> > The problem is that LEDs are really current driven devices. Go over the > current and the LED either degrades quickly or pops due to heat. And that > many LEDs in a sealed box will generate heat raising the ambient > temperature. > Good point, I did not consider this at all. Traditionally, people think LED's do not generate much heat at all, but it is still a conductor that has some resistance and has current flowing, therefore heat will be generated. > > The normal choice with just resistors is to design for under the max > rating of the LEDs. But the LEDs may very from device to device by as > much as a 10th of a volt and as they get warmer that changes too and in > the wrong directly. > > Say each UV light dropped 3.2V at 20 degrees C. If you put 3 in series > that's 9.6V. If your supply is 12V then you need to drop 12.0-9.6V= > 2.4V. Assuming you want 0.02A current you then need a 120R resistor. > If they get warm and their voltage drops to 3.1V then you need > 12.0-9.3=2.7V. To hold it at 20mA 135 Ohms. > > But with 75 Ohm 2.4V/75 = 32mA which may be above the device rating. And > if the temperature drops the LED forward voltage to 3.1V the 75R resistor > will increase the current to 35mA. That makes things worse and it > thermally runs away and pops your LEDs. > I am blindly following an Instructable, so that may be my first problem. The blindly following part, not necessarily the Instructable part. I am going to rethink some things based on your notes here (thanks by the way!). I think the box may work for a couple few weeks, but will most likely fail as you suggested. I want a long term UV exposure box with consistent results, so I am going to learn more about this based on what you've taught me here. > > If the voltage isn't stable but can rise up to 14V then the same problem > exists. 14-9.6V is 4.0V divided by 75 Ohms results in 53.3mA through > the LEDs. Then you'd need a 200 Ohm resistor to hold the current to 20mA. > > There are lots of constant current LED driver circuits on the WEB. Or you > can buy them premade. A simple version uses the LM317T as a constant > current source. Configure it for 20mA and use a higher supply voltage > like 36V or so. Then run 10 LEDs in series. > > I have a circuit that I designed for lighting up the cables on the Lions > Gate Bridge in Vancouver. It has 32 LEDs in series and runs off 120VAC > and there are 6 of those in each light fixture. Total power consumption > is 15W per fixture. See my home page link below. > > Again. thanks for the free education. I am going back to spice for a bit to do some work and prove out the design. Basic electronics cannot be overlooked! Chris > John Dammeyer > > Automation Artisans Inc. > http://www.autoartisans.com > Ph. 1 250 544 4950 > > -----Original Message----- > From: Homebrew_PCBs@yahoogroups.com [mailto:Homebrew_PCBs@yahoogroups.com] > On Behalf Of Chris Kleeschulte > Sent: Monday, September 26, 2011 3:50 AM > > To: Homebrew_PCBs@yahoogroups.com > Subject: Re: [Homebrew_PCBs] Two Videos demonstrating the Kinsten UV > Exposure Box > > Tom: > > Your questions lead to me to believe that I have no idea what I am doing > with this UV led exposure box. My design is not using a driver chip..just > a > 75ohm resistor to drive 3 leds. This circuit will be in parallel with many > other ones just like it. Would I need a driver chip? To secure the > artwork, > I was thinking that the glass plates would "sandwich" the artwork against > the glass plates. Would there be a problem with this? > > On Sun, Sep 25, 2011 at 9:23 PM, Tom Biery <judsquare@... > <mailto:judsquare%40yahoo.com> > wrote: > > > ** > > > > > > Are you driving the LEDs with a driver chip like a ZXLD1362? > > > > > > > > [Non-text portions of this message have been removed] > > [Non-text portions of this message have been removed] > > > -- Chris Kleeschulte KK4DYJ [Non-text portions of this message have been removed]
Message
Re: [Homebrew_PCBs] Two Videos demonstrating the Kinsten UV Exposure Box
2011-09-26 by Chris Kleeschulte
Attachments
- No local attachments were found for this message.