The I just found out this vendor sells this driver directly from their website too.
http://www.topledlight.com/led-driver-input-3v18v-for-phlatlight-luminus-sst90-led-light_p1235.html
What is the overall diameter of the toroid used on this driver?
Best guess is about 1 inch. I will measure tonight.
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.
Thanks! I figure if worst comes to worst the LED wires can go through the middle of the toroid. Are you sure the doughnut is rubber & not ferrite ?
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.
Going to crawl under a rock now.
Heheh. I’m sure I thought they were made of something else for years, the colors are misleading. 
So yeah, frankly I do not have a handle on exactly what the core does, but it is there for it’s magnetic properties. It is my understanding that this is the part that ‘saturates’ under certain circumstances in our buck drivers, this causes unhealthy voltage/current spikes or maybe DD.
OK, want to splice in a Attiny13a tonight. With the meter, I get a 123Hz at the FET gate and as the top pin closest to the circle/divit on the MCU. Does it make sense that that particular pin is connected to ground via a 30,000K resistor?
Another question. I would prefer to pull the stock MCU off the board. I don't mind if I lose voltage monitoring (if it even has that). I don't think that should interfer with the bucking function. Anyone feel differently?
Here are my voltage readings (with 2S King Kongs). Pin 1 is at divot and then they count down the side and then over to the other side and then up (so clockwise).
1 -0- 10 .618V
2 4.86V 9 4.86V
3 -0- 8 4.85V
4 4.86V 7 4.86/4.87V
5 -0- 6 -0-
EDIT: Corrected pin readings.
So pin 10 can be used for voltage monitoring, if desired. I plan on pulling the MCU, and wiring an Attiny13a to pins the HX-1175b pads 8 (+ in), 3 (Grd), and 1 (PWM).
I’m having trouble figuring out exactly what’s going on.
I guess I’d just remove the MCU and then probe with the DMM to find out which pad is VCC for the MCU. Once I found that, I’d pull Pin1 high and low. If I can turn the buck circuit on and off that way then I’m good to go. Pin1 certainly looks like the PWM pin.
EDIT: As far as I can tell, neither Atmel nor Microchip have an 8-bit MCU with 10 pins.
Thanks wight. I tried what I said above without success. I soldered wires as follows:
- Attiny Pin 4 (Gnd) to HX-1175b Pad 3
- Attiny Pin 6 (PWM) to HX-1175b Pad 1, and
- Attiny Pin 8 (+ in) to HX-1175b Pad 7.
I flashed the Attiny with MiniMo.
I wish I had thought to probe the MCU pads after removing the MCU. I think I picked the wrong pad for + input as I was getting a very unstable voltage to the mcu. It would fluctuate from sub 1V to 3.3V. Could have been a bad connection.
It's late. I will take another crack at it tomorrow. I wonder if I have to have the voltage divider pin (pin 7) on the Attiny13a connected for it to work. I would think that would depend on the FW. I don't think MiniMo monitors voltage.
Yeah, I don’t think MiniMo does monitor voltage. If it did, you’d want to pull the voltage monitor pin high (connect to Vcc).
When you get back to it I imagine you should be able to easily correct the situation by finding Vcc w/ your DMM.
Note that you may get better dimming results with an extremely slow PWM freq such as what we used in the Knucklehead driver thread. I wonder if I ever got around to compiling a less heinous firmware configuration for that? Probably not. I wrote about what was wrong with the current compiles and posted the code in that thread.
Sounds like good advice. I need to check out your Buck Driver thread. It was just starting when I last perused it.
Didn't get but a few minutes to work on this tonight. Definitely got different readings without the stock HX-1175b MCU in place. Some of the pins that I read + voltage on were getting the voltage from the MCU. I don't recall the readings right now. rewired the positive feed, but it still didn't work. I should have tested the MCU before wring in to the HX-1175b. Hopefully, I get a chance to take a crack at it this weekend.
My advice remains the same, please try what I wrote in post #30. Don’t try adding your own MCU yet.
You speak wise words chief wight. I tried that and got no light. I'll recheck my connections and reperform tonight and go from there. Thank you for reminding me to try that. 
I'll also pull the MCU off a "b1" board and post a picture of the underlying traces. I keep forgetting to do that. The "b" does not show traces because it is black and I have most of the bottom copper clad.
I happened across a tidbit of info about what probably allowed the driver to work when you accidentally swapped a diode for an FET. I’m not planning on looking into the matter any further, because it was definitely a bad thing to do, but it seems like what Olin Lathrop posted in this thread on stackexchange may explain why it worked at all.
Thanks wight. I tried to read it, but I'm in a rush. Sounds like it basically did not let any current flow through it. I may read it again later, but like you, I don't intend explore the possibility again. I swapped back in the diode once you made me aware of the issues.
I certainly didn't mean to swap in the FET originally. As you can see in the pics above (Post 17), with the label ground off, the diode looked just like the FET that was next too it. It was a good think that the label was not ground off on the "b1" version and that you noticed it.
I haven't had a chance to work on the driver any more. I hope to soon, but I'm starting to get concerned about finishing my scratch build light.
EDIT: Added the word "not" to the third sentence. Swear I typed it the first time.
UPDATE: The mod in this Post has a problem that will cause an over current and maybe over voltage situation. Please see the corrected mod at Post 50 below.
Finally got back to this driver tonight and got some great news. Found a good spot to tie in a PWM feed. Used Dr.Jones MiniMo FW on an Attiny13A. The HX-1175b had the stock voltage sensor resistors, but ComfyChair's 07N02 FET. Also, the HX-1175b's MCU was removed. Everything worked reliably. Did have one unexpected, but not an unwelcome result. For some reason, current to the emitter on high was twice as high as stock. Using 3S average laptop pulls I got 9.7amps to the emitter on high. Had moonlight set at 5pwm. Don't remember the current reading but I think it was like .05 amps.
This was quick and dirty work. Here is how I connected it. The red wire connected nearest the toroidal inductor is the PWM feed from the Attiny. Not sure what the component I connected to is. I think its a buck convertor. I connected to the pin closest to the inductor coil. The red wire above it that is connected to the pad next to that capacitor is the VCC feed to the Attiny. Feeds a little over 3 volts (sorry, forgot exact number).
