachbar,

You should see it as "vis" which has a range of -.250 to .250 V. With the solar sensor you should see "vis" vary from 0 to .250 V. So if you figure a max solar value of about 1000 W/m^2 you can multiply "vis" by 4000 and have a rough idea of solar radiation in W/m^2. I am still playing around trying to figure out how to calibrate it better, but that's how I currently have mine set up.

Jeff

Thanks. Perhaps I just don't have it exposed to enough light, which is why I am getting a zero. I will have to try it outdoors in the Arizona sun when I get a chance and see what happens. In the meantime, I will use it for temp and humidity.

Hmm, I must have to much sun! My reading goes to negative values in daylight and what I can make out from the schematic that should not be possible. I have not seen the 0 reading but I have been close to 0 in darkness. But in daylight the reading really takes off into something funny, any ideas?
Marten

 

[mramsey]

I'm using a similar setup. Yesterday, I added a 6 port hub as a power injector. When I mounted the H/T/S outside, I started getting unreliable results which I attributed to too much voltage drop on the long cable. Running with injected power seems to have fixed it.

I'm using OWFS and some of my own scripts to query the sensors, both in PHP and perl. Both scripts are getting the same results. The perl script gets a sample every 5 minutes and posts it to a mysql database. Here's a couple of hours of data from this morning. Overnight the solar reading went to zero. This morning it slowly increased. There are some trees to the east of the sensor so I expect the solar reading to fluctuate in the morning. When it gets to the maximum, though, the sign goes negative. It has stayed at the same value since. I'm thinking it might return positive once it starts to get darker. With full sun, the sensor might be saturated, but why the negative reading?

Mike

+---------------------+-------------+----------+-----------+
| time | temperature | humidity | solar |
+---------------------+-------------+----------+-----------+
| 2008-06-10 09:00:05 | 73.625 | 26.0959 | 0.0183075 |
| 2008-06-10 09:05:05 | 73.9625 | 25.4668 | 0.0185516 |
| 2008-06-10 09:10:05 | 74.1875 | 23.8943 | 0.0185516 |
| 2008-06-10 09:15:05 | 74.4125 | 24.2084 | 0.0187957 |
| 2008-06-10 09:20:05 | 74.75 | 24.2079 | 0.0207485 |
| 2008-06-10 09:25:05 | 75.0875 | 22.95 | 0.165256 |
| 2008-06-10 09:30:05 | 75.5375 | 21.6922 | 0.234824 |
| 2008-06-10 09:35:05 | 75.7625 | 21.377 | 0.233604 |
| 2008-06-10 09:40:05 | 76.4375 | 21.0619 | 0.246785 |
| 2008-06-10 09:45:05 | 77 | 20.4324 | -0.249958 |
| 2008-06-10 09:50:05 | 77.675 | 18.8603 | 0.0217249 |
| 2008-06-10 09:55:05 | 77.7875 | 19.1746 | 0.0229454 |
| 2008-06-10 10:00:05 | 77.9 | 18.546 | 0.0239218 |
| 2008-06-10 10:05:05 | 78.2375 | 18.5453 | 0.171114 |
| 2008-06-10 10:10:05 | 78.6875 | 18.2306 | 0.053702 |
| 2008-06-10 10:15:05 | 78.8 | 17.9161 | 0.105939 |
| 2008-06-10 10:20:05 | 79.25 | 17.9157 | -0.249958 |
| 2008-06-10 10:25:07 | 79.5875 | 17.2865 | -0.249958 |
| 2008-06-10 10:30:05 | 79.7 | 17.2864 | -0.249958 |
| 2008-06-10 10:35:05 | 80.2625 | 16.9717 | -0.249958 |
| 2008-06-10 10:40:04 | 80.375 | 16.6575 | -0.249958 |
| 2008-06-10 10:45:05 | 81.3875 | 16.6565 | -0.249958 |

 

[geoff]

I use OWFS and see the same phenomenon, but it is not an OWFS issue, I get the same result if I use the MAXIS 1-wire viewer to look at the device.

My HB solar sensor is mounted in full sunlight. During the morning the voltage sensed grows from zero up to 250mV by late morning. Depending on the season/weather when the sun is sufficiently bright the voltage sensed reaches 250mV and then the DS2438 "breaks". In my case it immediately flips to read negative 250mV. Why negative ? Who knows? But there it stays until mid-afternoon and the sun gets less bright. Apart from a few minutes in early afternoon when the sensor falls in the shadow of the pole upon which it is mounted!

The sun is pretty bright here (Australia) so my assumption is that the brightness is such that more current flows though the photodiode than was anticipated when the board was designed.

I have a three proposed fixes. Has anyone tried any of these or got any other suggestions ?

Option (1) a diffuser of some sort to reduce the light intensity (an old pair of sunglasses maybe ...)
Option (2) reduce the 390ohm resistor. This would drop the voltage, at the cost of increasing the load on the bus. As long as not trying to use parasitic power this should be okay.
Option (3) reduce the current. The saturation voltage is reached at around 0.6mA. A resistor of around 3k in the positive leg of the photodiode ought to cut the current by around 50% at the time the 250mV saturation point is reached. The linearity (or otherwise) of the output versus light intensity would be messed up, but this doesn't really matter to me.

 

[comet48]

I use xapmcs1wire to read the sensor - from http://mcssprinkler.com - at the bottom of the page under xAP resources. Check the browser output option when you have it up and running.

 

[skyewright]

Option (1) a diffuser of some sort to reduce the light intensity (an old pair of sunglasses maybe ...)

I've seen people use a table tennis ball, e.g.
Weather-above solar sensor set up details

 

[geoff]

Option (1) a diffuser of some sort to reduce the light intensity (an old pair of sunglasses maybe ...)

I've seen people use a table tennis ball, e.g.
Weather-above solar sensor set up details

After a few experiments I found a 270 ohm resistor in parallel with the existing 390ohms did the job perfectly. At noon on the brightest midsummer day (southern hemishere here) the sensor reads approx 150mV whereas before it saturated at over 250mV. No magic in the 270ohm value, just that was what I had to hand.

 

[skyewright]

After a few experiments I found a 270 ohm resistor in parallel with the existing 390ohms did the job perfectly.

Interesting.
A side benefit of the table tennis ball diffuser approach however is that it can help at least a bit with the acceptance angle, which can be a problem for use at higher latitudes. The acceptance angle of the sensor is 140 degrees, and where I live even the noon day sun doesn't manage 20 degrees above the horizon from the end of October through to some time in February.

A diffuser isn't a full solution for that, but it's the best I've seen so far.

I have wondered about trying to fit some sort of fish-eye lens above the sensor (maybe a half round clear glass bead?) but not got further so far than just thinking about it. If anyone has ideas on the subject I'd be very interested to hear them.

 

[makki]

Getting back on this topic Hobbyboards T/H/S with owfs:

Has anybody already measured/compared at which light-level (Lux) the max value of .25v reading VIS is reached ?

As far as I understood, if I get a negative reading it's saturated and the datalogger can assume 100% which would be fine for me (as long as it's high enough);
don't want to use diffusors, covers or change resistors..

Michael

 

[makki]

While testing, I just noted that I already have a THS with the new Diode, not clipping anymore. So most likely not comparable with the ones this thread started..

Compared to my other weather station saying 99 kLux (which is the max there), I get readings from vis of approx 0.165
Anyway, I'd really appreciate some hints on how to calculate vis->lux, what the range is or sharing some readings..

Michael