Led4power.com : LD-4 CC linear drivers, ILC-0/1 illuminated tailcaps, optics, MOSX, copper DTP MCPCBs...

707 posts / 0 new
Last post
giorgoskok
giorgoskok's picture
Offline
Last seen: 1 day 18 hours ago
Joined: 11/13/2015 - 10:46
Posts: 2651
Location: Greece

Received my MosX mcpcb’s today , that was fast !

Just tried 2 of them , with 2 Oslon black flat leds (LUW HWQP) .

The reflow was really easy , pads line up exactly and you can remove excess solder easily .

I made a simple setup in a Eagle Eye X6 , with a fet driver using a PSMN3R0-30YLDX fet (not really low resistance fet) . Soldering the wires on the mcpcb is tricky , thank god i have a soldering station with variable temp .

Unfortunately the light it emmits turns blue in turbo (when in direct drive) , so i guess as djozz mentioned the thermal pad of the Oslon isn’t big enough , and the black flat can’t handle the heat on a non dtp board .

I will run some more test tomorrow and report back.

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

I haven't tested black flat yet, but conclusion is that LEDs with very small die and small heat pad area really need copper DTP. Efficiency of LEDs drops at higher current because of two things: temperature of die, and current itself(why, still not fully explained by physicists) . For most big LEDs second cause is dominant at high currents because lumen stops increasing even if die temperature isn't even close to 150C, which is maximum allowed by most manufacturers. Maybe best example is Luxeon V with industry lowest thermal resistance per mm2 of die (0.8C/W,4mm2), it tops at around 9A, but if you calculate die temperature at 9A (LED on CuDTP, room temp. 20C), it is only      20C+ 9A*3.55V*0.8C/W=45.5C , not even warm in LED world. So second cause is dominant. This means even if you put such LED on non-dtp board which would cause for ex. 10C higher die temperature, that's still only 55C, difference in light output would be hard to detect. 

On the other side it seems temperature effect on black flat is considerable, simply because power density for tiny 1mm2 die is very high. Temperature of die with copper DTP at room temperature,5A  is 20C+5A*3.9V*4.3C/W=104C , this is quite high, if you take fact that typ. flashlight would have 40-50C temp., then die temp of oslon is very close to max. 150C. That's why every additional deg C caused by non-dtp board would have visible impact. Problem with black flat is small heat pad (3.5mm2, but it's questionable how good is heat distributed over that 3.5mm2 inside package), if black flat is in XM-L sized package, difference between DTP vs mosX would be much smaller.

As I said earlier, I designed mosX Osram/luxeonIR boards for infrared LEDs, where power density (~ max 2A) is much lower than with black flat, I expect no visible difference DTP vs. mosX in that case. 

EDIT: When soldering wires to mosX boards, best method is to put some solder on iron tip and try to put some solder on one of the big pads. At first nothing will happen, but as board heats up, solder will melt over pad when board reaches high temperature. So it's important to hold tip on pad steady for some time. After board is hot, sodlering wires on all pads is easier. Board must be on piece of thick paper or some other thermal insulator.  

giorgoskok
giorgoskok's picture
Offline
Last seen: 1 day 18 hours ago
Joined: 11/13/2015 - 10:46
Posts: 2651
Location: Greece

Thanks for your detailed explanation Smile

The_Driver
The_Driver's picture
Offline
Last seen: 54 min 28 sec ago
Joined: 10/20/2016 - 05:51
Posts: 1175
Location: Germany

Black Flats often turn blue at 5A, even on DTP PCBs. Limiting the current to 4.5A fixes the problem. The brightness difference is negligible.

The quality of the soldering will also have an effect on the performance.

khas
Offline
Last seen: 6 hours 53 min ago
Joined: 08/08/2014 - 07:20
Posts: 906
Location: Denmark
Jerommel wrote:
khas wrote:
I used one of these Convoy centering rings reamed out to fit the emitter
I use those too, but i pushed a square diamond coated file through it, so you have 4 corners filed away. Perfect fit.

That’s a great idea Thumbs Up

Blazer296
Offline
Last seen: 4 months 1 week ago
Joined: 02/15/2015 - 03:17
Posts: 169
Location: central illinois

Have been looking at these drivers a bit lately. Very interesting, with a lot of features I like. Couple questions I haven’t seen yet……..

When the double tap high (100%) is turned on, will it still be a part of the normal modes?
Ex= if 4 mode is set, normal mode changes go 2-10-30, then back to 2, with max only on double tap?

Does anybody know if the quad Luveon V will fit under the Astrolux S41 optics?

Thanks,
Jim

djozz
djozz's picture
Offline
Last seen: 25 min 14 sec ago
Joined: 09/07/2012 - 17:04
Posts: 13308
Location: Amsterdam

Blazer296 wrote:

Does anybody know if the quad Luveon V will fit under the Astrolux S41 optics?

Thanks,
Jim


I think that version has not been reported yet. It does fit under the quad Carclo, but the S41 optic is a bit different.

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

Blazer296
Offline
Last seen: 4 months 1 week ago
Joined: 02/15/2015 - 03:17
Posts: 169
Location: central illinois

Thanks Djozz, I will have to check on the Carlco quad.

I have their triples, but not a quad to check if I could make it fit. If I recall correctly the S41 optic is only 22mm, where the Carlco is 24. May be possible to shave it down a bit, 2mm is not much. Pics on MTN looks like I MIGHT be able to just get away with it.
led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

Blazer296 wrote:
Have been looking at these drivers a bit lately. Very interesting, with a lot of features I like. Couple questions I haven't seen yet........ When the double tap high (100%) is turned on, will it still be a part of the normal modes? Ex= if 4 mode is set, normal mode changes go 2-10-30, then back to 2, with max only on double tap? Does anybody know if the quad Luveon V will fit under the Astrolux S41 optics? Thanks, Jim

 

Double tap is just shortcut, so if you have 4 modes and 2x tap to high enabled, modes will work 1-2-3-4 just as usual,  2x tap mode jumps to mode 4 from any other mode.

 

About S41 optics, if it's different than carclo then it would not fit. I have several more mosX boards on my wish list, I could add S41 to that list too, and make them in next batch, but I need detailed drawing of such board.

 

Unrelated to this, I have soldered Luxeon V LEDs on mosX 4040,3xV,4xV with soldered LED wires ready to use, I just need to list them on my page.

 

 

 

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

Coupon codes for larger orders:

 

L4P-3of50        3€ for every order over 50€

L4P-10of100    10€ for every order over 100€

L4P-25of200    25€ for every order over 200€

L4P-45of300    45€ for every order over 300€

Blazer296
Offline
Last seen: 4 months 1 week ago
Joined: 02/15/2015 - 03:17
Posts: 169
Location: central illinois

@ led4power
Any idea when you will have the Luxeon V on 16mm DTP mcpcb back in stock?

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

~3weeks.

Mike C
Mike C's picture
Offline
Last seen: 2 months 3 weeks ago
Joined: 01/22/2014 - 08:03
Posts: 2009
Location: Sweden

I just placed an order for some quad MCPCBs with Luxeon Vs but I didn’t see any tick box anywhere for reflow service, neither did I see any place to add message on placing order. Is reflowing the LEDs onto the MCPCB included? If not, do you offer reflow service? What would it cost? I’d like my the LEDs reflowed if possible.

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

I'll put listing with soldered LEDs as soon as I find time, until then you can order LEDs and mosX separately,and leave note about soldering LEDs, no extra fee for soldering.

Mike C
Mike C's picture
Offline
Last seen: 2 months 3 weeks ago
Joined: 01/22/2014 - 08:03
Posts: 2009
Location: Sweden

It’s the order placed to Sweden at about 23:30. I can PM you order number if needed.

A question about the MOSFET option for future orders if I may. The MOSFET that you put on if option is selected, is it specific for your drivers? I’m wondering if I could use it for mine own drivers as a full on boost mode and get similar results as with the SIR800DP.

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

OK,I see it.

MOSFET is just that, mosfet, you can use it with any driver that use mosfet, DD or linear. It has 2,4mOhm resistance, one of the lowest in LFPAK33 package,and should be good in most cases.

Even with DD drivers, having mosfet on LED board is beneficial because:

-it saves space on driver board (more space for other things),

-Even on "DD" mosfet has finite resistance, if mosfet is on driver PCB at 10+Amps it would generate enough heat to make internal uC temperature reading wrong/problematic - with mosfet on LED pcb there is nothing left on driver pcb that could generate larger amount of heat, so temperature sensing with internal uC sensor is much more predictable.

 

 

Flashy Mike
Flashy Mike's picture
Online
Last seen: 12 min 48 sec ago
Joined: 01/14/2016 - 16:38
Posts: 1043
Location: Germany

What’s the current limit of this mosfet? Is it good for 25 Amps with single cell DD?

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

25A without problem. Power dissipation in that case is 3-3.5Watts, this is no problem on any mcpcb. On the other side 3+ Watts on driver's FR4 pcb would cause melting if there is no additional thermal path (silicone cubes etc.). Internal uC temp. sensing would be pretty much useless for thermal regulation/stepdown.

Flashy Mike
Flashy Mike's picture
Online
Last seen: 12 min 48 sec ago
Joined: 01/14/2016 - 16:38
Posts: 1043
Location: Germany

Thanks for the info!

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

 To find out resistance of spring that I have for sale HERE 

I used rigol DP811 in CC mode at 1,2 and 3 A and measure voltage drop with 4-wire method.

Djozz did similar test for few springs quite long ago: http://budgetlightforum.com/node/35228

I measured resistance for 3 other springs to check consistency: convoy steel spring, fasttech bronze spring, Intl-outdoor beryllium short spring.

Results are:

Convoy steel spring:

1A 43,5mV  :R=43,5mOhm

2A 86,4mV : R=43,2mOhm

3A 133mV : R=44,3mOhm

Avg. resistance: ~43.7mOhm

 

Fasttech bronze spring:

1A 22,5mV  :R=22,5mOhm

2A 44,8mV : R=22,4mOhm

3A 67,2mV : R=22,4mOhm

Avg. resistance: ~22.4mOhm

 

Intl-outdoor beryllium short spring:

1A 7,4mV  :R=7,4mOhm

2A 14,63mV : R=7,32mOhm

3A 22,5mV : R=7,5mOhm

Avg. resistance: ~7.4mOhm

 

Phosphor bronze spring 12x9x1mm

1A 13,4mV  :R=13,4mOhm

2A 26,82mV : R=13,4mOhm

3A 40,5mV : R=13,5mOhm

Avg. resistance: ~13.4mOhm

 

It's obvious that measurements are very consistent and independent of current.

At higher currents resistance of steel spring starts to increase due to temperature rise caused by power dissipation.

Intl.outdoor spring has smallest resistance, but it's also the shortest,and not very "springy".

 

Conclusion is that 12x9x1mm bronze spring is good for currents up to 6-7Amps without need for bypass, which should be perfect for medium (by today's standards) lights, or for SRK-like mulitcell lights.

Spring is big enough to add bypass of course, good thing compared to bypassed steel springs is in case of bypass failure spring has still very decent conductivity and probably won't melt,unlike steel springs.

 

 

 

djozz
djozz's picture
Offline
Last seen: 25 min 14 sec ago
Joined: 09/07/2012 - 17:04
Posts: 13308
Location: Amsterdam

^ Cool

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

MRsDNF
MRsDNF's picture
Offline
Last seen: 3 hours 28 min ago
Joined: 12/22/2011 - 21:18
Posts: 12385
Location: A light beam away from the missus in the land of Aus.

Thanks for the testing led4power. Thumbs Up

My current and or voltage measurements are only relevent to anything that I measure.

Budget light hobby proudly sponsored by my Mastercard and unknowingly paid for by a hard working wife. 

djozz said "it came with chinese lettering that is chinese to me".

old4570 said "I'm not an expert , so don't suffer from any such technical restrictions".

Old-Lumens. Highly admired and cherished member of Budget Light Forum. 11.5.2011 - 20.12.16. RIP.

 

The_Driver
The_Driver's picture
Offline
Last seen: 54 min 28 sec ago
Joined: 10/20/2016 - 05:51
Posts: 1175
Location: Germany

Very nice, thanks for testing!

I have to say, you’re doing an amazing job offering specialty products which precisely cater to the unique needs of modders on the flashlight forums! You fit in perfectly with some of the other guys like Clemence and Lexel (and maybe the Sandwhich Shoppe 10-15 years ago). Everybody has their own niche where they offer their unique products. You seem to be especially motivated though to offer stuff that has never existed before and is very practical. And all of that for such nice prices from a guy who knows what he is talking about. Wink

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

Finally first LD-B4HV test (Luxeon M 6V version as test LED on 20mm 4040 mosX, 6A LD-B4HV):

 

  

 

6A high mode, top side of mosfet is still cold:

Two 6V Luxeon Ms in series as simulation for 12V (4S) LEDs like XHP-35:

  

6A high mode, 16.4V input, around 100Watts input/output power:

I will try XHP50.2 in Convoy C8 host with LD-B4HV and 2x18350 Aspire cells soon. 

Cula
Cula's picture
Offline
Last seen: 6 hours 52 min ago
Joined: 11/14/2014 - 02:04
Posts: 357
Location: Croatia
led4power wrote:

Finally first LD-B4HV test (Luxeon M 6V version as test LED on 20mm 4040 mosX, 6A LD-B4HV):

Have I missed something or is this your new 2S/3S driver? Side switch capable?

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

Yes, this is 1S-4S variant of LD-B4 (1S only). Side switch will be available later, it's  supports clicky switch for now.

thijsco19
Offline
Last seen: 4 hours 40 min ago
Joined: 12/14/2012 - 16:18
Posts: 1247
Location: Nearby Rotterdam, the Netherlands

It is capable of driving a 3v led with 1s, 2s, 3s and even a 4s battery setup? Without needing to change components on the driver?

Or is 1s still limited to 3v LEDs and 2s to 6v leds and 4s to 12v LEDs?

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

It can drive 3V LED with any voltage up to 20V, but this is linear driver,it burns extra voltage into heat. This driver is good when voltage of LED and battery are similar, like 2S cells and XHP50.2, XHP70.2.

In combination with mosX MCPCB generated heat in mosfet is not problem at all,since it's well cooled on mosX. Second generation of XHP50 and XHP70 have very low forward voltage, so they draw too much current with DD drivers which can lead to LED death or at least low efficiency because of too high current. Efficiency is equally bad in all modes that are derived by PWM of DD mode. Also 7135 regulators have max. operating voltage lower than 8.4V, so reliability is questionable when they are used in 2S setups.

This drivers provides constant current on all modes, so there is no problem with too much current on high mode, LED efficiency is max. possible in all modes. Disadvantage of CC method was extra heat generated in driver(mosfet to be precise) - that is completely solved with mosX board because mosfet is on same MCPCB as LED.

 

djozz
djozz's picture
Offline
Last seen: 25 min 14 sec ago
Joined: 09/07/2012 - 17:04
Posts: 13308
Location: Amsterdam

Are you planning someday to develop a buck driver, to run leds from a battery set-up with much overvoltage, like 2S for a 3V led?

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

led4power
led4power's picture
Offline
Last seen: 4 hours 34 min ago
Joined: 12/29/2012 - 09:48
Posts: 999
Location: Croatia,EU

Yes, but I have to have linear drivers too, they will always be one of the options because of small size and relatively low cost/Watt.

Pages