Let me start by saying I’m no electronic engineer, just a persistent tinkerer. Take all this info at your own risk.
If there is anything you might add to this feel free to do so or correct me if I’m wrong. I can’t just look at this and know what it all is. I spent many a hour reading and studying this design and components just to feel like I maybe understand it.
SWM uses a Pam2803 with a Fet (AO3416) and a 316BCG (MLX90316) Rotary position sensor.
Datasheets Pam2803 AO3416 MLX90316.
The pam2803 has been used here on blf alot. Wight designed a board similar to what SWM used but instead of the attiny13 to control brightness they used the 316BCG rotary position sensor. https://budgetlightforum.com/t/-/31239
Thanks wight for giving a lot of insight into this IC, it helped alot.
The way I see it, its basically direct drive controlled by a rotary position sensor tied to the gate of a FET if the battery voltage is over the vf of the Led. If its not the Pam2803 Boost the voltage.
Here you’ll see the driver as it came out from the light. This is a two sided board, this is the side towards the Led.
A few minutes of scraping and alcohol.
The components we are interested in on the Led side of the driver.
Most of the stuff to the right of the BCG316 it uses to operate the rotary sensor. In the Pam2803 datasheet you will find three components that are in series with the Led. The inductor the schottky diode and the sense resistor.
These all have resistance that’s in series with the led. Reducing this resistance allows the led to see more current without burning through components or the pam2803. You can still just change the sense resistor (FB) value to gain current but at some point it becomes to much for the components in its path. It comes stock with a R150 .150 (0805) ohms sense resistor. Changing that to .100 ohms is doable for a small gain in output. The sense resistor is on the battery side of the driver.
In the pam2803 datasheet it also describes the capacitance needed at high currents to reduce the output ripple. 10-22 uf is whats recommended in X5R or X7R type capacitor.
The SWM driver comes with one 10uf (guessed value). We need to change that to 2 10uf X7R capacitors in parallel to give us 20uf on the output side. I removed the existing capacitor and scrapped a little of the masking trace off beside the output capacitor pad to solder on a second capacitor. I wanted to use two brand new capacitors going back on that I knew were 10uf.
Two X7R 10uf (0805) soldered in parallel for 20uf to reduce the output ripple. Some people reported the output jumping around when setting perfectly still, I suspect this may have been where the problem was at. Bad capacitor or just wasn’t enough to reduce the output ripple.
The inductor can be changed for a much larger one with a lower DCR and still fit if you sand or file slowly the top of the inductor. Go slow and watch and feel for the core, its much harder and starts to show if filed down to it. Its a lot easier than it sounds but you will need that extra clearance to get the driver to lay down flat inside the pill area. I also use a drop of Arctic silver on top when installing the driver back in, helps dissipate the heat from the inductor into the head of the light. In this pic I had’t figured out yet that sanding the inductor top would gain just enough to make everything fit right. I have also already added a better Schottky diode in this pic. I went with the Schottky diode the pam2803 datasheet recommended and seem to work well. I tried the BAT60A I got flickering at higher currents and the MBRA210LT3G had alot less resistance. It’s bigger and not easy to fit in such a small space but it’s doable.
Now for the crucial part, the sense resistor, I used two in parallel (stacked) and went up in size to 1206. This helps with handling the extra wattage. You’ll need to check for this if you change to 1206, the aluminum ring that sets over top of the driver and holds the driver in with a screw, can hit the resistor if stacked. What will happen is you just have high mode and the control ring will not change a thing. If you experience this look for a component touching the aluminum driver ring. Ask me how I know :FACEPALM: . Its best to keep the stack as close to the brass battery stud as possible without soldering to it.
I used two .100 resistors stacked parallel for value of .050 sense resistor. Ends up at about 2.25 amps to the led measured on a bench power supply. This seems to be about the best for reliability. Makes about 800 OTF lumens with a LH351D 4000k at start up and around 700 OTF lumens after 30 seconds. I have used a .100 and a .050 ohm resistors stacked for .033 sense resistor. Increases current to the led at 2.5 amps at start up and a little above 900 lumens OTF with a little over 800 OTF lumens at 30 seconds. This is currently the way I have one V11R setup. I have used it for over a week or so and everything seems to be working fine. Even let the thing run continuous while it was stuffed in the lumen tube watching to see the decrease in lumen output. It holds the lumens fairly well for a 16340 light. The limiting factory in this setup is the battery. When tested on a bench power supply and the input voltage close to 4.0v the current to the led is around 3.5 amps and the driver doesn’t like that current to much. But with a Trustfire IMR16340 (best 16340 I have tested) the voltage sags to around 3.6v right from start up and falls quickly until around 3.5 where it slowly falls until depleted. So the led never can see more than 3.6v for its vf. I even tried the Efest 14500 and got pretty much the same results.
This mod increase current using 1 or 2 AA’s as well. When the voltage is low say 0.9v there’s not hardly any increase over factory but once it reach’s 1.2v input there’s a 50 ma increase. At 2.5v input there’s a 400 ma increase. Outputs almost 1 amp at 2.5v in.
The components I used are all from Arrow Electronics because their cheap and offer free overnight shipping.
Inductor SRP5030CA-4R7M https://www.arrow.com/en/products/srp5030ca-4r7m/bourns
Schottky diode MBRA210LT3G https://www.arrow.com/en/products/mbra210lt3g/on-semiconductor
Capacitors C2012X7R0J106M125AB https://www.arrow.com/en/products/c2012x7r0j106m125ab/tdk
The sense resistors I’ll leave up to you.
I bought a few of these just encase I let the magic smoke out and cheap enough.
Pam2803 PAM2803AAF095 https://www.arrow.com/en/products/pam2803aaf095/diodes-incorporated
I intended to make this short but that somehow doesn’t ever seem to work out for me. :FACEPALM:
This is not a easy mod replacing all the components listed, the inductor and diode just barely fit but can be filed a little to make room. Hope this helps and if I messed up somewhere by all means correct me. I’m here just to help others in this fun hobby and learn along the way.
In my long search to figure out how people up’ed the current on this driver (never found much info) I soon realized why. It’s kind of complicated and takes a long time to explain. Changing the current sense resistor just doesn’t cut it with this one.
Let me start by saying I’m no electronic engineer, just a persistent tinkerer. Take all this info at your own risk.
Reserved for Lighted tail switch mod.
The almost non increase when lowering the sense resistor indicates you are running the regulator way outside safe operation levels and more like DD the thing
Yelp, and all 16340’s have to much voltage sag to matter.
Now that I have had a few months to play around with this modded light, everything seems to be working fine even with a Olight 16340 10C battery. The best lows of any light I have and a decent beam with a LH351D 4000k 90 cri using the stock reflector.
Jon_slider had problems with the lumens fluctuating when he rotated the light as seen here. Sunwayman V11R on Ebay, $18.99. 2 left. - #270 by jon_slider
With all 3 modded V11R’s I have, none experience this issue when rotated in my lumen tube. I don’t see more than a 7 lumen difference on high while rotating the light 360 degrees and this could mostly be the slight change in angle when trying to hold the light straight up and down while rotating. In my lumen tube if the light is tilted just slightly down towards the tube path the lumens reads slightly higher, just a few lumens. So none of my modded lights appear to experience this issue.
Very helpful instructions and pictures!
I have receintly acquired a Sunwayman V10r Ti and replaced the LED with my prefered Nichia 219b sw45.
I, of course, neglected to measure the drawn amps before the LED replacement, but with the 219b the light draws 1.82A at when it turns on.
It gets warm soon and hot after a couple of minutes. I feel it could go up to 2.2A safely, but the thermal properties of the light, especially being Ti, do not seem good enough to dissipate the heat from more current going through the not-very-efficient 219b.
moderator007, do you remember the Amps on the unmodded light? I have ordered the BOM you are suggesting, with different capacitors, as the ones you are linking are not available (The only difference is the capacitors you were listing had 20% tolerance and I got ones with 10%). I also have some 0.1ohm and some 0.082ohm resistors rated to 1w. Would 1w be enough or would I still need to solder two in paralel? I do not want to go too low on the resistor, as I do not want to exceed 2.2A.
I am a bit comprehensive as I do not know if I will be able to find a replacement circuit board if the original gets damaged beyond my ability to repair.
My V10R untouched everything factory light draws 2 amps from a vapcell 16340. From a fenix ARB-L16 it only draws 1.4 amps. So what battery you use will change the output since it’s basically direct drive with the voltage over the vf of the led. The vapcell 16340 is the best high drain battery in it’s size that I have tested.
If using the 0805 resistor I would use a stack but at 2.2 amps it might be alright since mine factory is drawing 2 amps with the vapcell. Moving up to 1206 1 watt should work great since the light came with a 0805 with a 1/2 watt or less resistor.
Try the .100 ohm resistor first and see what it draws. The current draw isn’t the same as the current to the led even though it’s direct drive. You still have all the other components resistance in the path. Changing to better lower resistance components allows the led to see more current. You also have to take into consideration that the output drops as you use it as the battery voltage drops. The better the battery the higher sustained output you’ll have as it falls. Once it goes below the vf of the led the PAM2803 takes over and the better components will give you a slight increase in output as well from being more efficient. With mine, I don’t use it on high that much at all, but it’s there if I need it.
If you use the larger diode the side of the diode near the inductor can be filed a little to give a little more clearance.
Thanks for all this information!
I have some 16340 vapcells in the mail, but not at hand yet. I was told they are not protected. Does the driver offer LVP or do we just need to be careful?
I am going to use 1206 1W resistors, so I will go with one. That should eliminate the touching the anuminum cap issue.
I am thinking of replacing the other parts with the lower resistance ones, and then regulate by selecting the correct resistor, for my taste in Amps. Isn’t that what you are implying would be the best way?
Exactly, it makes the driver more efficient either way and should be more reliable with the higher rated components… Less resistance, less loss in the components and more current to the led.
I was wondering if you might be able to help me. I posted this (below) elsewhere and was directed to you here and it looks like you may know the answer to my problem. I am guessing that my problem may be the 4.7uH inductor. Does it sound like it is just that or might it also be the component (in your first picture) at the bottom with the glob on it because on mine that looks like it is also an inductor and it may be damaged because I can see windings exposed and there looks like the top of it may have a small piece broken off.
Sunwayman V11R repair?
I have contacted Sunwayman already about this, I filled out a contact form and I am waiting for them (in China) to reply.
I have had a V11R for a few years and really like it but just the other night it stopped working and it appears to be the V10R V1.4 control board inside the flashlight, the parts on that board that react to the variable intensity ring being moved seem to have disintegrated. There were parts floating around in that part of the flashlight. They appeared to be magnetic. I was wondering if is it possible to get a schematic and parts list so I can try and find the broken parts and replace them myself. There are 2 devices on the board that look like electro magnets and they seem to be physically broken. I assume that they interact with the magnetic ring to change the intensity of the light.
I am sure that I have dropped this light a few times but the latest drop seemed to break those little parts inside the light. I have a used one on its way to me now but I would like to fix this one since I think it is something that will break again.
I just now saw your post. Sorry I didn’t get back here sooner.
I’ll try to help if I can, I think I understand the output side of the driver but all the components for controlling the rotary position sensor I never tried to figure out because it only controls the amount of output in relation to the magnets positioning to the 316BCG, not the peak output.
The second smaller coil on the bottom in the my first pic, I think is just part of a boost circuit to supply enough voltage to the 316BCG. The 316BCG from the datasheet requires 4.5v min to operate. If it’s bad then there will be no light. I replaced one that had that same problem.
There are two magnets glued in the control ring that move with the control ring that changes in position in relation to the north and south axis of the 316BCG. As they move the 316BCG pics that movement up and relays the position to a mcu which in turn outputs so much current according to that position, which I think is programmed in the mcu firmware. The jetbeam rotary lights seem to have the same circuit but offer a strobe mode and sos with a double turn to high and a triple turn to high for sos. Pretty sure thats just being picked up in firmware just like a double tap for high in Narsil with a e-switch.
I would check the two magnets in the control ring and the small coil if you think it’s not connected. If you can reconnect it even if you can just solder the winding wire to the pads to see if it will work might help. Then you could find a replacement coil and reflow the new one on.
The magnets also have a north and south, if you get them out of sink or in the wrong way the position sensor will not work correctly. Takes a little experimenting to figure out which way they go. There tiny too, so be careful where you open the light up and try not to lose them.
Just shoot me a PM if you need any further help.
Great post! Was thinking about to get one V10R…
The V10R has been discontinued for some time now. You can find them used every now and then on ebay or here on the forum.
The V11R can still be found in a few places. Sunwayman V11R CREE XM-L2 U2 Variable Output 1 x CR123 / 1 x AA 500 Lu?
Capacitors C2012×7R0J106M125AB https://www.arrow.com/en/products/c2012×7r0j106m125ab/tdk
Looks like not avail at arrow anymore. What would be next one in their offering? I’m a huge noob in electronics…
You may order Samsung CL21B106KOQNNNE at Arrow instead.
You would need only one of these 22uf instead of using two 10uf paralleled, just swap the original out. Should work just as good, two paralleled that microa recommended will work also.
Thx guys! Cart has been updated.
Gotta question regarding V11R’s…they are XML reflector lights, so how is the beam with XPG size emitters? What’s the size of the MCPCB?
Just to know what to expect when i get “11”