just wanted to post this since im sitting home under a level 3 weather alert i found the board design on here but for got by who ? i dont no yet if it works but looks great
i used 380ma AMC7135,SOT-89 Package,
and sorry for the crappy phone pics will post better one’s soon
Interesting, the only one I knew of was texaspro’s Custom SkyRay King SRK Driver.
There might be a much easier option for squeezing more chips when designing the board. The amc7135’s can be pretty much stacked end to end because the back tab is the same as the center front tab. Every one laid in a row where each leg is in parallel with the next. Its not my idea, I just came across the picture not to long ago. The board could be made much smaller or more amc7135’s added on the same real estate. I thought it was pretty cleaver. Click on the picture at the very bottom of this page.
. http://www.swiatelka.pl/printview.php?t=10422&start=0&sid=1c6db3c77f1302ad15173585ada51f0d
You might be right, scaru’s does look more like yours.
About that link, I thought that was pretty cleaver, very easy traces and very efficient use of space.
If you used that layout on a 17mm board I wonder how many you could get on one board back and front.
An AMC7135 is 4.3mm long (pin ends) and 4.6mm wide (body), so 4 of them are a tad too big for 17mm. I think you could easily get 7 on one side.
BTW, here's my design of a 46mm 4*9 AMC7135 driver for King/Kung/UF-T90 etc.:
I hope the boards will arrive soon from the fab... and that they'll work out. It's my second PCB project (the first was a 17mm RGBW driver).
You sir DrJones are a very clever person. I wish you all the best on your new driver. I’m looking forward to see it in operation. What will the maximum voltage input of the driver be?
It's a typical AMC7135 setup, so it's the same as for the NANJGs, best used with 1s?p Li-ion. I didn't have a 2s->MT-G2 setup in mind, so there's no voltage regulator, but the zener mod should work the same (and battery monitoring wouldn't work).
Yeah, but heat is there anyway, most of the heat is from emitter, conducted via pill, and a fiberglass pcb in not especially conductive (for heat) so it’s not like you get better heat transfer if contact surface between AMC and PCB is bigger.
That makes sense only if/when temperature of AMC is greater than surrounding temperature, I doubt this is the case in small C8-like pills.
Anyway, slightly increased temperature may decrease AMCs efficiency but since you can stack much more of them in upright position, mA loss would be almost insignificant.
More chips for higher output current means each chip has a smaller share of the excess voltage to burn off. Also higher current means a higher Vf which means there will be less spread between Vin and Vf that needs to be dissipated.
The pcb is not enough to handle the heat from a couple of 7135, they need more cooling than that, at least when the battery is full.
When the 7135 reach 150°C, it starts to reduce current, this toke less than 30 seconds on a pcb with 4 volt from the battery in my driver test.
Trying to check for the current drop with an ammeter in series with the battery will not work. The voltage drop in the ammeter will reduce the heat in the 7135.
Put the battery in the light and use a luxmeter on the light is a much better solution, some of the brightness drop will be from heat in the led and some can be from the 7135 reducing current.
Yes but it will only stay in regulation for a fraction of the total runtime depending on the internal resistance of the cell as well as all other resistances in switch, springs, wires and so on. Put enough of those 7135 chips on and it won’t be in regulation at all due to voltage sag, effectively making it direct drive. The whole purpose of using a constant current source (linear driver) supplying the led is to get some runtime with constant output.