Mountain Bike Reviews Forum banner

Niterider Minewt 250 cordless - deconstructed

11K views 31 replies 14 participants last post by  rfxc 
#1 ·
I haven't seen much about this light here, and was ordering something else from Perfomance, and had a coupon....So I ordered one to play with. My plan was to try to 'turbocharge' it by upgrading the LED and maybe the battery. I figured others here might be interested as it looks like a decent all-in-one light. It was of course on backorder, but finally arrived yesterday.

So I dug in today to check out the guts and see what could be upgraded. Below are the first pictures.

The LED looks like a Seoul P4 to me, on a round 20mm PCB. There are some spots for resistors and stuff on the board, but they don't seem to be populated. Maybe they are used on the 150 lumen version?

The reflector is also 20mm diameter and about 16mm long.

The upper PCB has a a small pushbutton switch and small red and green LEDs on either side for power/battery info.

The lower PCB has all of the driver electronics on it, with 2 wires going to the LED PCB. The charge connector is on the back of this board, and is one of those mini USB connectors that most digicams use these days.

The battery is a single lithium 18650 cell, labeled 2800Mah. So it looks like the battery is not worth upgrading until higher capacity units are available. What are the current best ones, 3000Mah? The battery is not soldered and can easily be replaced once the unit is apart. Not something you would want to do while riding, as it requires removing 3 M2.5 screws.

The battery is still charging, so no beamshots or voltage specs yet, but they will come.

If anyone wants more info or measurements, just let me know!

BTW, how much current would it take to get 250 lumens from a P4?

I'm thinking an XM-L might get me more light and even longer run times. Run time is quoted at 2:30 on high.

Mark
 

Attachments

See less See more
4
#2 ·
Nice info and pics bikeny,
I still have a few SSC P4 lights from my first builds (years ago) and they still work just fine. The Max input current they can handle is 1 Amp and that should give you 240ish lumens, don't give them any more or they will go blue.

As far as I'm aware 2900mah is the highest capacity 18650 cell ever and they are rare (the real ones that is).
 
#3 ·
Nice dissection :thumbsup:

I'm working on a Pace 310 tonight i hope - just gotta get a beamshot for prior upgrade. I have a Cygolite Milion 200 I was going to check out with the same thought you had. Issues with that one is it has a modular Li-Ion battery that slides in from the rear for quick change.

I'll be watching to see how your's comes out.

Ed
 
#4 ·
NitroRC Ed said:
Nice dissection :thumbsup:

I'm working on a Pace 310 tonight i hope - just gotta get a beamshot for prior upgrade. I have a Cygolite Milion 200 I was going to check out with the same thought you had. Issues with that one is it has a modular Li-Ion battery that slides in from the rear for quick change.

I'll be watching to see how your's comes out.

Ed
I just replaced the led in my Cygolite expilion 250. I did not recognize the original led and there were no markings, but the dome of the xpg r5 I replaced it with is larger. It was pretty straight forward, but Cygolite just smears thermal paste all over the place in there. Overall, the Niterider looks to be a better build.
 
#5 ·
anthonylokrn said:
I just replaced the led in my Cygolite expilion 250. I did not recognize the original led and there were no markings, but the dome of the xpg r5 I replaced it with is larger. It was pretty straight forward, but Cygolite just smears thermal paste all over the place in there. Overall, the Niterider looks to be a better build.
Overall, it is a decent build, and so far no thermal paste at all. I hope there's some under the LED PCB! The battery finally charged all the way, took about 4 hours. I did some quick voltage measurements on the three levels. I don't have any replacement LEDs at the moments, so no un-soldering yet to get the current. Voltages as follows:

High: 3.5V
Medium: 3.05V
Low: 2.4V

So now I am confused. According to the Seoul P4 spec sheet, 3.5 volts equals about 450mA. But 450mA only equals about 135 lumens. What am I missing? Maybe it's not a P4? I guess I need to measure the current sooner than later;) .

Mark
 
#6 ·
upgrading that's going to be awkward as the back of the P4 will be thermally coupled to the housing with some paste (was on the minewt150 I took apart a while ago) and all the other bare LEDs (XP-G, XM-L) have the + and - solder pads on the back.

Your best bet may be a XP-G (or even better, XM-L if you can find one) on a 10mm star, though you'd have to check total height so that the optic still fits with the cover on. That way you can thermally couple the LED star to the housing and still have a smidge of room to solder leads from the star to the pads on that board thing.
 
#7 ·
mattthemuppet said:
upgrading that's going to be awkward as the back of the P4 will be thermally coupled to the housing with some paste (was on the minewt150 I took apart a while ago) and all the other bare LEDs (XP-G, XM-L) have the + and - solder pads on the back.

Your best bet may be a XP-G (or even better, XM-L if you can find one) on a 10mm star, though you'd have to check total height so that the optic still fits with the cover on. That way you can thermally couple the LED star to the housing and still have a smidge of room to solder leads from the star to the pads on that board thing.
Yea, the plan is to get am XM-L on a 20mm star and replace the whole PCB. I will have to work out the location of the wires, the height to work with the current optics, and the mounting. I think it can be done without too much trouble.

Mark
 
#9 ·
mattthemuppet said:
ah, I see now - there aren't any components soldered onto that board, so replacing it won't be an issue. Missed that 1st time round :)

I'm puzzled by your readings too, but a direct current reading should clear things up
Yeah, maybe the board is populated on the 150 version? Anyway, I got a chance to measure the current this afternoon:

High: 740 mA
Med.: 550 mA
Low: 310 mA

So here are my calculations: Current LED (assuming P4 U bin): 190 lumens @ 4.0 volts

With a replacement XM-L: 320 lumens @ 2.9 volts.

Those are guestimations from the Cree and Seoul data sheets, so please tell me if I did something wrong!

So if I can get the XM-L to fit, it should give a big boost in output as well as run time.

Of course this is all assuming the current LED is a P4, which I cannot substantiate. Do any other LEDs look like this?

I'm also confused on where Niterider is getting the claimed 250 lumen output, any comments?

I just ordered some XM-L 20mm stars from Cutter for this light and to upgrade my homemade triple XR-E. Should be a nice upgrade to that light as well if I can find suitable optics.

Wish me luck!
Mark
 
#10 ·
the board is indeed populated with some resistors and stuff on the 150 version and that LED is almost definitely a P4 - I've had a couple in my hands that look identical and NR use P4s a lot, plus it doesn't look like the LEDs they used in their older lights.

Your calculations sound about right. As a very rough guestimate (driver losses not included, battery capacity is what it says it is etc) your run time should go from ~3 1/4h to over 4h, which is pretty tidy.

As for NR's output claims, I'd take any manufacturers claims (except Lupine, they seem honest) with a large pinch of salt :)
 
#12 ·
Interesting thread - I have an Exposure Joystick Maxx that has a tired battery, and I'm keen to replace that and also upgrade the LED at the same time.

I was going to replace with an XPG, so interested to see you are going to use an XML. What are you going to use as a driver?

I've not seen anyone take apart a Joystick yet, has anyone else?
 
#13 ·
biggusjimmus said:
Interesting thread - I have an Exposure Joystick Maxx that has a tired battery, and I'm keen to replace that and also upgrade the LED at the same time.

I was going to replace with an XPG, so interested to see you are going to use an XML. What are you going to use as a driver?

I've not seen anyone take apart a Joystick yet, has anyone else?
My plan is to leave the driver alone, and just change the LED to an XM-L. That should give a nice boost in lumens as well as run times. Win-win!

Mark
 
#14 ·
mattthemuppet said:
the board is indeed populated with some resistors and stuff on the 150 version and that LED is almost definitely a P4 - I've had a couple in my hands that look identical and NR use P4s a lot, plus it doesn't look like the LEDs they used in their older lights.

Your calculations sound about right. As a very rough guestimate (driver losses not included, battery capacity is what it says it is etc) your run time should go from ~3 1/4h to over 4h, which is pretty tidy.

As for NR's output claims, I'd take any manufacturers claims (except Lupine, they seem honest) with a large pinch of salt :)
The only thing that has me questioning my results are the claimed lumens are run times, which are 250 and 2:30. The lumens seem high and the run times low.

BTW, what's the best way to measure current? What I did was unsolder one of the wires going to the LED, and clip the meter between that wire and the PCB. I used one of the wires going from the driver PCB to the LED PCB. Sound right?

Mark
 
#15 · (Edited)
bikeny said:
The only thing that has me questioning my results are the claimed lumens are run times, which are 250 and 2:30. The lumens seem high and the run times low.

BTW, what's the best way to measure current? What I did was unsolder one of the wires going to the LED, and clip the meter between that wire and the PCB. I used one of the wires going from the driver PCB to the LED PCB. Sound right?

Mark
I suspect your current measurements are off because the output of the driver isn't purely DC current. I don't see any inductors on the board which means it's not a switching regulator. I bet they're playing some games with PWM to control brightness levels even on the high setting. The best way to see what's going on is to put a 0.1 ohm current sense resistor inline with the LED and look at the voltage across that resistor with a scope. A meter that assumes the current is DC, is going to give an inaccurate reading.
 
#16 ·
bikeny said:
My plan is to leave the driver alone, and just change the LED to an XM-L. That should give a nice boost in lumens as well as run times. Win-win!

Mark
Since the board holding the LED looks like a normal PCB, as opposed to an MCPCB, I suspect this light is designed like the other Minewt lights. The case has a metal post that sticks through a hole in the PCB so the LED slug is in direct contact with the metal case. You can't easily do that with either the XPG or XML since the thermal slug on those LEDs is designed to be reflowed soldered directly to an MCPCB. If this really is how it's setup, your best bet will be to machine the case to create a flat surface for mounting the MCPCB with the new LED.
 
#17 ·
bikeny said:
My plan is to leave the driver alone, and just change the LED to an XM-L. That should give a nice boost in lumens as well as run times. Win-win!

Mark
Maybe or maybe not. Depends on the design of the driver. I swapped out a Luxeon 1 for an XPG in my backpacking headlamp. I saw a big increase in light output but reduced runtime. The lower Vf of the XPG it turned out was allowing an additional ~100mA current from the "driver". I suspect it is similar in operation to what MtbMacgyver described above. Ultimately it really does give longer runtime though as I used to run it on high most of the time. Rarely do I use high now. The XPG on low setting is about the same as the Lux 1 on high.

P.S. @MtbMacgyver, Shouldn't the last few words of your post #15 say "going to give an inaccurate reading."
 
#19 ·
MtbMacgyver said:
Since the board holding the LED looks like a normal PCB, as opposed to an MCPCB, I suspect this light is designed like the other Minewt lights. The case has a metal post that sticks through a hole in the PCB so the LED slug is in direct contact with the metal case. You can't easily do that with either the XPG or XML since the thermal slug on those LEDs is designed to be reflowed soldered directly to an MCPCB. If this really is how it's setup, your best bet will be to machine the case to create a flat surface for mounting the MCPCB with the new LED.
Good info there! I did not remove the LED PCB yet, so I cannot confirm either way, but you are right that it looks like a normal PCB, as opposed to an MCPCB. If needed I can machine the case. Or, as I suspect, the LED will have to be raised up to align correctly with the reflector, maybe I can just machine a spacer with a cutout for the post. Once I receive the XM-Ls I will did a little deeper.

I am also starting to wonder what will happen to the driver when switching to an LED with a much lower voltage.

After looking at the data sheet more and playing with LED Pro runtime calculator, it seems to me the real specs are 1000mA at 3.8V, which gives around 2.5 hours (90% driver efficiency) runtime.

Mark
 
#20 ·
So it seems MtbMacgyver is correct. I opened the light up again and lifted the LED PCB. There is a hole in the middle of the PCB where the LED is, and it sits on a raised post from the metal housing. And I finally found some thermal paste under there!

Anyway, once the XM-Ls arrive I willplay with the LED and reflector combo to see if I can get a nice beam and then work out the heights.

On a side note, I'm not sure why Niterider continues to use the Seoul P4 LED when there are many newer brighter AND more efficient LEDs available. The P4 is like, what, 3 generations old? And this Minewt cordless is a brand new product!

That's all for now,
Mark
 
#22 ·
ballsofcottonwool said:
I suspect they are sticking witth the p7 because the tooling for casting the casings cost a fortune and would need modified to make an easy mount for an XM-L or XP-G.
The only problem with that thinking is that this light uses a brand new housing not seen on any other Niterider lights. So the question is why would they design a new light around a completely outdated LED?
 
#23 ·
perhaps they bought huge stocks of them? Could also be that they've done all the work getting the driver and the optic right for that LED and changing them would cost more than it's worth. They do have a very specific "format" (LED on post through hole in a PCB) that few other light manufacturers use, probably for this very reason.

R&D on drivers and optics is also why big light manus lag so far behind LED technology - they have to get it right 1st time, whereas we can pull something apart and swap out an optic without too much hassle.
 
#25 ·
bikeny said:
The only problem with that thinking is that this light uses a brand new housing not seen on any other Niterider lights. So the question is why would they design a new light around a completely outdated LED?
Maybe because XML was not available in 2010 and Cree could not make reliable supply of XPGs to Niterider back then?:thumbsup:
 
#26 ·
bikeny,
It is P4 LED, just a special bin designated for this product. And yes, it is 250 lumens measured at the integrated sphere. Replacing this LED with XML is a great idea and you will see over 400lm. if you will use T50 bin.
DO NOT REPLACE THE BATTERY. This is not a cell, this is battery pack with build in safety circuitry calibrated for this unit. Check with Niterider sales, they may have 2900 packs available.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top