Topic: Davis Solar Radiation Reading never reach 1800 W/m2? Pls help me out.

[Stefan]

Dear All,

We are customizing our visibility sensor to be plug and play in ISS solar radiation RJ11 Port.(10~1800 meters correspond to 10-1800 W/m2), that's why we really need to reach "1800" reading.
But strange thing happen, and Davis console never show up a 1800 W/m2 reading at its solar radiation port.
I am posting here to find out if you guys notice this?
 
We have studied the way how ISS sample data from wind cup/ wind vane,  and we successfully finished customization of Davis VP2 compatible ultrasonic anemometer last year.
In our understanding of ISS:
The wind speed signal is transmitted in pulse
and wind direction, solar radiation signals are transmitted in variable resistance(digital potentiometer) to ISS.
 
Three methods had been used to test if ISS/Console can recognize maximum reading of solar radiation.

Method one.
When our HY-SLV2E output 3VDC voltage, the reading of console is keep "----", we think it trigger ISS(sensor suit transmitter) limitation of detection of voltage.
After we slightly reduce the voltage as 2955 mV, we got a reading: 1782 Watt/m2
Below is our test result for your info.(the output voltage resolution of our digital to analog converter is about 0.7 mV.
Voltage below is actual voltage measured by multimeter, we tried to change various output voltage.
But we found the smallest addtion of voltage we increased based on 2955mV all led to one single result: a null display on Davis console "----"
  [ You are not allowed to view attachments ]
Below is screenshot taken from Davis solar radiation sensor manual:
 [ You are not allowed to view attachments ]

 
Method two.
We use a potentiometer to directly output/reflect the voltage supplied by ISS to ISS itself.
By changing its resistance, we are able to simulate a solar radiation reading to cheat the ISS.
But the reading we got from Davis console is still constrained to its maximum 1782 W/m2. (we test it both on cabled VP2 and wireless VP2)

Method three.
Davis sensor suit transmitter is considering "wind direction" and "solar radiation" sensor as potentiometer/resistance and measuring how much voltage is distributed on them from the 3VDC reference voltage.
Interesting thing happened during our test, photo attached IS for your reference to explain you what we found.
No matter how hard I tried to adjust angle of Davis wind vane purposely "wrongly" plug at Solar Port at Davis sensor suit transmitter  (around 350°), the maximum reading of solar radiation I can get at console is still 1782 Watt/m2, which substantiate my previous view that console itself may not be able to recognize maximum reading of solar radiation before the upper limitation Davis programmed lead it to output a "----" state.(just like method 1,2)
 [ You are not allowed to view attachments ]

Look forward to hearing opinion from members at this forum.
Thanks.

[Stefan]

I conducted more test to find out where went wrong.
The phenomenon happened as below:
I did the same test again.
I rotate wind vane to make it as 180°, 346°, 347°,  348°,  349°,  350°, and insert RJ11 connector to solar radiation while keep wind vane stable.
Below is what I got.
 [ You are not allowed to view attachments ]

Clearly, we can see that Davis VP2 fail to reach higher solar radiation reading.
I also test wind vane, we can get 359° reading by carefully rotate wind vane.
As for solar radiation, it just stuck around 1781, any higher value just be reported as "----"
The only conclusion I can draw now is that resolution of ISS' solar radiation port is lower than wind port.
I believe this is the only reason that solar radiation stuck at 1782 W/m2.
Anyone who have interest can try it.
I badly hope I was wrong and this problem can be fixed...
 [ You are not allowed to view attachments ]

[johnd]

Why not just use a precision 20k potentiometer? That way there'll be not interference from the extra resistors in the anemometer circuit.

Anyway, I don't know what the accuracy of the A/D converter in the solar input might be, but you seem to be seeing ~1% error which I wouldn't be too surprised about.

[Stefan]

Hi Jonhd,
Yes, we've tried precision potentiometer before, samething happened.
That's why I used wind vane (a standard potentiometer of Davis)
1782/1800=99%...
This 1% error is driving me crazy.
It's weird that the A/D converter of wind didn't show such error, I always succeed to get reading of 359°.

Talking about interference from the extra resistors in the anemometer, we don't need to worry about that.
Since we already knew wind speed is output via separated cable in pulse signal.

[Mattk]

The old 99/1 rule, spending 99% of the effort and time trying to find 1% of the problem that generally isn't a real problem anyway. Nothing is always 100% with a spec that has a +/-, there is always some +/- factor.

[Stefan]

But the wind direction conform to their description- 359°...
I think Davis could have done a better job.
Obviously they didn't take the upper limitation of solar radiation into consideration seriously.

[mcrossley]

wind direction, solar radiation signals are transmitted in variable resistance(digital potentiometer) to ISS.

I'm no electronics expert and don't really understand what a digital potentiometer is, but I think wind uses a physical potentiometer and solar uses a common rail op amp? Dunno if that makes a difference?

[DaleReid]

Stefan,
Considering the solar radiation will NEVER get that high I can see how the engineers would not worry too much about having it be true on the screen at the extreme high end of the measurement.

It is what it is, and without the design circuit to see how they did it, and with component variation, it is unlikely you'll ever make your station adapter to your needs.  How about changing your project design to be 1500 rather than 1800?  That is under your control, while the inner working modification of the Davis is not.

[Stefan]

wind direction, solar radiation signals are transmitted in variable resistance(digital potentiometer) to ISS.

I'm no electronics expert and don't really understand what a digital potentiometer is, but I think wind uses a physical potentiometer and solar uses a common rail op amp? Dunno if that makes a difference?

Yeah, you're right, wind vane is a true physical potentiometer. However for solar radiation, since it doesn't have "rotation part" can physically change resistance  while rotating, it can change resistance by digital signal(from MCU).
At the end, they have same function and work as a variable potentiometer to inform ISS their current status.
Thereafter, you get a reading after conversion/calculation on console.

[Stefan]

Stefan,
Considering the solar radiation will NEVER get that high I can see how the engineers would not worry too much about having it be true on the screen at the extreme high end of the measurement.

It is what it is, and without the design circuit to see how they did it, and with component variation, it is unlikely you'll ever make your station adapter to your needs.  How about changing your project design to be 1500 rather than 1800?  That is under your control, while the inner working modification of the Davis is not.

I was thinking exactly in the same way as you did, DaleReid.
Truth is that NOBODY cares about little error of high solar radiation reading
We rarely happen to encounter such high solar radiation in daily life.

[Stefan]

More correlationship between solar radiation value and wind direction value for your info.
 [ You are not allowed to view attachments ]
It seems it's not just high edge of solar radiation value but also low edge has same problem.

[mcrossley]

Also don't forget that solar values less than 5 are truncated to zero by the console.

[kobuki]

As others noted, the wind direction is sensed by a simple potmeter that has a dead zone at North of about 20°. The solar sensor input is specified as a fixed voltage per W/m2 and you need to actually supply the appropriate voltage on the output of your visibility sensor. A digital potmeter is not a good fit and I'd rather use a cheap and accurate DAC (maybe one in the MCU of your custom sensor?). Note that you cannot exceed the voltage supply of the SIM board with that output. If you do the ADC measurement will flat out at supply voltage and you might damage the chip on the SIM or its ADC. Also, getting solar radiation on the surface of the Earth over about 1100 W/m2 is not really possible (well, sometimes the "edge of cloud effect" can spike a bit higher than it were otherwise possible). The console probably simply shows an error above that. If you'd be willing to lower the resolution of your sensor to say, 2m, that would solve most of your problems. Of course, the value displayed on the console wouldn't be a direct read then.