Good point on filtering, actually I would like to test the ZBW3 as I think it allows more light to pass than the 2 or the 1.
I didn’t do the soldering, it was done in China (probably by the engineer’s assistant) but it’s just temporary for testing different components.
That’s a cool looking light on the left! Love the switches. How many different emitters are in that?
Regarding the size of the pcb. It needs to be around 16mm diameter. I’m trying to understand what direct thermal path means in terms of the pcb. Does it mean it doesn’t have the dielectric coating on the emitter side?
For the driver/regulator. Any other ideas that would be cheaper than a step down?
I need something that costs cents rather than bucks. Of course the local Chinese wholesale price will be a lot less but I still think it will have too big an impact price wise on the end user.
Yes, this is roughly what I was thinking. I have room for one around 10mm square. Thanks
depending on the heat sink I think it would more than double the surface area exposed to the airflow. So that could make a significant difference. I’ll try a copper pcb and 600mA first to see what the temperatures are looking like. If still too high then heat sink is probably the next step.
Lookit the I-V curve of any diode. Tiny blip in voltage makes a huge blip in current.
There are thousands of cheap but crappy ways to regulate LED current, even a ballast resistor, or a bunch of good but pricier ways to do it.
Choose wisely.
Last line means that there are assloads of UV lights already on the market that are quite good already. Many quite excellent.
So what sets yours apart? Ie, what about it is going to make people want to pick yours over 1000 already existing “me too!” lights?
Burning off half the battery capacity to act as a hand-warmer? Mmm, maybe, in cold climes. But you do you.
Just saying…
Thanks, so a big blip in current is bad because it can burn out the emitter?
Here’s a pic of the main pcb that shows the two wires at the top going to the emitter’s mpcb.
I’m not manufacturing a flashlight. It’s a new novel product that happens to incorporate a UV function. Due to all the costs involved with the rest of the device, I’m only able to allocate a few bucks to the UV components. Which is why I’m glad Budget Light Forum exists!
the reason we recommend against lm 317 (if anything you want lm 117, it is 1.5A and probably cheaper than 317, you do not need 3A chip if all you need 700ma) because it regulates voltage, not current, like I said you can build a circuit with lm that will regulate but you need 2, (among other parts like capacitors, and resistors) one for current one for voltage, and both will need heatsink, voltage part will have to convert 7+v to 4, and those 3 has to be dissipated as heat, you will be basically wasting ~40% of your battery heating up heatsinks, and if they are inside, they will add to heat problems. you need a switching regulator, not a linear one.
Yeh, easily. UVLEDs are even more sensitive than white (ie, blue).
A 317 might not even have the headroom to regulate down from a pair of cells as the voltage drops. The 78xx series needs at least 2.5V, not sure about the 317.
I do like the hand-warmer idea, though…
So you are using existing board that has 2 wires already marked “led” so most likely it already has current regulation, have you tried connecting a led to those wires?
The reason we use drivers with leds, not voltage regulators, is those limit current, and led takes as much voltage as it needs, you can connect it to 100v but if you limit current the led will only pull needed voltage, and work just fine. despite 96v more i n the line than leds need.
Wait, do you think you’ll get away with just a star aka mcpcb for 600ma? lol, nope, you’ll fry the led in under a minute, 60ma is all a copper direct heat path star can take. my led stars (copper mcpcb with direct path) are sitting on a 1/4in thick AL base, about 3in in diameter, and only 1 or 2 work at the same time.
Sorry for not being clear. That board was custom made for my device. It’s hard to communicate with the factory, especially as I myself have so little knowledge of electronics. Which is why I really appreciate the advice here and hope to learn as I go.
But yes, from what I understand the current on that specific board is adjustable and set to 435-440mA for the LED. It is connected to an emitter and mpcb similar to the first picture I shared.
It seems to work fine in short-term testing that I’ve done but I would like to try to increase brightness while also managing heat that comes with any extra current.
I’m a bit confused. The first couple post shows the LEDs that were tested for five minutes each at 700mA.
Edit - Also my convoy s2 is on a 16mm board and is running at 1050mA?
Hand sanitizer AND warmer all in one?!
You mean you ran those 2 leds on the pic, on those non direct heat path boards, with nothing under them to take the heat, at 700ma for 5 min??? did you actually measure the current? in my experience they would not last a minute under those conditions.
Some convoys ran on 3040ma, but their 16mm star bolted or pressed firmly onto the AL body of the light, some thru a brass pill, some sit directly on the body, which acts as a great radiator/ heatsink.
The factory ran the tests using those two pictured emitters. Note that they were not tested inside the body of my device so they were open to the air.
One of them is very budget, about 30 cents and the YingFeng one costs about $1.05. if you buy 1000.
At 700mA the cheap one got much hotter. My preference is to use the YF even though it costs three times more. I don’t trust the ultra budget one.
Here is the data resulting from Ying Feng test:
Voltage: 3.2 - 3.5
Current: 0.691- 0.721
Power(W): 2.245 - 2.392
Risistance: 4.691 - 5.056
E_Quantity(kWh): 0.047
E_Capacity (Ah): 15.478 - 15.537
Max temp: 72C
edit - I get that the convoy S2 has much better heat syncing ability being made of aluminium. Which is why I’ve resigned to a lower current and taking other steps to control heat.
second edit:
One correction as two tests were performed. These changes were made on the second one:
E_Quantity(kWh): 0.051
E_Capacity (Ah): 16.964
Max temp: 82C
check continuity between middle pad and back of the star.
The factory confirmed that there is a direct thermal path
OK, but what does your test confirm? did you check continuity?
