The 7135’s may have been cutting out due to too much heat. In the following thread, HKJ’s testing showed the 7135 cutout out when it over heats. Under a graph showing current drop over 4v, he said:
“. . . The driver will reduce the current when it gets to hot (8 linear drivers on a small PCB do get hot very easily), as can be seen at 4 volt and above.”
Sorry but I am confused!
Vestureofbloods maglite dropins ran fine on 4 D Nimh cells at around 3 amps for around 50 minutes during his testing if i recall correctly…
Edit: he only ran it for 20 mins, but output didnt really drop in that time…
Did you just have a faulty driver? Everything else sounds ok. Its easy to say when its not my money being spent(!) but I say dont give up yet! You could try again with a new driver… maybe check it works correctly without fitting the extra chips first, to save time if it doesn’t?
AA NiMh and D NiMh are different animals. The open circuit voltage means very little, what matters is the voltage when loaded, and smaller cells will sag more than the big ones.
yeah I tried the same host and batteries with a different 8x board it was still buggy.
If it is a heat problem?….PilotPTK did some gerbers for some single sided AMC7135 boards that would have a chance of dissipating the extra heat. You could arctic alumina a board like that to your mag heat sink, and run it from a 105C as a slave. Pure speculation on my part :), but you could contact PilotPTK his experience?
The make of the cell will also play a big part in how much voltage sage the cell has. I have run 4D size Nimh’s to a amc7135 driver before with great results. But the cells where probably 5 years old and good cells of there time. I tested the cells under a 3amp load and found that they dropped so much voltage under the load, it was acceptable for a amc7135.
Might point is you need to consider the actual make and voltage sag under the load. 2 completely different brands of Nimh AA’s can have very different voltage sag characteristics. If they are quality cells and have little voltage sag under a load, 3 Nimh’s maybe best with a amc7135 driver. If they are cheap cells and have a high voltage sag under a load, 4 Nimh’s will work fine. I suspect something when haywire when you added extra chips. I have experienced problems when adding extra chips to mine several times, usually flickering in the lower modes.
Here are the cells I’ve used. It also flicked when using a different 7135 board with no modification. If I can’t use and xml 7135 combo any other suggestion on what could go in a 2 d mag.
I’m with nickelflipper. If it is heat, you may be able to better connect the chips to the heat sinking. I’ve used 7135’s with both 2S 18650’s and also with 6S 4/3F (Ni-mh that are about 18650 sized) with no problem. The chips were mounted on a piece of copper pipe separate from the pill. So they were nicely heat sinked.
EDIT: I used the above to drive a MG-T2. Sorry for not specifying that.
I once tested one of my 3D Mags with four blue Tenergy D cells, and instant POOF... no more XML. It was using a 8x7135 105C at the time (now has one of those "5A" V10+ drivers).
I'd suggest that given the working voltages with 4 cells, a real buck driver would be more compatible. Otherwise do you have a way to only partially charge the 1/2 D cells? That's less than ideal but would be enough to keep it alive.
How about trying it with 3 x 1/2D and a 1/2 dummy cell?
I tried that before in the same host with a xpg2 before I did the xml swap. It flicked with 4 cells and worked fine with 3 cells. I was hoping the high amp xml2 swap would eat up some of the extra volts. I guess I was mistaken.
2 of these or 2 18650’s with a buck driver seems to work great for me in a 2D mag. The 32650”s will give you almost twice the run time.
I used this driver and a XM-L2 on a copper sink pad in the defiant 3C super thrower, worked great with 2 18650’s. Pretty sure it fit a Maglite. What are using for a heat sink in the Maglite?
Granted, I’m no EE but both from what I’ve read and tests I’ve done 7135’s are capable of handling more excess voltage but not when crammed altogether on one miniscule board. The specs call for up to 1000^2 mm of trace area for each chip when burning off 700mW or more and there’s no way that there’s even a fraction of that for 1 chip let alone 8 or 12 on a 105C. If you want to push them hard then consider ways to isolate the chips from the heat generated by the led and provide more additional heat sink capacity for the chips themselves. It can be done but it’s a pain.
So I rebuilt this light again still with the 1/2D cells and 14x 7135’s. light out put is great but it still overheats in about 45 seconds. Please excuse the ugliness of the heatsink it was in adn out about 600 times and had no less than one burnt up emitter hacked out of it.
Damn that is one hulk of a pill.
I still say 4 big NiMh are gonna be too much. The 4v you have left when the cells are fully flat would be good for getting full output through the discharge but the 5.6v or so when freshly charged is just too much for a 7135 to deal with (note: not the total input voltage, but the difference between the input voltage and the LED's forward voltage, the difference between those two is what the 7135 has to 'burn off' directly to ground to keep the LED alive).
Also keep in mind vestureofblood's 4 cell lights were using his own custom programmed driver, and he may have tweaked the code somehow to have the high mode at something less than 100%, or done something else as a workaround.
Have you checked how it runs with 3 cells plus a dummy?
Perfectly fine on 3 cells. I’m probably gonna swap them out for some sub c cells
I'd keep the 4 cells and switch to a buck driver... there's enough meat in that pill to be bored out to take nearly any size flashlight driver on the market.
I wonder if adding a few more amc7135’s might help with 4 Nimh’s. Raising the vf of the led from the extra current and lowering the voltage difference. There should be more battery voltage sag too at a higher current.