Analog UV Light Sensor Breakout - GUVA-S12SD - Increase sens

[max86]

Dear Adafruit community,

I have a question to the Analog UV Light Sensor Breakout - GUVA-S12SD. I connected this part to the analog input of my Arduino Uno R3 and I can read some signal/voltage from the OUT Pin when I illuminate this sensor with strong UV light 1x15W bulb (365nm). However, I would like to use this sensor to detect the light of UV diodes with lower UV light intensity - currently, the sensor cannot detect the light of my UV diodes (no difference in OUT voltage when the UV diodes are switches ON). Therefore I would like to ask if there is any possibility to increase the sensitivity of this sensor? I already set the reference voltage to 1.1 V using "analogReference(INTERNAL)" - this slightly increased the intensity values of the 15 W UV bulb - however I still cannot sense mir UV diodes. It would be really great if someone would have an idea how to increase the sensitivity of this UV sensor or knows a UV sensor with higher sensitivity.

Thanks a lot for your help and best wishes,
Max

[adafruit_support_rick]

What is the wavelength of your UV LEDs? According to the product page:

Please note, our UV LEDs are 400nm, outside the range of this sensor, so if you're trying to test this sensor, don't use them!

[max86]

Thanks for your reply. The UV LEDs have a peak wavelength of 310 nm - so it should be in the range of the detector...However, the sensor is not responding to the UV light. Is there any possibility to increase the sensitivity? Thanks a lot for your help.

[adafruit_support_rick]

Sorry - there are no adjustments

[carsonroberts]

I recently purchased two of these sensors, and I cannot get either to work. Is there some kind of trick?

I initially soldered on the three-pin header, and then soldered wires to the pins on the header.

I attached the + an - to a voltage source at 4.5V, and measured from Out to -. The + and - pins are clearly getting power.

I tried initially with a 3W mercury lamp, which gives out 0.16W of UV at 254 nm. No voltage change on Out.

I also took the device into the sun, but again, no change in the voltage from the Out pin.

I did set up a chopper and lock-in amplifier, and saw a signal on the Out pin that varied with the mercury lamp intensity, but this output was independent of the voltage applied to the device.

Is this device somehow extremely sensitive to soldering or static? I did all my work on a grounded workbench, and I work with static sensitive stuff all the time.

So:
1.) Should this device respond to the mercury lamp at 254 nm? the diode seems to.

2.) Are there any tricks to make it work?

[adafruit_support_rick]

Please post clear, detailed pictures of both sides of the breakout, showing your soldering and wiring

[carsonroberts]

I doubt the problem is with the solder, but I have taken some photos. I purchased two of the devices, and both fail in an identical way. I can measure the input voltage between + and -, but get no voltage out between Out and -.

One device was wired up with three wires, Red +, Green - and purple Out. The header was soldered in from the bottom of the PCB:

DeviceOne_Header_Connections_1.jpg (205.72 KiB) Viewed 967 times

Here is a close-up of the solder connections on the top side of the first device:

DeviceOne_Header_Connections_2.jpg (135.13 KiB) Viewed 967 times

Here is a third photo of the first device, showing the header, solder connections to the PCB, and the three colored wires:

DeviceOne_Wiring_1.jpg (212.94 KiB) Viewed 967 times

[carsonroberts]

The second device was wired with two wires to connect to a battery of three AA batteries, to power it with ~4.5V. For this one, the header was inserted from the top of the PCB, and connections made to the three long pins. The Out signal was measured with clip leads connected to the Out and - pins

Here is a photo of the second device, with the red + and black - wires, as well as the battery connector:

DeviceTwo_Battery_Connections.jpg (258.16 KiB) Viewed 965 times

As shown in the photo above, I mounted the device to an optical post with a 2-56 screw. Is it possible that mounting the board somehow damaged it? I have tested both devices without the mounting screw as well. The result has always been the same: Voltage gets to the op-amp, but no signal comes from Out, neither with 0.16 W of 256 nm Hg light, nor with direct sunlight.

The power connections were made to the long pins on the header on top of the PCB. The wires were flowed with solder then tacked onto the pins. The solder flowed well onto the outside of the pins, but not all around. Again, the power is getting to the board, and appears on the two top pins of the op-amp. Here is a close up photo of the wiring to the header on top of the PCB:

DeviceTwo_Battery_Connections_3.jpg (186.34 KiB) Viewed 965 times

The header was soldered from the bottom of the PCB, here is a closeup of the solder joints on the bottom of the second device PCB:

DeviceTwo_Header_Connections_4.jpg (540.96 KiB) Viewed 965 times

[adafruit_support_rick]

According to the datasheet, 254nm is pretty close to the end of the responsivity curve. Peak is 360nm
https://www.adafruit.com/datasheets/1918guva.pdf
Still, it should work in direct sunshine. Not sure what's going on. I'll check with somebody else.

[adafruit2]

are you measuring it with an oscilloscope? just cuz at least that way you could see any spikes

[carsonroberts]

I have tried the mercury lamp with an oscilloscope, and gotten nothing at all on a 2 mV scale with the light on and off. With the same settings, I was able to see a signal with a chopper and a lock-in, but this was independent of the power applied to the PCB. I did not take the oscilloscope outside, so the sunlight measurements were done with a digital voltmeter.

On the scope, there was noise at a level of ~ 1mV, but no change at all in the signal with the sensor ~2mm from the light, and the light turning on and off.

I realize that 254 nm is near the lower limit of the diode, but from the data sheet, the detectivity is about 1/8 of the maximum at this wavelength, and the light is about three orders of magnitude brighter than the UV levels in sunlight, as far as I can tell; ~ 500 W/m^2, compared to 0.25 W/m^2 for a UV index of 10.

[carsonroberts]

So I have two of these devices that do not work. I know they are small and inexpensive, but are they tested before shipping? Is there any chance of getting one that has been verified to work?

[adafruit2]

we test every single one with a UV lamp before shipping, so I think replacing wont help. if they just arent working you can email support@adafruit for a refund.

[carsonroberts]

I took one of the boards, and soldered very fine wire wrap wires onto the diode (still on the PCB).

Then, I built the circuit in your documentation:

guvaschem.png (16.87 KiB) Viewed 910 times

but using an LM324 comparator instead of the MCP6001 op-amp.

I connected the wires from the diode on the PCB to my circuit, powered it with 4.5V, and I get 0.7V output from my mercury lamp, and ~3V output in full sunlight. That is, it works the way the original was supposed to.

So, nothing wrong with the diode, nothing wrong with the circuit design, but neither of the units I purchased work as built.

I seem to have a solution, but it would have been much easier if the boards I bought had worked.

[adafruit_support_rick]

Why don't we replace one of them and see if it works for you. Please email [email protected] with a link to this thread for a replacement GUVA