UV-C Sensor

[Marvo]

Hi all, I joined a while ago and have only lurked in the background since.

I've got a small project at my home I'd like some help with please. Where I am we are in the middle of a drought that will possibly see the water supply being turned off in the next month or two but thankfully I have a borehole that I've now plumbed to my house as an alternate supply. For most of the house such as bathrooms I'm simply filtering the borehole water but for the kitchen where we require drinking water I have installed a UV-C steriliser. It's a very simple stainless steel tube that contains a 29 watt T5 UV-C fluorescent tube 260nm that kills harmful bacteria and the tube is driven by a generic external ballast. The unit cheap and nasty but it works fine.

Because it's drinking water I wanted to make an alarm to indicate tube failure. My initial approach was to use an off-the-shelf current monitor but the results have been hit and miss. I suspect the current difference between the tube being lit and the tube failing is too small. My next idea is to use a UV sensor but all the options I've found are hellish expensive so I'm going to make my own UV sensor. I'm after ideas how it can be done using fairly basic components you might find in a third or maybe second world country.

[mikeselectricstuff]

Is there a tube failure mode that means the tube emits visible but not UV? I would imagine the tube emits some visible that you could detect.
Alternatively use some fluorescent material to convert to visible.
 

[Marvo]

I don't think the tube emits any light other that in the UV-C spectrum. Do you think the standard photocell sensor in a day/night sensor operate to light in the UV-C range?

[mikeselectricstuff]

I don't think the tube emits any light other that in the UV-C spectrum. Do you think the standard photocell sensor in a day/night sensor operate to light in the UV-C range?

I'd be surprised if there isn't some visible. An issue is protecting a sensor from UV damage long-term, so some glass in the way would probably do.

As UVC is ionising, you may also be able to make an ionisation detector.

[Marvo]

Thanks for the thoughts Mike.

Would glass that protects things from UV damage not block the UV and make it more difficult to sense? Would UV-C light cause ionisation of the water flowing through the sterilizer? Would this ionisation be easier to detect that the UV light itself?

[Marvo]

Here's a few pics of what I have;

[mikeselectricstuff]

Thanks for the thoughts Mike.

Would glass that protects things from UV damage not block the UV and make it more difficult to sense?

No because you are sensing visible , not UV

Would UV-C light cause ionisation of the water flowing through the sterilizer? Would this ionisation be easier to detect that the UV light itself?

I'm not a chemist but don't think so - UV is just killing bugs

[Someone]

Because it's drinking water I wanted to make an alarm to indicate tube failure.

Based on your other replies this appears to be the least of your problems. You can't just shine some arbitrary amount of light at water to make it safe to drink, the dose rate needs to be carefully controlled and interlocked to prevent any water passing through before it has been adequately disinfected.

I don't think the tube emits any light other that in the UV-C spectrum. Do you think the standard photocell sensor in a day/night sensor operate to light in the UV-C range?

Buy a system that is built by people who know what they are doing.

[Marvo]

The dose rate isn't arbitrary, it's predetermined by the manufacturer who states a maximum flow rate for the device. As long as the flow rate is below 8 US Gal/min they claim it is effective. I'm not looking to operate the device outside of these perameters, I sized the unit according to my max water demand flow rate when I bought it. I'm just looking to give myself a visual indication alarm should the lamp fail.

[thm_w]

Yes Marvo your idea is sound.

You want something like this: https://www.digikey.ca/products/en/sensors-transducers/optical-sensors-photodiodes/543?k=uvc and a transimpedance amplifier.
Note the sharp cut-off at 280nm.

Interesting that the call the supply "20-40W automatic control", I wonder how its modulated? We were using temperature sensing on the glass to shut down the lamp in case the water stops moving, as the water can overheat if you are pumping 40W into it.

Things to watch out for:
- properly water sealing the sensor from the water, sapphire or quartz window can be used
- do not look at or expose your skin to the UV light, even for a short time

[jct]

My suggestion would be as simple as a timer circuit. All UV lamps (according to data sheet/s) have a limited operating time frame where the light intensity drops to 80% or below - usually 8000h. Make a simple timer circuit that counts the operating time and audibly and/or visually notify when approaching the 8000h operating time. It will not indicate a broken lamp or ballast etc, but at least notify when its time to replace the lamp.

[David Hess]

Is there a tube failure mode that means the tube emits visible but not UV? I would imagine the tube emits some visible that you could detect.

Alternatively use some fluorescent material to convert to visible.

They make little plastic panel inserts which have a fluorescent dye for visually confirming that UV is present.  UV EPROM erasers used to have them and I have seen them in other applications also.  I have no idea what they are called or where to find them.

[BravoV]

Is there a tube failure mode that means the tube emits visible but not UV? I would imagine the tube emits some visible that you could detect.

Alternatively use some fluorescent material to convert to visible.

They make little plastic panel inserts which have a fluorescent dye for visually confirming that UV is present.  UV EPROM erasers used to have them and I have seen them in other applications also.  I have no idea what they are called or where to find them.

Most of those Glow in The Dark thingy is sensitive to UV and turns bright when exposed to UV, and then coupled it with ordinary light sensor for detection.

[Zero999]

I don't think the tube emits any light other that in the UV-C spectrum. Do you think the standard photocell sensor in a day/night sensor operate to light in the UV-C range?

I'd be surprised if there isn't some visible. An issue is protecting a sensor from UV damage long-term, so some glass in the way would probably do.

As UVC is ionising, you may also be able to make an ionisation detector.

Yes, UVC lamps produce some visible radiation, mostly in the blue region. It should be easy to detect that with a CdS cell or photodiode.

[David Hess]

I used to run across inexpensive flame sensors which detect UV on Mouser all the time but now I cannot find them.  They looked like neon bulbs but with a special envelope (quartz?) and perhaps a different gas fill and electrode material which detects the UV emitted from a flame and not ambient light.  They were relatively inexpensive compared to similar bulbs from Hamamatsu.

[Marvo]

Thanks guys, some interesting info and ideas. I'd particularly like to explore the flame sensor that detects UV. It would be great if anyone has more info on this.

[Zero999]

Thanks guys, some interesting info and ideas. I'd particularly like to explore the flame sensor that detects UV. It would be great if anyone has more info on this.

After doing a bit more research, I think that's a good idea.

According to Wikipedia, a fluorescent lamp can fail, due to loss of mercury, which can happen when all of it is absorbed into the glass tube. If this happened in a germicidal lamp, it would still continue to emit visible radiation, but would no longer emit much UVC, which would fool a visible light sensor.
https://en.wikipedia.org/wiki/Fluorescent_lamp#Loss_of_mercury

How about using an ordinary, visible light detector, behind a piece of glass, with the phosphor, from an old fluorescent tube? If the sensing threshold is set high enough, then it won't be triggered by the dim light generated from a tube suffering from mercury loss.

Another option could be both a timer and a visible light sensor? The tube could be routinely replaced,. after 8000h has elapsed or it no longer emits and visible light. Mercury depletion won't happen after just 8000h. It takes much longer than that and isn't common because the filaments normally fail first.

[thm_w]

My suggestion would be as simple as a timer circuit. All UV lamps (according to data sheet/s) have a limited operating time frame where the light intensity drops to 80% or below - usually 8000h. Make a simple timer circuit that counts the operating time and audibly and/or visually notify when approaching the 8000h operating time. It will not indicate a broken lamp or ballast etc, but at least notify when its time to replace the lamp.

In addition to not showing a broken lamp or ballast, it will also not indicate if the lamp glass sleeve has become dirty or obstructed.
But if you have a pre-filter, I assume this wouldn't be much of a problem.

[David Hess]

Thanks guys, some interesting info and ideas. I'd particularly like to explore the flame sensor that detects UV. It would be great if anyone has more info on this.

The ones I saw looked like a normal neon bulb and just cost a couple dollars.  I searched but did not find them online but I am probably using the wrong search string.

[John Heath]

Yes Marvo your idea is sound.

You want something like this: https://www.digikey.ca/products/en/sensors-transducers/optical-sensors-photodiodes/543?k=uvc and a transimpedance amplifier.
Note the sharp cut-off at 280nm.

Interesting that the call the supply "20-40W automatic control", I wonder how its modulated? We were using temperature sensing on the glass to shut down the lamp in case the water stops moving, as the water can overheat if you are pumping 40W into it.

Things to watch out for:
- properly water sealing the sensor from the water, sapphire or quartz window can be used
- do not look at or expose your skin to the UV light, even for a short time

This one is looking good as the nasty UV that is taking care of business is in the far UV 250 n meter range. The suggested sensor will not detect near UV from the sun so you will not have to shade it from daylight. It will only see far UV from the sterilizing lamp. Next question is can you buy a Digi-Key Part Number    SD012-UVC-011-ND or equivalent sensor? How to turn that into an alarm? Many here that know how. In the mean time put a new white t-shirt or new white socks next to it. If they glow in the dark chances are the lamp is good. I would ware UV glasses for that test.

I was wondering if I could tax you on a question that is unrelated. In Cape Town did you or your friends know of Sixto Rodriguez and his songs. Did your parents talk of him ? 

[NiHaoMike]

Since it's a one off, what about program a bit in an EPROM and check how long it takes to erase? No idea how many erase/write cycles an EPROM will endure, but maybe check it only once a week or even less often since sudden failure modes (e.g. bulb cracking) tend to be easily detectable in some other way.

[BrianHG]

Since it's a one off, what about program a bit in an EPROM and check how long it takes to erase? No idea how many erase/write cycles an EPROM will endure, but maybe check it only once a week or even less often since sudden failure modes (e.g. bulb cracking) tend to be easily detectable in some other way.

You cant rely on the behavior of 1 bit, but, maybe, program a whole eprom & keep on reading it under the lamp and see how long it takes for 50% of the contents to erase.  This might be a little more accurate, but, not that different programmers may burn the eprom with slightly different burn speeds, number of burn cycles & burn VPP voltage affecting the erase speed.  Also note that the distance between the eprom and lamp will radically affect the erase speed.

[David Hess]

Since it's a one off, what about program a bit in an EPROM and check how long it takes to erase? No idea how many erase/write cycles an EPROM will endure, but maybe check it only once a week or even less often since sudden failure modes (e.g. bulb cracking) tend to be easily detectable in some other way.

40 or 50 cycles was common in my experience but that is with only a limited range of UV erasable EPROMs.

[NiHaoMike]

I was thinking make the EPROM part of the unit and have it programmed by the unit itself for doing a test. My idea of using just one bit is so you can use a different one when the first one wears out.

[floobydust]

Solar Light’s PW254 germicidal lamp monitor has a teflon body. I guess you could put a UV-C photodiode in a teflon enclosure. Other plastics degrade under UV which is bad for drinking water.

The Hamamatsu UV TRON tubes are extremely sensitive, they can detect a candle flame across the room, even 75m away  They would be great for a kitchen fire alarm. But the tubes need 400-500VDC.