47 replies to this topic

[WiSi-Testpilot]

Since a few days I have the ARTCAM-2020UV-USB3 camera with the GSENSE2020BSI sensor.
http://www.artray.us/usb2_uv.html
It was cheaper than the price that was mentioned here in the forum, but please don't ask for the price. Now I want to show a first noise and leak test.

Recently I learned that gas light emits UV-C, see page 4.
https://www.dguv.de/..._gasbrenner.pdf

Therefore I used a simple lighter as UV-C source. The flame was recorded in front of a cloudy sky, so that the leak of the filter can be estimated.

The lens is the UV2528B, the aperture was 2.8, the shutter time was 1 s, the Global Gain was about 50% .
I used this filter:
https://www.ahf.de/p...-uv-et-bandpass









The vignette is because the size of the sensor is 1.2” x 1.2”, whereas the lens is made for 1” sensors. However it is not visible with lower gains, see for example the image with my 313 nm filter. Shutter time was about 10-100 ms. I did not memorize it.
https://www.ahf.de/p...-bandpass?c=138


Best regards,
Wilhelm

[Stefano]

Your UV2528B filter seems to have around OD 5+ as average, which is good but maybe not perfect. It is very difficult, basically impossible to reach below 300 nm using only sunlight, but your results look promising.

What is the dark object in the last image, a glass pane?

[WiSi-Testpilot]

Yes, it is a glass plate. With a Baader U filter (320-380 nm) the glass plate is transparent.
Below 300 nm is no sun light. The images represent the leak of the 250 nm filter at high gain.
Best regards,
Wilhelm

Edited by WiSi-Testpilot, 18 October 2020 - 20:23.

[dabateman]

Wow this is super cool.
Your all set with a quartz lens and what looks like a great sensor. I can see you having some serious fun with that.

[colinbm]

Wow this is serious UVC photography. Looking forwards to seeing some more.

[JMC]

Interesting. Was it easy enough to remove the sensor coverglass?

With regards to the cost of the camera, I'd be interested to know how much it was? While I'm not looking to buy a new camera at the moment, one with known high UVB and C sensitivity might be of use next year. I know you said please don't ask, but the link doesn't give a price and it would be good to know. If you'd rather send it by PM then that's no problem.

[WiSi-Testpilot]

In the manual I found: Cover glass’s material: Labo-USQ. (Can be changed to no glass model or AR coating cover.)
My version has no cover glass.
Please write to the company for a current price. They always answer promptly.

In the next few days I would like to try the following in daylight:

Edited by WiSi-Testpilot, 19 October 2020 - 16:11.

[UlfW]

The next step might be to get a stronger light source for proper illumination in UV-C
https://www.jstage.j.../1/23_1_53/_pdf

[JMC]

Ok Wilhelm, will check with the vendor for the price. No cover glass will be great for sensitivity, but be careful with that sensor. Might be worth getting a quartz or fused silica window to be safe.

Edited by JMC, 29 December 2020 - 16:45.

[Andy Perrin]

Wow, that is one nice toy! I look forward to your further experiments. Did you have a work purpose for this, or did you just get one to play with? You can't be putting it on a drone with no UVC in sunlight to visualize, unless you are doing powerline inspection by drone?

[dabateman]

I emailed the company, but haven't heard back yet.
Does the software work in Linux or more specifically a Raspberry pi computer. Debian, I think is the pi Linux flavor.
How are you running it?
How are you powering it?

Does the USB 3 port also provide power or do you need an input?

Sorry for all the questions. You have peaked my interest again.

[WiSi-Testpilot]

Andy Perrin, I'm retired and do it as a hobby. But I’m still interested in everything that you can't see with the naked eye.
The camera is not suitable for use on a drone, because it is controlled by a Windows program and has a rolling shutter at long exposure times. Corona discharges on high voltage lines are the main reason for the purchase, but have not jet a save 12 V regulator.
Speaking of save, Jonathan, I will never remove the lens.
Dabateman, you have a PM.
The software only runs under Windows, as far as I know. The camera needs 12 V from a power supply.
Best regards,
Wilhelm

Edited by WiSi-Testpilot, 20 October 2020 - 08:59.

[dabateman]

Thanks for the update.

12V only might be doable with a battery pack. There are some astro cameras that run similar with there cooling options. I will have to see if my external battery can also output 12V. I got it mainly for my spectrometer, which needs 5V. So it might not.

Only Windows is a harder problem. I did hack a version of Windows 10 to run on my pi camera system. Buts its very slow and was too unstable to support Starlight live, to run my Lodestar X2C camera. Wine in the pi Debian flavor didn't work at all with Starlight live.

I guess a tablet might work. Looks to be a tricky solution again.

Edited by dabateman, 20 October 2020 - 10:48.

[WiSi-Testpilot]

Here are two images of a candle flame at 250 nm indoor. The shutter time was set to 300 and 600 ms. All other settings where default. The UV light comes from electronically excited molecules. It is not blackbody radiation. Very different from the visible flame is the bright spot on the top.

Edit:
I have added a colourized image.
Edit1: 21.10.2020
I have added a vis-image.

600 ms


600 ms, color scheme gem


300 ms


Vis-image. It was made shortly after the UV images.

Edited by WiSi-Testpilot, 21 October 2020 - 10:39.

[Stefano]

I wonder how deep into UV you can get. As far as I know, David holds the record in this forum for imaging the mercury 185 nm line: https://www.ultravio...l-185nm-attempt

Judging from the sensor graph in the datasheet, you have ~16% quantum efficiency at 200 nm (it is "Q.ExFF", which is similar to Q.E I guess?), and you may reach 190 nm. Usually silicon sensors are said to respond between 190 and 1100 nm.

That's an amazing sensor. You can easily shoot UVB under sunlight, something I dream to do one day. Seeing dark glass because it absorbs UVB is very cool. You may try to image something like a glass of water if you want.

It also seems your sensor can go a bit beyond 1100 nm, as in the graph you still have ~5% sensitivity there, a third of what you have at 200 nm, but still not zero. You may have a tiny bit of sensitivity at 1150 nm.

[Andy Perrin]

To be honest, I have one of the Sirchie KSS lens and image amplifier setups that can do UVC, but the main reasons I haven't yet are that I haven't figured out a good way to do it safely and also the lack of UVC in sunshine to image scenes with. It's a dark dark world out there below 300nm.

[dabateman]

If you want to image deeep uv, the problem is a good clean light. My 193bp filter leaks into 254nm and even ozone bulbs don't stay at 185nm for long and very expensive.
I am hoping the 207nm lights come into fashion and 222nm becomes cheap due to COVID. That will help with the lighting problem.
Best controls are water (dark less than 200nm), isopropyl alcohol (dark below 210nm) and acetone dark below 300nm.

Actually acetone maybe a good control for your lighter UVC. See if you see through it, next too water.

[WiSi-Testpilot]

Stefano, here are two images of a glass of water. Today it is dark and rainy, so I will repeat it, when the sun shines. The spot is a bubble in the glass. The water is a little bit darker. The settings are default, but the shutter time was 0.9 s, contrast was increased to 30. The filter was the 313/25 nm, see above.






Now I have to mend the rear tire.
Best regards,
Wilhelm

Edited by WiSi-Testpilot, 21 October 2020 - 11:05.

[Stefano]

Darker water, more or less transparent glass. Kind of the opposite of what I expected.

[WiSi-Testpilot]

It depends on the thickness of the glass. The plate was thicker. The bottom of the water glass is almost black in the UV.