I have been looking for something like that. I see that it may have a sense resistor that could be adjusted. Would be a fun one to play with in a light that has the option of one or 2 batteries.
Once I get this installed and running with better heat sinking, and find it runs steady at 2.2-2.5A, how would I go about lowering the current to 2.A.
It’s probably fine as it is, but I wouldn’t mind playing with some options. Like I said, once I deal with the heat sag from my quickie tests, I may be dealing with a slightly higher current.
If R100 is a sense resistor, one would increase the resistance by a small amount. I would use trial and error, there are some here that could give you a good idea how to calculate it. My interest lies in it from the standpoint of a 2x18650 red rifle light as well. Something like a C8 with extension tube.
How do you think a red XP-E2 will handle … 2.5A?
I have not pushed that far yet, mounted on copper with a good heat path... maybe? I have one running 2.1 amps, did not have more emitters so did not push any harder.
Are you measuring tail cap current or emitter current? The driver could be eating the extra current.
I’ve just got it wired up. I’ve got my DMM spliced into the LED neg wire.
That’s a lot of current to eat up. I’ve tested it at 12V.
At a glance it looks like the venerable and common QX9920 buck IC, same as used by Solarforce and many others. Here is an example of SF’s application with pictures and some small amount of discussion.
I think I built a flashlight like that. Red XP-E2 on a Noctigon with a 7x7135 driver for around 2.45A. Seemed fine to me. See here: djozz - blue and green XP-E2, output tests
Single cell to single emitter it might have been ~10-15% driver losses but certainly from 2 cells you would expect Ib to be closer to 1A. In any case it will be less efficient as the voltage differential increases.
I use this for 1.5A
I’m getting the 2-2.2A readings at the emitter, from the Neg. wire between the LED and driver. Reading between the cell and driver is 1A with two cells.
Thanks wight, and glad to see you’re back. I’ve been gone for a long while myself and need to shake the cobwebs out. I read the link you gave for the QX9920 and the SF. That board is two sided, whereas mine is only one sided and probably more basic. I realize it’s the nature of the beast for a buck driver to give off heat, and generally they have double sides or two boards with larger toroids, so I’m just wondering if is should worry about the heat this little driver puts out, or just ignore it and let it do its thing. I mean bad things can always occur in any situation, but do you foresee any real problems cropping up using two CR123s with this driver. Is the heat it produces really a big deal, or should I just “let it bleed”.
My cobwebs are in full force, so I might not always give the best advice. I don’t think you should worry about it based on your description in the OP. If you can keep your finger on it indefinitely you are OK. 2xCR123A primaries is probably this driver’s ultimate sweet spot. They should be around 2.5v each at 1amp for 5v total input - That’s plenty to keep the driver regulating for a red LED, but not too much extra. Very efficient.
FYI I wouldn’t say that SF’s board is really a more complex implementation of QX9920. Note that the PCB you see is around 15.6mm and uses a contact board. Their inductor has around an 8mm diameter base. The back has an MCU and stuff if it’s the multi-mode version.
Okay, back to this thread. I was having problems understanding my tail readings so I started a Separate thread here.
Now I’m back to heat. I looked closer into the heat coming from the driver. It’s coming from the R100 resistor. I powered the driver with one and two 18650s and two CR123s with my fingers clamped down on the resistor. It’s seems to get equally hot no matter the cell configuration. Real hot. Can’t keep my fingers pressed down on it. How much heat can a resistor bleed off? Is it going to fail, or are these tough little buggers. I built the light anyway, and left it running for quite a while. I’m sure that resistor must have been really hot, but the light didn’t fail.
I don’t hunt, but these red lights are for hunters and I’m not sure if they’ll have them running for extended periods, or just for one to five minute bursts.
Resistor package size determines the power they can handle.
That resistor should be the load-side current sense resistor and will see the same current, regardless of the input voltage.
At 2.2 A, the R100 resistor will be dealing with 0.484 W.
A 0.5 W resistor will too close and will get too hot. A 1 W resistor will do better. It also depends on PCB layout, fat tracks connecting the resistor helps to sink some heat.
Maximum rating for resistors is usually much more than 60 C (60 C is about the limit to keep your finger on it), but are linearly de-rated above 70 C. So at 100 C, a 1 W resistor may be only rated for 0.65 W, for example.
Random google hit of resistor size/wattage:
I suppose a driver like this based on ATtiny 13 does not exist? Otherwise all would have been sorted? Richard has this, which is what I’m currently using, but its not cheap. ATtiny 13 is a must for me for custom mode setups.
There is room on the board for an extra resistor soldered along the side of the r100, so if you are worried about the heat, you may swap the r100 out for two parallel r200's.
Do you mean remove the R100 and squeeze two R200 in its place? Would stacking an R200 on top of another be the same.
Will this change anything else in the drivers operation.
Let me know. It’s early morning here, and I have the day off. I’d like to try this first thing…after I change the cats litter box.
The R100 resistor seems to be a 1206 form factor (size: 0.126 in × 0.063 in || 3.2 mm × 1.6 mm)
These resistors are rated for a typical power rating of 0.25W, although if they are connected to the PCB using thick traces they can handle more than that.
Assuming the R100* resistor is the current sense resistor, about 2.2A should flow through it, we can calculate the power which will be:
P=(I^2)*R = 2.2^2*0.1=0.484W, as DEL mentioned in the above post.
*R100 stands for a (0.1 Ohm resistor)
0.484W is higher than 0.25W, but the traces seem to be thick and the back of the PCB is covered in copper which should help with the heat dissipation.
Resistors are usually pretty tough and will handle temperatures of more than 100 C.
If you want to lower the output current, increasing the resistor value will give you a lower current, usually the resistor value is proportional to the current.
For example, a 0.15 Ohm resistor will give you (0.1/0.15)*2.2=1.47A.
Doing this will also lower the heat dissipation to 0.32W