Help with Current Sense Resistor Modding

103 posts / 0 new
Last post

Pages

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA
Help with Current Sense Resistor Modding

I bought this bike light, which runs on 4.2v (and does not run on 8.4v).  The driver looks like so:

1

 

So the R100 & R120 are the current sense resistors, right?  In stock form I am getting 1.8A pulled from the battery (voltage @ 4.06v) - so about 0.9A to each LED.  What value resistor(s) would I add (or replace with) to bump current up to about 3.0A @ the battery for approx. 1.5A per LED? 

Thanks,
Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 1 month 1 week ago
Joined: 01/04/2014 - 06:47
Posts: 5071
Location: H-Town

You could stack the same value resistors over the top of each of them putting them in parallel (effectively 1/2ing the resistance/doubing the current) or replace those resistors with 1/2 the resistance (R060 and R050 resistors)

Either way you will need to replace them, just bridging them will make it more or less direct drive

With no buck capabilities the driver can run the led’s w/ 4.2vdc

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

Ok, I was assuming stacking the same values would give me double the current.  That's at least a start.  I don't think I have R100's, but I know I have R120's. Using 2 R120's would get me a little less than double, correct?

I definitely do NOT want direct drive in this light.

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 1 month 1 week ago
Joined: 01/04/2014 - 06:47
Posts: 5071
Location: H-Town

Yes, w/ a 120 over a 100 it will be a little less than half of a 120 but a little more than if you had two 100’s stacked, but still good to go

approx 54.5454 mOhm

ryansoh3
ryansoh3's picture
Offline
Last seen: 3 months 4 days ago
Joined: 08/04/2012 - 08:21
Posts: 3880
Location: US

WarHawk-AVG wrote:
Yes, w/ a 120 over a 100 it will be a little less than half of a 120 but a little more than if you had two 100’s stacked, but still good to go

approx 54.5454 mOhm

Yup, what he said. Smile

This one should be fairly straightforward.

BLF ≠ B-grade Flashlight Forum

 

bibihang
bibihang's picture
Offline
Last seen: 3 weeks 1 hour ago
Joined: 11/10/2011 - 09:32
Posts: 2387
Location: Malaysia

Remove both of the R100 and R120 and use three R050, you will get approximately 2.94A @ the battery.

Or, stack two R050 on top of either R100 or R120 and you will get approximately 2.86A @ the battery.

Anyone is welcome to check my math though.

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 9 months 5 days ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

Are those sense resistors or limit resistors? Is that driver even regulated? What is the current at 4.15V, 3.8 V? It just looks more like a DD driver with modes to me.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

Well at 3.89v I get 1.53A. Does that help?

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
fellfromtree
fellfromtree's picture
Offline
Last seen: 5 years 1 month ago
Joined: 07/25/2014 - 15:14
Posts: 470
Location: spelunking

It looks like to me a dd driver with those 2 resistors going from ground to the source on the fet. Limiting like that.

hmm, I’d probably take the .100 off and put in a .120- then lay a couple of .120s on top of that, if that’s all you have. Should give you ~3.2-3.3a or so.

Aiming at .032 for 3a from my math anyways, at 4.06v.

Try the four .120’s and see where it’s at.. If you have a .180 that, stick that in over a .120 and it should be right at 3 according to my calculator least.

I’m sure wight will be in here

bibihang
bibihang's picture
Offline
Last seen: 3 weeks 1 hour ago
Joined: 11/10/2011 - 09:32
Posts: 2387
Location: Malaysia

Rufusbduck wrote:
Are those sense resistors or limit resistors? Is that driver even regulated? What is the current at 4.15V, 3.8 V? It just looks more like a DD driver with modes to me.

Good point. There is no torroid there, I guess it is just DD driver. If it is really DD driver you can just simply bridge the resistors.
ImA4Wheelr
Offline
Last seen: 3 weeks 4 days ago
Joined: 02/03/2013 - 14:51
Posts: 7918
Location: SC

Like RBD and FFT said, they appear to be current limiting.  If they are, their resistance in and of itself is limiting current flow.  Are there any components on the other side of the driver (Other than the switch)?  Also, I can't follow some of the traces due to part of the board being in shadows.

bibihang
bibihang's picture
Offline
Last seen: 3 weeks 1 hour ago
Joined: 11/10/2011 - 09:32
Posts: 2387
Location: Malaysia

garrybunk wrote:

Well at 3.89v I get 1.53A. Does that help?

-Garry


4.06V – 1.8A
3.89V – 1.53A

Pretty sure it is DD driver then.

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 9 months 5 days ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

The use of the resistors would still be the same but the values might be different since those resistors just add to the total of the circuit. In other words halving the resistance probably won’t double the current. Also, if it’s DD then it’s not regulated across the battery charge so output will start to dim from the moment you turn it on. Don’t know if you care about that.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 9 months 5 days ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

I’m no EE and could well be wrong but if not you might be better off just replacing it with a 3A Qlite or whatever firmware 105C floats your boat. Then you would have better efficiency and consistent output over most of the charge cycle. Better still would be a Led4power LD-01 which is even more efficient in the low modes than a nanjg type and doesn’t have pwm. It’s not as cheap and has fewer firmware options but I can’t wait to try one out for myself. If the board has an integral switch then another option would be to pull the resistors and the FET and replace them with an 8×7135 slave board and keep the existing mcu and firmware.

As it is that driver seems to be running with one foot on the gas and the other on the brakes.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

Rufusbduck wrote:
If the board has an integral switch then another option would be to pull the resistors and the FET and replace them with an 8x7135 slave board and keep the existing mcu and firmware. p. As it is that driver seems to be running with one foot on the gas and the other on the brakes.

Hmm . . . interesting idea, but I'd need my hand held the entire time as that is well beyond my ability to handle.  Can you sketch up a diagram on how I'd wire it up? 

Thanks,
Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

Here is another photo of the driver showing the traces a little better:

2

(I didn't post this one because the FET wasn't as legible.)

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
antiparanoico
Offline
Last seen: 3 weeks 3 hours ago
Joined: 07/14/2013 - 19:17
Posts: 1164
Location: Wandering the world.

I can’t see any photo.

If anything can go wrong, it will.
Matter will be damaged in direct proportion to its value.

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

antiparanoico wrote:
I can't see any photo.

Not sure why not.  Can you access my Photo Gallery here?

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
fellfromtree
fellfromtree's picture
Offline
Last seen: 5 years 1 month ago
Joined: 07/25/2014 - 15:14
Posts: 470
Location: spelunking

I would just bridge it if it was me garry. I mean think your pushing 4s in parallel into two xml2 u2 emitters in parallel from the looks of it. You can run in medium, low modes and have the high for when you need it. Maybe reflow onto a couple of noctigons and redo the tim. You should have plenty of light in the lower modes.

Or put in the 120s you already have and see where high takes you. That’s the cheapest way right now. Those u2 aren’t going to blow up on dd- just keep it in the lower modes until you can get the noctigons and tim mailed out to you

Just my 2 cents anyways.. Silly * disclaimer I take no responsibility if those emitters are mounted badly or have no pill and go blue on high

But seriously the four 120s should be good even with crappy sinks, that’s less than half what those leds can take.. You can then add more resistance or take away after playing around.

antiparanoico
Offline
Last seen: 3 weeks 3 hours ago
Joined: 07/14/2013 - 19:17
Posts: 1164
Location: Wandering the world.

garrybunk wrote:

antiparanoico wrote:
I can’t see any photo.

Not sure why not.  Can you access my Photo Gallery here?

-Garry

Yes. And… now I can also see the photos… Dwarf mischief Big Smile

If anything can go wrong, it will.
Matter will be damaged in direct proportion to its value.

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

fellfromtree wrote:
* disclaimer I take no responsibility if those emitters are mounted badly or have no pill and go blue on high. . .

2

 

1

 

That's all I have for heatsinking in stock form!  Emitters on a 2mm thick aluminum board only making contact at those triangle shaped contact points (not even around the edge of the board!).  I plan to fully "pot" the area behind the board out to the driver.  Even then I don't want to produce a huge amount of heat.

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
ImA4Wheelr
Offline
Last seen: 3 weeks 4 days ago
Joined: 02/03/2013 - 14:51
Posts: 7918
Location: SC

^ At least it appears to have a copper layer to spread heat and transmit to the host.  Might not be all that bad.  Need to check for thermal sag to know.

I looked at the photo gallery.  It looks like those resisters have traces going to more than just the FET Source Pin.  Are there any components on the switch side of the board?

fellfromtree
fellfromtree's picture
Offline
Last seen: 5 years 1 month ago
Joined: 07/25/2014 - 15:14
Posts: 470
Location: spelunking

Yeah I would bridge it man.. You don’t have to worry about the limiting resistors going out. I think, well from what I’ve gathered here on the forums.. That’s not the correct way to limit power. I’m no expert by any means but I know what those emitters can do. Try bridging it and put it on medium- check the heat. The pwm should keep it cool and you’ll have high for blasting your buddy in the face when he looks back

*Edit yeah 4wheel guess could cut the trace to the source and ground the source on the FET.. That would be the safe way to do it

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

ImA4Wheelr wrote:

It looks like those resisters have traces going to more than just the FET Source Pin.  Are there any components on the switch side of the board?

Ugh!  Do I really have to tear it down again?  I figured the backside didn't look like it had anything important.  Should have known I should have taken pics.  Will try to get to it soon (like tonight).

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

fellfromtree wrote:
Yeah I would bridge it man.. You don't have to worry about the limiting resistors going out. I think . . . but I know what those emitters can do. Try bridging it . . .

I like the way you think!  All or nothing!

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

BTW - am I okay gauging LED current by the input current on this driver? Or do I really need to check current at the LEDs?  I guess I could pull that emitter board back out and screw it down on a big CPU heatsink while trying to tweak the driver.  That way if I hit 4A/emitter it doesn't go "POOF".

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
fellfromtree
fellfromtree's picture
Offline
Last seen: 5 years 1 month ago
Joined: 07/25/2014 - 15:14
Posts: 470
Location: spelunking

You should be ok in dd.. I was thinking like this but someone might have a more elegant solution that involves giving you the 1.5a per emitter your looking for

Cut the trace at the red line, connect a wire from the fet to gnd, (blue line). That’ll keep everything like it was and put it in dd

*edit actually just unsolder the leg on the fet- probably a better idea before connecting the wire

ImA4Wheelr
Offline
Last seen: 3 weeks 4 days ago
Joined: 02/03/2013 - 14:51
Posts: 7918
Location: SC

GB wrote:

Ugh!  Do I really have to tear it down again?  I figured the backside didn't look like it had anything important.  Should have known I should have taken pics.  Will try to get to it soon (like tonight).

You don't have to for me bro.  I am curious how you planned to resistor mod the light without opening it back up. Wink

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 9 months 5 days ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

Unless you are a tinkerer and willing to risk failure it’s probably best to start by just stacking or bridging the resistors to see what that gives you. However, if like me, you can’t leave well enough alone I’m willing to lead you astray.

This is essentially an adaptation of the master/slave mod developed by Techjunkie and posted here by Oldlumens.

First a bit of identification.

On that FET the big tab at the top is called the drain and it’s connected to Led-.

The smaller pin bottom left is called the gate and is how the mcu turns the FET rapidly on and off(what’s called pwm). The amount of (on) time is called the duty cycle and gives you your modes.

The other small pin bottom right is called the source and is connected through those resistors to V-.

7135’s have the same 3 connections but they have different names. Don’t ask me why, maybe an EE could tell you.

The large tab is instead called ground and is equivalent to the source pin.

The bottom right pin is called Vdd and has the same function as the FET gate pin.

The bottom left pin is called Led out or Led -.

The middle pin is the same as ground and is often left unconnected without consequences.

All the 7135’s on a typical driver are connected in parallel so if you make those 3 connections between this board and any of the 7135’s on the slave board then all the chips are used.

What you would need to do is remove the FET and the two resistors and solder in 3 wires to connect the slave board there instead.

If you’re not comfortable soldering and desoldering smallish stuff then you shouldn’t try it but here’s why you should consider it. The difference between the two is in regulation and resistance. The FET is either fully on or fully off and will always pass as much current as the voltage allows during each (on) cycle with the maximum controlled by those resistors. As the battery voltage falls since the resistors are fixed value the current falls as well and the LEDs get dimmer in every mode. With the 7135’s, when they are on each one passes a fixed amount of current with the maximum controlled by the number of 7135’s. As the battery voltage drops the resistance of the 7135’s drops as well to maintain current level. The lights won’t begin to dim until the battery voltage drops down close to the forward voltage of the led and stays constant until then in all modes.

The FET/resistor combo would be more similar to the 7135’s if instead of fixed value resistors a pot were used instead that automatically lowered it’s resistance as the batteries drained.

If you’re interested I can post a sketch later showing the connections but you would need a slave board or a stack of 7135’s to do this.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

Thanks for such a clear detailed description RBD!  I understand how to wire a slave board up without a sketch.  Now the question is, do I want to try adding resistors, or try the slave board method?  Hmm . . . I think I'll try the slave board method, but I need to make an order.  RMM stock those?  I'll go have a look.

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
garrybunk
garrybunk's picture
Offline
Last seen: 23 hours 8 min ago
Joined: 10/31/2011 - 09:25
Posts: 6050
Location: Johnstown, PA

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.

Pages