Looking for help - built a COVID-19 Killer, UV-C bulbs are not powering up.

Flashlight Modding and DIY Parts
21

I was not able to figure out how long you need to leave the mask in there to sterilize it. This Wikipedia page was interesting, probably a good read.

22

Source: https://techshop.lk/products/uv-c-ultraviolet-light-bulb-3w-253-7nm-uv-sterilizing-lamp

The above I guess it presumes 50 Hz mains, with some simple math half voltage means double capacitance (9.4 µF), and since your bulbs are 4 W instead of 3 W, the capacitor should be 9.4 µF × 4/3 = 12.53̅ µF. But since your mains rate is 60 Hz instead of 50 Hz, the capacity should be reduced a bit to 12.53̅ µF × 5/6 = 10.4̅ µF. This is the cheapest solution, and works because the bulb is being fed with AC and therefore always receives sufficiently high voltage each swing.

23

Thanks, so no need to worry about 25v startup to 11v constant. Will the bulbs adjust voltage accordingly?

24

Why are you not using standardized light tubes that you can run like normal fluorescent tubes with the right mains power ballast known from normal cft bulbs?


https://www.aliexpress.com/item/4000004427069.html


https://www.aliexpress.com/item/4000240853360.html


https://www.aliexpress.com/item/32655705803.html

There are 2 terminal and 4 terminal light tubes, that need different ballasts

Generally wiring those tubes parallel should not work, even your small ones as they does not ignite at the same time

25

Hey Trailhunter,

What you have there is in fact a low pressure mercury arc UV lamp. As a result, it's not straightforward to drive like your setup. In fact, while it's possible to run on DC, I'd highly recommend running on AC since the cathode of any gas discharge lamp will have significantly more wear and tear. The bulb works when you first apply a current through the incandescent filament. This filament begins to heat up and ionizes the mercury amalgam inside, and once struck, the arc begins radiating UV. At this point the resistance drops significantly and a ballast is required. Depending on the specific bulb you have, it may be necessary to have a slightly higher voltage to initiate the arc to strike. Likewise, depending on the bulb design, the initial pre-heating duration can vary.

It's possible to get working with a DC power supply and resistor, but you'll need to choose the appropriate resistor and make sure it can handle the dissipated power. I wouldn't recommend this because of the wear on the electrodes.

You can also get it working with a AC supply (I'm not sure what voltage or ballast capacitor required, depending on the specific bulb you use) - it'll depend on what bulb you have and what line voltage it is.

I'd recommend purchasing specific drivers for each of your bulb, or better, just get one of these or similar on eBay:

Unfortunately, it looks like it comes from China, so shipping may take a while. Other members have recommended other options.

Personally if I were building one of this, I'd go for something quick, cheap, and all integrated:

The only question is the quality of the actual tube. If you're concerned and want a quality bulb, you can get a Ushio G6T5 (6W) or G8T5 (8W) tube from Ebay for fairly cheap to replace the stock tube. Regardless, I think it'll suffice for your purpose.

Good luck and remember to wear eye protection (and skin protection).

26

Mmm no. The bulbs do not adjust voltage, they're the load. DC/DC CC/CV boost/buck/boost-buck modules do that. CC/CV is an acronym for Constant Current / Constant Voltage. Once set, both current and voltage are limits in these DC/DC converters. If the CC trimpot is set at 0.38 A, the module will raise output voltage until it meets the set current, 0.38 A, or until maximum voltage. Your Drok module works this very way!

Since your Drok module has a nice display, could you go find some 24 V supply at hand and supply power to the module with it? With a 24 V supply connected to the Drok module, and making sure the CV trimpot is fully turned up (the output voltage should be around 24 V, or close), turn down the CC trimpot all the way down, usually counter-clockwise, and connect one UV-C lightbulb at its output; just one. Then slowly raise the CC trimpot until you can read the desired current flow on the Drok display. This way the Drok will be set to drive one bulb with constant current. This way you can power down the thing, connect a different, cold bulb, power it up again and observe the Drok's behaviour in its screen. The voltage should raise, the bulb power up, and the onscreen output voltage will gradually dial down while the bulb heats up. The output current should remain around the set value all the time.

As I say above, to properly do this you will need one module per bulb. If you observe exactly the same behaviour with all bulbs starting at room temperature, setting them in parallel could work (raising the Drok output accordingly), but its risky. In a flashlight many people drive leds in parallel succesfully, but this is due to optimal thermal coupling, which you don't have with the bulbs. And if a bulb dies, the remaining will have to bear with the additional load.

Or a capacitor in series with mains AC for each bulb. Very simple and completely KISS.

P.S.: a 19 V supply like a laptop's brick may suffice for the above testing purposes. Very likely.

Thu, 04/02/2020 - 21:47

27

You violated basic BHL standards and UVC interior sterilization or not, this will be a magnet for cross contamination and thorough exposure. This, although only 85-90% certainty, is based on the Wuhan Virus remaining active and viable on surfaces within set temps and humidity. There’s a particular setup you have to do. And any light, even a tiny amount, is EXTREMELY dangerous with that bulb. Honestly, it was a cool move to do something that obviously took a lot of thought, gathering, and effort. But this device is officially dangerous my man. Otherwise, still a cool, even if unsafe, way outside the box….box.

28

What makes it light extremely dangerous with this bulb mean? This bulb is put into a home wall outlet, I dont see any recommendations on the box that it’s extremely dangerous? Will it explode on my face or something?

29

The danger is with the UVC exposure. You really need to make sure you take all precautions when operating the bulb because this wavelength is very ionizing and will easily cause eye damage and skin burns very quickly. This means absolutely no looking at the bulb without eye protection when it is operating, and the box needs to be light-sealed very well. At this radiation level, oxygen is also easily converted to ozone gas, which is itself very reactive and can be toxic above a certain (low) limit. Please take care.

30

Thanks brother, I have a bench power supply that I can test with, will give it a shot. Highly appreciate your feedback, very informative and valuable!

31

Ah yeah, i got that - very aware of the dangers. The bulbs dont emit ozone or at least, that’s what the manufacturer said

32

Quite probably.

I had some panel-mount neon-bulb indicators. Normally, they have the ballast resistor included, and are then marked for “120VAC” or “240VAC”, etc. These didn’t have any resistor inside. Hmm, no wonder they weren’t marked with any voltage.

Connected one to the mains, and the bulb blew apart into pieces and shredded the plastic “case” it was in.

Good thing I wasn’t holding it when testing it.

33

visible light isn’t dangerous, uvc is, but it is blocked in his box, even clear pc will block any uv below 400hm, i do not see any danger unless he runs the box opened

34

The bulbs don't emit ozone, but assuming they are true germicidal bulbs of about 245nm frequency, you will get some amount of oxygen photolyzation to form oxygen atoms which combines with O2 to form Ozone (O3). If you have this running in a box, ozone concentrations can build up, and ozone is highly reactive and is an extremely strong oxidizing agent. Just make sure the area is well ventilated.

35

Thank you for this awareness factor. Will keep in mind

36

You can put them in a oven for 30 minutes at 70 degrees Celsius to destroy the virus. https://iheartintelligence.com/stanford-researchers-n95-masks-sterilized-reused-low-temperature-heating/

37

Don’t use your home oven!

https://aim.stanford.edu/covid-19-evidence-service/

38

FYI, other methods that have been studied in the past for mask re-use:

If it does not have a metal strip on it, the microwave is probably going to be one of the easiest to implement.

Note that the page references different studies for different methods. None of them are specific to SARS-COV2, and many were tested against bacterial microbes, not viruses, but some of them did test viruses like influenza.

The suggestion someone in this thread made of waiting three days is also discussed in the link I posted…actually the CDC suggestion is for 5 masks to be issued to each worker.

39

the article talks about legal reasons, not that it wont work. assuming the oven can hold 70c accurately, gas ovens can not, but electric one can. some food dehydrators actually can get to 70c and hold it steady and accurately,

40

that is really simple solution, but does he have to run each bulb with its own capacitor? can he run 2 in series with 1 capacitor, or even 4 in series with 1 capacitor? or 2 or all in parallel with 2 or 1 capacitor? what formula would he need to use to calculate capacity,