EDIT: it turns out hauling my mockup of a 4x8 board alone is a huge PITA that I opted for something smaller, like a 3' x 5' corkboard on a canvas easel...plus I have a buckpuck in my drawer which makes the plan seem much more appealing
the LED adjustment pot unfortunately is a 25 turn 5K that sets really fine voltage increments...according to my multimeter each rotation increases or decreases the voltage by 0.003! [no wonder several turns down the road and neither of the LEDs respond tell me they are brighter or dimmer
] Eventually I set it so the green LED receives 2V flat, the yellow and red are now seeing 1.7 and 1.8V respectively. Green LEDs have much higher of a Vf than the other two so it looks starved of power when lit, yellow and red are quite bright even though I bought "standard power" LEDs. I haven't bothered to disconnect the LEDs and see how much current flows through them but I'm quite sure they won't heat up for the life of my fixture:thumbsup:
I may not be seeing something, but... if I understand correctly, you have the LM3914 connected so that it will disable the C-C Driver when the battery drops to 10.5Volts. Correct?
A 10.50 Volts - this is considered 100% 'dead'.
I built it slightly different than the drawing initially laid out, the LM3914 is independent of the CC driver's function, in fact I tested the driver without the buckpuck soldered in place
Its an indicator of its own, completely independent of the driver, drawing only 10ma it will be essentially an OCV meter on a 18AH battery [my jumpstarter, SLA] and even more so on my 40AH car battery [not sure if its SLI or Deep cycle]
EDIT: I just realized with the buckpuck the input output differential is 2V, which means if I'm pushing a 3xLED load of about 9.9V it should cut itself off at 11.9V...that was a totally unrealized combination of redundancy. :laughing: It will still be of value as I am going to use header sockets to mount the buckpuck so when the time comes in which a boostbuck is necessary I can swap it out easily
EDIT: I decided to use just 2 LEDs and hard wire the buckpuck. If the time comes to use a boost puck I'll build a identical one
Due to the limitation of my power supply I cannot make fine calibrations, I will be borrowing a friends power supply to update the LM3914 settings to a more realistic 12.6V+ green, 12.0V red, setup. I'm pretty sure 11V may damage the battery, 10 would really be unacceptable.
It might be better to set the LM3914 to LVD-cutoff at around 12.50 - which is 50% dead.
+1, I hope to set it somewhere near there too, just in case where I am I may have to wait a bit before recharging can take place. The general rule of thumb is never let the SLA drop below 60% capacity. My current plan for the buckpuck is to drive three XREs dimmed to a certain level [<350ma], at 12V assume 90% efficiency would mean it'll draw 320ma from the battery at 350ma output. It should take awhile for a charged SLA to drop below 12.5V [point where the red led goes out].
Since the LED driver part has alot of loose parts waiting revisions, I decided to make an identical one, calibrate that, then mirror things over using my digital multimeter by measuring resistance across the trimpots then aligning them to the matching trimpot in this driver. I had an extra 3x2x1 radioshack enclosure I had kicking around and figured those parts I bought for spares oughta be put into good use.
The original design planned for 3mm LEDs, apparently my queue list is still on the refrigerator so out goes that plan
the 13.8-15V, 12V, and 10V are the only output modes I get on my power supply, so for starters I went with it. I asked around for some advice on battery levels and I got two different answers: 12.7V for 100% and 11.7V for 0%, 12.6V for 100% and 12V for 0%. I'm pretty certain 12.6V+ for upper and 12V for lower should encompass most things. The switch was left out to ensure that the red light be seen when the battery gets low, but chances are I'll shut things down if it drops to red:huh:
