You crack me up On Fri, Jun 2, 2017, 5:58 PM <[email protected]> wrote:
> Crap, 324 watts input. I guess the LED driver efficiency is a bit less > than > I was expecting. But 324 vs 1000 watts per fixture, I will take the 67.6% > power reduction any day. > > -----Original Message----- > From: [email protected] > Sent: Friday, June 2, 2017 4:54 PM > To: [email protected] > Subject: [AFMUG] LED project > > Made some significant progress on a high bay metal halide LED replacement > project today. > Can you tell which light is mine in the photo? If you look on the floor > you > can see some labels we put down for a grid. I measured the LUX at each > grid > point with metal halide and then with my LED. > > If I integrate all the points on the grid, my LED is only 16 LUX lower but > at less than 30% of. The 300 watts is the max rating of my LED drivers. I > have not actually measured the input power yet, but I am pretty sure it > will > be lower. > > In an industrial setting you get hammered for your demand charges. The > highest power draw in a 15 minute window each month. I pay $14.62 per kW > demand. My warehouse burns 24kW just for the lights. So just turning on > the lights for 15 minutes one time in a month the bill is $350. Plus the > energy charge of $158/month for running them 8 hours a day. > > This will take 16.8 kW off the demand which will save $245.62 each month. > Energy charges will be reduced $110 each month for a total savings of > $355.62/month if I convert all the fixtures. Not sure what it will cost me > to make each fixture but I might be able to sell them for that or a bit > more. Once done, if I turn this into a released product and someone > converts a warehouse like mind, the payback will be 12-18 months. > > I am pretty jazzed. I have some heat issues. Right now I am using forced > air with muffin fans but we all know how reliable those things are. I hope > to have a heat pipe version working soon. No moving parts. Silent. > > Then comes a dimmable version, a network version, motion sensor versions > etc. I can be pretty busy on this if I choose to do so. I am envisioning > a > phone app where workers can turn on a light in the back corner if they need > it and either turn it off when they are done or it will time out after 30 > minutes or some such thing. We can do time of day programs etc. > > Best of all, my old metal halides are the type that take 15 minutes to come > back on if there is a power bump. These will be instant on. > >
