Cheapest way to check if a UVC bulb is as advertised?

[LaserSteve]

For measuring lasers above about 15-20  mW, we use what is basically a Absorber mounted on a  Pelter element in reverse, which is the core of many laser power meters.  While it is possible for a Hobbyist to build one, the issue is that they respond to pretty much  "EVERYTHING" around them.  Which is why my post above stresses finding the wavelength.

 Carbon black in the form of Black Krylon Engine Paint is generally a good start for the absorber, with a decent instrumentation amp circuit on the Peltier. Calibration, however, is an art form, with the classic means being a heater coil.  You need a massive heat sink on the other side of the array.

UV low pass glass filters or dichroric filters don't grow on trees, making this task a bit difficult. Fluorescence conversion can be anything but linear unless you have in-organic phosphors.  SO:

 If you've got 100$ to spend you can get a GaP photodiode from Thorlabs.  If you go GaP, I hope you can find a decent Ampere per Watt curve for a representative device.

Part number is FGAP71

Photonics is easy, but accurate quantitative photonics is difficult and expensive.

Steve 

[Zero999]

UV low pass glass filters or dichroric filters don't grow on trees, making this task a bit difficult.

I would have thought you'd want high pass, not low pass filters, unless your source has considerable energy above the useful germicidal frequencies, although it would produce a lot of ozone, if that were the case. I think lots of those cheap UV lamps will really be black light and there will still be a significant thermal IR blackbody, which would both be picked up by a bolometer. Cooling would reduce the blackbody emission, but it will affect the efficiency of the lamp.

[LaserSteve]

I have attached an  optical "Shortpass" typical graph from the Edmund Optics website. Optics uses the terminology just a little differently:

[pipe2null]

What do you think of these photodiodes?

$30-ish on digikey with reduced responsive range, no filter required for UVC/germicidal detection range but only "0.076 mm2" active area:
https://media.digikey.com/pdf/Data%20Sheets/Photonic%20Detetectors%20Inc%20PDFs/SD008-2171-112.pdf

$350-ish on digikey, has "1 cm2" active area but has normal response to visible light, requires filter etc.:
https://optodiode.com/pdf/UVG100.pdf

[Zero999]

What do you think of these photodiodes?

$30-ish on digikey with reduced responsive range, no filter required for UVC/germicidal detection range but only "0.076 mm2" active area:
https://media.digikey.com/pdf/Data%20Sheets/Photonic%20Detetectors%20Inc%20PDFs/SD008-2171-112.pdf

$350-ish on digikey, has "1 cm2" active area but has normal response to visible light, requires filter etc.:
https://optodiode.com/pdf/UVG100.pdf

The big sensor has a greater response in the near IR, than UV, so it would be a challenge to filter.

The current from the small UV-only sensor is tiny, but the data sheet specifies it with a 306nm source, which it has little sensitivity to: only 68fA at 10mW/cm². Going by the response graph, it should be 25 to 30 times more sensitive to germicidal wavelengths, but it still isn't very much. It will be fun designing a circuit to work at such low currents, an ultra low bias current op-amp is a must a lens or parabolic mirror could be used to provide more gain, but the former might be expensive/difficult to get hold of and the reflectivity of some metals might not be as good at such short wavelengths, making the latter difficult to do.

[pipe2null]

So...  I'm thinking I need to split this into 2 different projects to consider, and then choose one to proceed with, depending on cost, benefit, and likelihood of success building this myself.  In either case, I am assuming that the measurements/readings would be obtained with a scope and/or cheap dmm, 'cause that's what I have  :

Solution 1:  Cheapo UVC detector: $30 photodiode + parts for ultra low bias current amplifier (haven't messed with op amps in a looong time)
https://media.digikey.com/pdf/Data%20Sheets/Photonic%20Detetectors%20Inc%20PDFs/SD008-2171-112.pdf

The only real benefit I can think of for this is declaring the UVC LED products available on Amazon as scams.  Or not?  Seems likely, but don't know for sure since nobody's bothered testing them.

Solution 2:  Germicidal Power "probe": $350 photodiode + $225 BP filter + parts for whatever glue is needed to get a decent measurement using scope/dmm
Use 10mm x 10mm active area photodiode:
https://optodiode.com/pdf/UVG100.pdf

Plus a bandpass filter similar to:
https://www.edmundoptics.com/p/254nm-cwl-40nm-fwhm-25mm-mounted-diameter/22476/

I could be off my rocker, but I'm thinking you could reasonably measure the actual power of germicidal UV emissions with something like this?  Can it be done without a UVC test source with known output (reference) for calibration?


Thoughts?  Any better alternatives I should consider that hasn't been posted?

[Someone]

I could be off my rocker, but I'm thinking you could reasonably measure the actual power of germicidal UV emissions with something like this?  Can it be done without a UVC test source with known output (reference) for calibration?

Those ideas still can't measure the output power of the source as:
a) its spectral distribution is unknown
b) its spatial distribution is unknown

You really need to read some introductory texts on optics and/or spectroscopy and/or radiometry.

There are specific instruments to measure irradiance at a target surface, and more specialised tools that apply weightings for specific measures:
https://www.intl-lighttech.com/products/ilt2400-uvgi-0
You can't just magic something up out of parts unless you understand what the parts do and most importantly, what you are trying to measure. This thread is like the rush of people making radiation measuring equipment after the Fukushima Daiichi accident, you can measure "something" but its unlikely to be either what you think you are measuring, comparable to any specific units, or useful to assess safety.

[Smokey]

Seriously....  just stick to the easy problems like making an open source ventilator!

[james_s]

The easiest way is probably to sniff for ozone, the UVC lamps I have produce rather copious amounts of it. You can't cheat on the spectrum of a mercury discharge, the only corner they could cut on a "fluorescent" tube is to make the envelope out of glass rather than quartz. If it's glass it won't produce ozone so if it does you can be reasonably sure that it's producing UVC. Be careful as ozone is fairly toxic in high concentrations, along with the UV radiation being hazardous to biological organisms. Don't forget to keep your pets from being exposed too. It will also fade and degrade fabrics and polymers.

[NiHaoMike]

Re: the OP's UV intensity comparison problem, probably the cheapest hobbyist friendly option would be to time the 'bit-rot' of a UV erasable EPROM.  Program it all '0' then time till half the bits have become '1'.  However you mustn't apply power while the EPROM is being exposed to UV, so its going to require you to rig up some sort of servo controlled shutter to obscure its window while you are powering and reading the EPROM.

I remember an experiment that involved repeatedly reading an EPROM while it's being exposed to UV. I forgot how much UV was used, but the EPROM was constantly powered during the whole process.

The easiest way is probably to sniff for ozone, the UVC lamps I have produce rather copious amounts of it. You can't cheat on the spectrum of a mercury discharge, the only corner they could cut on a "fluorescent" tube is to make the envelope out of glass rather than quartz. If it's glass it won't produce ozone so if it does you can be reasonably sure that it's producing UVC. Be careful as ozone is fairly toxic in high concentrations, along with the UV radiation being hazardous to biological organisms. Don't forget to keep your pets from being exposed too. It will also fade and degrade fabrics and polymers.

Some UVC lamps are filtered to block the really short ozone forming UV.

[pipe2null]

In all seriousness, I am REALLY not trying to beat a dead horse.   

a) its spectral distribution is unknown
b) its spatial distribution is unknown

a)  AFAIK (please correct if I am wrong), it doesn't matter if there are 1 or 15 different emitted wavelengths as long as they are all within the germicidal region.  If there is emission outside the germicidal range, then it doesn't matter for this application.  There is the issue of ozone creation at <200nm wavelengths which is outside of the band pass region of the filter, but for simplicity can leave that one single "measurement" to a sniff check.  For wavelengths that get attenuated toward the edges of the band pass region, well, I'm still thinking on that.  The intended purpose of the measurement for this application is to establish the minimum W/cm2 I can expect from an off the shelf product that doesn't come with a datasheet, so error due to attenuation that results in a measured value that is lower than reality is on the the conservative side of the line.
b) Well...  Isn't this an issue no matter what you are using to measure, DIY or NIST traceable light meter?  Most (but not all) of the UVC sources that might be useful are cylindrical, so hopefully that will reduce the complexity a little bit.  You make a good point though, relating point measurements back to the overall output pattern...

You really need to read some introductory texts on optics and/or spectroscopy and/or radiometry.

I wholeheartedly agree.  If you have any recommendations on a technical ref book on the measurement of radiating sources/patterns/etc, especially if it includes the simplified equations for typical geometries (point samples -> total), I would appreciate it.  I ordered a technical reference for UVGI and its application to air and surface disinfection earlier this week.  It should have some basics on UVC source performance data, but I don't think it will have much on measurement.

Thanks for the pointer to the ILT2400.  Yep, that would take care of what I'm trying to do (at least once I get the tech reference info).  $3000 is a bit much for short term use, but I found a place that will rent it.  $420 for 2 weeks seems a bit steep, and I'll likely need it for longer than that, so the rental price is getting pretty close to the cost of the bandpass filter and fancy photodiode.  Hmm...

[LaserSteve]

Check your PMs, Pipe2Null.  I just sent you a link to overstock from a reputable optical filter manufacturer.

Steve

[Zero999]

So...  I'm thinking I need to split this into 2 different projects to consider, and then choose one to proceed with, depending on cost, benefit, and likelihood of success building this myself.  In either case, I am assuming that the measurements/readings would be obtained with a scope and/or cheap dmm, 'cause that's what I have  :

Solution 1:  Cheapo UVC detector: $30 photodiode + parts for ultra low bias current amplifier (haven't messed with op amps in a looong time)
https://media.digikey.com/pdf/Data%20Sheets/Photonic%20Detetectors%20Inc%20PDFs/SD008-2171-112.pdf

The only real benefit I can think of for this is declaring the UVC LED products available on Amazon as scams.  Or not?  Seems likely, but don't know for sure since nobody's bothered testing them.

Is there any reason why your application requires LEDs i.e. having many on/off cycles, free from mercury (by the way low pressure germicidal lamps contain tiny amounts of mercury), no glass to shatter etc.? If not, just use mercury lamps. They're much cheaper and more efficient, than LEDs, given the same power ratings.

[NiHaoMike]

Are there cold cathode or even capacitive coupled germicidal lamps available? The filaments are the main failure points of fluorescent lamps, and with no phosphor to degrade, a capacitive coupled germicidal lamp should last "forever".

[james_s]

There's no reason one couldn't be made however I've never seen one. Properly cared for cathodes (filaments) can last a very, very long time. If you are going for long life, avoid instant start as that blows material off the cathodes every time it starts. Better electronic ballasts have what's called programmed start, they preheat the cathodes then gently apply the high voltage to strike the lamp, a good quality lamp on one of those can last tens of thousands of hours.

I've used electronic ballasts designed for those triple tube compact fluorescent lamps to run small germicidal tubes with good success. I replaced the failed ballast chokes in the ozonator in my hot tub with one almost 15 years ago and it's still going.

[pipe2null]

Thanks everyone for the great feedback!

Check your PMs, Pipe2Null.  I just sent you a link to overstock from a reputable optical filter manufacturer.

Steve

Awesome, thanks!  I'm checking it out.

Is there any reason why your application requires LEDs i.e. having many on/off cycles, free from mercury (by the way low pressure germicidal lamps contain tiny amounts of mercury), no glass to shatter etc.? If not, just use mercury lamps. They're much cheaper and more efficient, than LEDs, given the same power ratings.

My purpose related to UVC is 2 parts:
Part #1) There are a lot of people out there desperately trying to hack together something to sterilize one thing or another.  There is a pretty long thread started by a guy who was trying to effectively sterilize the face masks his (medical professional) wife has to re-use.  I am not in any way an expert on these things, but I can work to find the right information and post it so those with no other choice but to hack something together might have a reasonable chance at succeeding.  I've seen posts where someone says to bake it at 60 degrees C for a while, but I found a study on SARS that required a minimum temperature of 75 degrees, so...
As far as the UVC LED products on Amazon go, I have been told by multiple people that it is not possible to produce a UVC LED light product with a useful LED count that can sell as low as $50 or $100.  If those products are scams, not only are desperate people getting cheated out of money, but the sterilization they believe is occurring is NOT, so, if they are indeed scams then the continued sales of those products actually promotes the spread of COVID-19.  I kinda think that's a bad thing...   

Part #2)  I had a crazy idea to use a powerful enough UVC source to sterilize exhaled air in psuedo-real-time from infected people who need to use CPAP/BiPAP/etc for one reason or another.  This includes (possibly healthy) people who normally use CPAP on a daily basis but have stopped due to warnings related to COVID-19 being expelled as an aerosol.  Theoretically at least, it might be possible to do this with zero filters, which are in short supply globally.  I should have a "textbook" on UVGI for surface and air sterilization tomorrow, so that should have most of the important info needed.  The problem is availability of UVC sources with minimal lead time, and which ones might be suitable.  I freely admit that it is Bass Ackward to design a "UVC light in a can" aka "bio reactor"  based on what UVC source "just happens" to be locally available instead of the other way around.  Which brings us back to products currently listed on Amazon.  And back to Part #1 since anyone attempting to use UVC for something important has the same problem I'm currently facing.


As for the current thread: I would appreciate recommendations on a "technical reference/textbook" that contains the right information to measure the output of UVC/radiating sources, preferably one that contains the "cheat sheet" equations to use for typical geometries so you can calculate total output based on point samples taken at specific positions around the UVC source.  Hopefully I will not end up needing it, but there is a lead time so I should order it asap.  Worst case it will be a welcome addition to my library.  Decades ago, I vaguely recall I got a degree in EE, but after many professional years only on the software side of things...  Well, saying that I have a lot of relearning and catching up to do is a gross understatement.

[james_s]

LEDs are pretty much a non-starter, at least right now. Currently nothing we know of comes close to a low pressure mercury discharge for generating copious amounts of UV. The lamps are essentially identical to conventional fluorescent lamps but without the fluorescent part, or at higher power densities there are high pressure mercury discharge lamps available up into the kilowatts. They are quite efficient at producing UV, one of the reasons mercury vapor lamps have been considered inefficient is that much of their output is UV which is not useful for illumination.

[helius]

In another thread on UV-C disinfection, a poster mentioned that the base of an EPROM eraser could be removed, transforming it into a shortwave ultraviolet lamp. I have one of the same units, so I did the same.
One of the things I noticed is that its ABS plastic housing fluoresces bright green under the UV-C rays. This is at least an indication that the proper wavelengths are being output from the tube. The rated power of the tube appears to be 8 W (the ballast is rated for 2, 4, or 8 W), and at 30% spectral efficiency that is about 2.5 W of UV radiation.

BTW, the discussion above about "instant start" vs "programmed start" is good. This unit has "manual start"! The power switch does preheat as long as it is held down, and kicks when released.

[Zero999]

In another thread on UV-C disinfection, a poster mentioned that the base of an EPROM eraser could be removed, transforming it into a shortwave ultraviolet lamp. I have one of the same units, so I did the same.
One of the things I noticed is that its ABS plastic housing fluoresces bright green under the UV-C rays. This is at least an indication that the indicated wavelengths are being output from the tube. The rated power of the tube appears to be 8 W (the ballast is rated for 2, 4, or 8 W), and at 30% spectral efficiency that is about 2.5 W of UV radiation.

If you have a known good lamp, then I agree the fluorescence of the plastic is a good indication that it works, but plastics often fluoresce at longer non-germicidal UVA wavelengths, so it's not a test to prove an unknown lamp doesn't just produce UVA and no UVC, unless you're sure it only fluorescences under UVC.

There's no reason one couldn't be made however I've never seen one.

Googling for electrodeless germicidal brings up a few microwave excited mercury vapour lamp.

LEDs are pretty much a non-starter, at least right now. Currently nothing we know of comes close to a low pressure mercury discharge for generating copious amounts of UV. The lamps are essentially identical to conventional fluorescent lamps but without the fluorescent part, or at higher power densities there are high pressure mercury discharge lamps available up into the kilowatts. They are quite efficient at producing UV, one of the reasons mercury vapor lamps have been considered inefficient is that much of their output is UV which is not useful for illumination.

Yes go with mercury vapour for UVC. By the way nearly all modern mercury vapour lamps used for visible lighting use phosphors to boost the efficiency and are therefore really fluorescent lamps, even though they're not known as such. This is evident from the fact they're opaque/translucent white when unpowered. See the example linked below:
https://www.bltdirect.com/venture-mercury-blended-lamps-delux-mbf-mbfudl-50w-e27?adcid=pla&gclid=EAIaIQobChMInZ_Kh5PS6AIVCPlRCh1EQwIDEAQYBSABEgIr2PD_BwE

Part #1) There are a lot of people out there desperately trying to hack together something to sterilize one thing or another.  There is a pretty long thread started by a guy who was trying to effectively sterilize the face masks his (medical professional) wife has to re-use.  I am not in any way an expert on these things, but I can work to find the right information and post it so those with no other choice but to hack something together might have a reasonable chance at succeeding.  I've seen posts where someone says to bake it at 60 degrees C for a while, but I found a study on SARS that required a minimum temperature of 75 degrees, so...
As far as the UVC LED products on Amazon go, I have been told by multiple people that it is not possible to produce a UVC LED light product with a useful LED count that can sell as low as $50 or $100.  If those products are scams, not only are desperate people getting cheated out of money, but the sterilization they believe is occurring is NOT, so, if they are indeed scams then the continued sales of those products actually promotes the spread of COVID-19.  I kinda think that's a bad thing...   

How about a non-electronic solution? You want to test it kills bacteria so how about testing it on some harmless bacteria, such as yeast? Culture a small amount on an agar dish, expose it to the lamp, then see how much it continues to grow afterwards.

Part #2)  I had a crazy idea to use a powerful enough UVC source to sterilize exhaled air in psuedo-real-time from infected people who need to use CPAP/BiPAP/etc for one reason or another.  This includes (possibly healthy) people who normally use CPAP on a daily basis but have stopped due to warnings related to COVID-19 being expelled as an aerosol.  Theoretically at least, it might be possible to do this with zero filters, which are in short supply globally.  I should have a "textbook" on UVGI for surface and air sterilization tomorrow, so that should have most of the important info needed.  The problem is availability of UVC sources with minimal lead time, and which ones might be suitable.  I freely admit that it is Bass Ackward to design a "UVC light in a can" aka "bio reactor"  based on what UVC source "just happens" to be locally available instead of the other way around.  Which brings us back to products currently listed on Amazon.  And back to Part #1 since anyone attempting to use UVC for something important has the same problem I'm currently facing.

That's not a crazy idea. I like it. There are small UVC lamps available. Something the size of a small neon indicator lamp would be good, but I've had a quick look and couldn't find anything that small. The smallest I could find is a 2W filament pre-heated mercury vapour lamp, which won't be that efficient, but still much better than LEDs. You should avoid lamps that produce ozone, which is bad for the lungs, unless there's some kind of filter to remove it.
https://www.amazon.co.uk/GTL-2-Screw-Germicidal-Light-GTL2/dp/B06ZZ8GH24

As for the current thread: I would appreciate recommendations on a "technical reference/textbook" that contains the right information to measure the output of UVC/radiating sources, preferably one that contains the "cheat sheet" equations to use for typical geometries so you can calculate total output based on point samples taken at specific positions around the UVC source.  Hopefully I will not end up needing it, but there is a lead time so I should order it asap.  Worst case it will be a welcome addition to my library.  Decades ago, I vaguely recall I got a degree in EE, but after many professional years only on the software side of things...  Well, saying that I have a lot of relearning and catching up to do is a gross understatement.

Sorry, I can't help you there. I'd just go with data sheets. If you're going to make enough of them, it might be worth talking to a manufacturer, but I expect they'll be very busy, given the current crisis.

[james_s]

Yes go with mercury vapour for UVC. By the way nearly all modern mercury vapour lamps used for visible lighting use phosphors to boost the efficiency and are therefore really fluorescent lamps, even though they're not known as such. This is evident from the fact they're opaque/translucent white when unpowered. See the example linked below:
https://www.bltdirect.com/venture-mercury-blended-lamps-delux-mbf-mbfudl-50w-e27?adcid=pla&gclid=EAIaIQobChMInZ_Kh5PS6AIVCPlRCh1EQwIDEAQYBSABEgIr2PD_BwE

Yes I'm well aware of this, there is a long history to phosphored mercury lamps, you can see the reddish or pinkish glow when they first start up before they warm up and appear white. The phosphor is not perfect though, only some of the UV is converted into useful light and the phosphor itself does absorb some visible light as well. You can be fooled too, I have at least one inexpensive Chinese mercury vapor lamp which is printed with the /DX (deluxe white) suffix but the coating is only a diffusing powder, it does not actually fluoresce.

The best lamp related website I know of is www.lamptech.co.uk. It's an excellent site all around IMO, very good information density, clean layout without any scripts. Lots of good info without any fluff.