Thanks ImA4Wheelr. We’ll see about DIY versions. The specs look pretty similar to the new-version DRY driver. From what I can tell they are using different buck driver ICs though. This thing is interesting to me from two perspectives; either it’s another implementation of the
Oh. I just realized that the ON Semiconductor 35CLG (big TO252 package) is not an FET. It’s just a big 10A schottky diode. duh. Therefore this circuit is definitely not a synchronous buck circuit. That explains why the connections looked weird, but it also makes this driver way less interesting - it’s probably no more capable than the QX5241 based ones. This is just scaled up with a big inductor and big Schottkey diode.
What I’m really looking for is an inexpensive high-power synchronous buck driver IC that’s suitable for drivers like this. That type of buck requires two FETs, but some non-synchronous drivers use two FETs as well. A synchronous buck setup can improve efficiency.
Out of curiosity, what would be your goals with a replacement PCB for this driver?
A more efficient driver sounds better as long as it can handle high current well. This guy (The “b”, not the “b1”) is driving a MT-G2 at at least 11.3 amps and not complaining. I don’t know the limit yet on it. I haven’t tested the “b1” version yet, but it looks promising.
I want a custom PCB because this driver may have already been discontinued and because I’d would like a MCU that I can program and flash. I’m going to learn how to program PIC’s, but most folks here like the Attiny13a. I will be piggy backing a Attiny13a on to my drivers in the mean time.
That’s interesting. I didn’t notice that was a diode either. When I swapped out the FET’s on my “b”, I think that one had the label ground off. If I remember correctly, I thought it was odd that they ground off one, but not the other. I replaced both units with 07N02 FET’s. It works great that way.
I believe I still have the old components. I will check them out to night and report back on that.
EDIT: I also have 4 more of these “b1”’s, I will check to see if they also have diodes in that spot.
Does that 35CLG diode make sense? It’s supposed to let current flow from either Small Leg to the Big Tab. It appears that both Small Legs are connected to Bat+ and the Toroidal Inductor that flows Bat+ to the L+ Pad. The Big Tab to the Drain for the 06N03 FET. The Drain for the 06N03 FET would be flowing Negative.
So wouldn’t it basically short positive to negative?
Sorry if this is a dumb question, but I have no formal electronics training.
At a glance it looks fine to me. I think you have gotten confused though. The small legs are connected to LED-. The big tab is connected to BAT+ and the torroid. So when you apply power and leave the FET switched “off”, no electricity will flow across the diode.
My grasp of this stuff is not good enough for me to explain the role of the diode in this circuit, but it is properly connected. I believe this diode is sometimes called a freewheeling diode and it allows the inductor to discharge it’s energy while the FET is switched off. When you look at the sawtooth graph of a buck converter, the period where the inductor discharges and the diode comes into play is the downward part of the sawtooth. Here is what I consider a very explicit diagram of a nonsynchronous buck converter. See Figure 1. I have not read the article.
I’m afraid that your replacement of the diode with an FET is doing something bad. I can’t quite figure out how the circuit is working without failure right now, but I suspect that it’s hard on one of these 3 things:
the FET that replaced the diode
the real FET (if the energy from the toroid isn’t being allowed to flow through the normal path then the FET is probably being stressed)
the LED (the MT-G2 can take some real punishment as I understand it. Maybe it’s taking a beating from the modded driver and continuing to operate)
I’d definitely put that diode back where I got it. Maybe another member can chime in on why the circuit works at all after the mod.
Thank you wight. You’re right. I switched the big tab and the small legs in my head. I will check out the circuits on my original “b” and the new “b1” drivers and report back. Sounds like you may have identified a problem I didn’t realize I had or the circuit on the original “b” driver is different (or this circuit is not totally known to us yet and the diode is supposed to be a FET).
So unless there is a surprise under the diode, it looks like we have a flashlight driver that can handle high currents (a rare and good thing), but not as efficient a design as what you are seeking.
Hhmm - give me a few more applications I can use this driver. Though I hate the classic Chinese 5 mode, the power seems awesome!
It will work for a single MT-G2, single XM-L2 I assume, from 1 to 6 cells, 2 min, for an MT-G2 I suppose? What other lights? I'll have to check my multi-cell, rear switch throwers to see if this driver will fit.
Tom E - I think the problem is being able to purchase it! It seems that people are often sent something other than this driver when they attempt to order.
Hhmm - I thought this link in the OP: ebay.com/itm/5pcs-5-modes-Led-Driver-Input-3V-18V-7A for qty 5 were for the b1's -- yes, ImA4Wheelr did confirm with the seller first, but those b1's look good so far. Right? $50 is a fairly big risk of course, but of course I've lost that much on a gamble before with drivers... If I were to order 5, I'd like to use like 2-3 off the bat, or give/sell at cost, etc., and check with the seller first like he did.
It looks good to me Tom, but I'm just a hack. I think he sells individual drivers too.
Here is what I did tonight. Removed the large diode and FET and compared the traces between the "b1's" and the "b". Every trace I can see looks identical. I tried to take photos of the "b", but the black pcb reveals very little in photos. I did get some good photos of the "b1" which I put in the beginning of the OP at full size. The toroidal inductors look identical too. Only difference I see is the FET's Source Pin resistor bay allows for much larger sized resistors, but underlying trace for both looks the same. I did not compare components though. I will try to do that tomorrow as it's late. As pointed out in the OP, I do see at resistor where there is a small capacitor on the "b".
Here are the 2 components I thought were FET's that I pulled off the "b". I can't tell if the one with the label ground off is a diode, but I think it is based on wight's analysis above.
Went ahead and swapped the FET to a 07N02 (Comfychair) FET and added 2 R10 resistors. Go 11.5amps in high mode and 5.x amps in Medium mode. There was no Low or SOS mode. Ran for a about 1 minute on high. Driver got pretty warm, but is was in free air and isn't pressed into a large brass ring like the T90 driver.
I'm sure I missed something, but I have to call it a night.
This driver with an attiny13a piggybacked would be pretty cool.
Wight, I've been searching for a long time for a good PWM dimmable synchronous buck IC...haven't found one yet that fits the bill. I have had a few ideas about converting some other board designs but as always time is an issue.
I'm thinking the large diode is going to be the weak link for very high currents. Last night after reinstalling it, heat was generated in that part of the driver (Running at 11.5 amps with 3S King Kongs). It will probably be fine if heat sinked, but it is only rated for 10 amps. I may piggy back another low voltage drop out diode on top of it for the time being. Anyone have a high current diode recommendations?
Thanks. I’m sure you realize this, but what I’m really looking for is the smallest diameter tube the toroid will fit into. So anything that can be sqeezed in doesn’t count, but measuring the shortest side also doesn’t help.
Diameter ranges .805" to .825" on the sample I measured. They're make of the typical copper wire around a rubber doughnut. So it's slightly compressible if needed. Don't forget to for the LED wires and a rubber/Kapton tape barrier (if you are like me and don't totally trust the enamel coating). Just in case, but I'm thinking you already have all that covered.
OK, I'm even more of a dufus than I thought. It must be some kind of ceramic coated/impregnated ferrite material. Magnet definitely grabs it. I take it that is needed for a magnetic field?. For some reason it felt squeezable. Guess it was just he copper wire giving way. Sorry for the mis information and thank you for catching that.
