Anyone interested in playing around with the PT4115 and coming up with a sub-15mm 1A buck driver?
I have seen that IC, my circuit design fu is lacking… that sucks.
Interesting IC, sad it is limited to 1.5A. I wounder if it could drive an external mosfet and a larger coil for virtually unlimited current capability?
It’s been around for awhile, Texaspyro used it some. My slowly growing ability to parse data sheets allowed me to uncover that it can be pwm’d up to 20k hz which is what is used by our Attiny mcu’s. I’m pretty sure everything needed can fit on either side of a 15mm pcb (hopefully 14.5mm). It’s a very simple circuit with only an input cap, shottky diode, inductor, sense resistor, and buck ic. All of that should fit on one side with an Attiny mmu and its assortment of parts on the other.
Fitting all of that on a 15mm driver might be possible but it would not be easy, just fitting the MCU and parts to make it work takes up basically the whole top side leaving almost no room for the buck parts. The coil in particular would be a tough part to fit.
I did see an instruct able where it was used to drive an external MOSFET for 2A or more output but even the 1-1.2A it’s good for on its own is enough wattage for a AA Minimag and an order of magnitude better than stock. I’ve no idea what changes to the inductor or other parts would be needed to increase current, I’m mainly concerned with it being regulated and drop in ready.
For a stock level 1-1.5A output I don’t really see the point over the existing 7135 based drivers like the TA15.
I see this more as an option for “real” buck drivers with output in the 5-10A range if it could drive an external mosfet.
The tiny10 fits everything but D1 on one side and I found a 47 microhenry inductor rated a 2A that’s only x 6 mm.
It must be using the DFN footprint tiny13. I don’t see a point to using the tiny13 myself unless there are positively no other options. Heck even the tiny25 is too small for what we are doing now days.
It would be interesting to see someone do it, just not something I see a point in doing myself.
Now a 17mm version that is able to handle say 5A+, that would be very interesting indeed.
Minimags are two cell so you either give up one cell or use 6V LEDs with Zener modded 15’s. 10-15A is a bit unrealistic for a light that size. Buck or boost is the only way to drive a 3V led in a two cell light. Eventually I’d like to see both options. When I joined the 17mm 4A buck driver was just a dream too. We may be a bit jaded but 1A still produces plenty of light for a small torch.
Good point, for that particular light it would make a good driver.
15mm? Nah. Match-mod the critter, use a D26 pill. 
Yeh, 17mm and lots of headroom up top.
There are lots of drivers well suited to particular applications and not so well suited to others. A Zener modded 5 x 7135 would certainly do the job with an xhp 50 but not be a viable choice for an xpg or Nichia. A while back I asked what drivers where missing from the selection and in some cases it was just a matter of size where in others a matter of type. Kd,dx, et al don’t seem to create new drivers, just copy what others produced, I’d like to give them something else to copy.
Trying to come up with something that doesn’t require any drilling or cutting of the host and that means putting the driver where the bulb holder used to go. There are a lot more people who are willing to deal with a drop in than a Match mod. To get there we need a driver and a pill. Mcpcb’s and reflectors/optics are a done deal.
Sure seems a waste PWMing a buck, but not a complete waste.
For the Texas buck we had a current contrallable IC, but that only worked with a PFET (NFETS are more efficient).
There's a thread here:
https://budgetlightforum.com/t/-/44347
With a buck IC that's designed only for hysteretic voltage control but it seems they used some kind of amp or comparator to put in a sense resistor instead and it seems it's probably doing non-PWM'ed current control. We discussed such ideas for NFETS with the texas buck, and I'm still not sure it isn't a better way to go. I think there's been a stigma on hysteretic control because RMM implies it's the reason for the ripple in his buck, but I'm not sure that has to be true. "Hysteretic" itself isn't the problem I think. Sensing on voltage output is the problem. I don't see why you can't sense on the inductor current though with a circuit similar to the one above (they don't), but with the sense resistor before the output cap. I don't think it would go into discontinuous mode, so that's a limitation maybe on how low it go before adding pwm.
There are probably many simple small IC's that could work well that way, but bucks are kind of hard. It's real analog electronics.
I'm sorry, that one is a PFET... but anyway, the idea might apply... converting a voltage controlled IC into a current controlled one, to get more options for NFETS. It's nothing remotely like a novel idea.
Whatever gets the job done and fits the space available is fine by me. This just seemed simple enough for even a non EE to get a grip on. Two other buck chips I know of, the ax2002 and qx9920 don’t pwm high enough.
Hm I'm not quite sure if I follow; so what exactly is the goal here? Are you hoping for a 15mm driver powered by 2 cells (AA or Lithium?) and what sort of power levels (how many A?) and output voltage (3V or 6V LEDs?) were you thinking of? I'm currently thinking of drivers to make for the next project so I'm looking around to find out what would be most useful for the community :)
Hi Loneoceans, was hoping you might see this thread but anyone with eagle knowledge and interest could work on this. Here are some rough specs.
14.5 mm OD with sub 7mm total height(board and both sides)
2- cell lithium buck driver
1A into 3V led
Attiny mcu(current firmware compatible), possibly 10 pin mmu.
Found this 47microhenry Panasonic 6.5 x 6.0 x 4.5m h inductor rated 1.8A. Series PCC-M0645M, part no ETQPM4M470YFN. There’s a 5.5 x 5.0 x 4 package at 22 microh and 1.9A. Link
The pt4115 has 6-30V input so it could conceivably drive 3V, 6V, or higher Vf LEDs but at only 1.2A max. Given the small size of lights it might go into 1A seems plenty and better suited to lower wattage 3V LEDs. It’s small with a minimum of external parts and has pwm input up to 20k Hz. The size constraints are those posed by the space occupied by the minimag bulb holder and the pill that would replace it. It’s a chicken and egg problem since there isn’t a pill yet but if the driver can be designed then the pill can be made. If the driver cannot be designed within those size parameters then there’s no point in making the pill. Years ago there’s was both a pill and driver, I have one of the pills but no driver for it. None of the current aftermarket led drop ins are even remotely close to this. All other mods require significant alterations to the host not available to many. Eventually it would be ideal to have two versions, one buck and one boost, either of which could fit and provide a drop in ready or kit mod for the thousands of minimags out there. Attiny compatible makes it possible that it might become available through one or more of the member run sales sites as a prebuilt option or kit with our favorite firmware preloaded. Gotta start somewhere though.
Played around with some paper cut outs on a 15mm circle and came up with this. 
After dinner I’ll fiddle with the mcu side but everything there should fit easily enough whether with Zener or LDO.