Mag625R-4d (full write up NEW PICS)

[Flakey]

FINALLY! we have the light operational, VERY operational i might add. Mag625R-4d its a long name, but there is alot here, It is a 4d mag body, with 4 D size lithium ion cells, The light is digitally regulated and protected by an AndrewWynnRouse hotdriver. The Regulator is really nessicary because with this light running on unprotected (PTC only) D cell lithium ion batteries, the threat of overdischarge, or uneven discharge, resulting in a VWF (venting with flames) situation ... is not something that i like to play around with. The advantage of the hotdriver, is that this bulb is being pushed to the ABSOLUTE limit, and held there. the regulation is set to 12.79V at the pins, earlier today i ramped it up to 12.98V and blew a bulb. This thing is riding on the very edge, but it can take it, partly because the Voltage is limited, and partially because of the 500 milisecond softstart that AWR's regulator provides. We are looking at sub 10 hour bulb life here, but when this light is used in 10 second bursts, 10 hours is a LONG time. So lets talk about light, how does an honest 3000 TORCH lumens sound? (4050 bulb lumens) ill tell you how it looks, AMAZING, how it feels? FREAKING HOT! absolutely dazzling, this much light is ... simply stunning. it puts out FIVE times as much light as your ROP or surefire M6, and almost 4 times as much light as a typical mag85. AND IT DOESNT DIM! Total runtime is around 20 minutes, (an unheard of number with this kind of power) However doing a full run test would just not be possible, maybe if i put it in a cooler full of ice. did i mention that this thing gets HOT! in reality though the heat is not an issue, because i use this light in 5-20 second intervals. when i do night hiking i carry 2 lights, one for seeing what im about to step on and one for seeing a way down the path. With this kind of usage, 20 minutes goes a LONG way.

Ill put the picures where my mouth is for a moment.

The light

The light and its competition, an ROP LE 550 Torch lumens

Shot 1 ROP (lighting up a tree)

Mag625R-4d (same tree now overexposed)

How about the front of my house? ROP

now with the Mag625R

ITS THAT BRIGHT! if i can light it up with the ROP the 625 will overexpose it! it is tremendous!

Ceiling bounce ROP

Ceiling bounce Mag625R-4D

My computers 100 class ambiant lighting (2 flourescent lights and window spill)

Same class with mag625R doing a little ceiling bouncing =)

hehehehehe this thing is so much fun! and it looks just like a normal maglight too!!!

Funny story, i was out in the park last night, finishing a night hike, and i saw a guy walking his dog, we exchanged the basic pleasentries, Me: "How you doing this evening?"
walker: "Pretty good how about you?"
"Doing alright, I see you have the 3D cell maglight there"
"Yeah its a nice light" ::he proceeds to shine his light at a tree::
"Cool, i got the 4D maglight" :: i light up the tree and the entire park::
He stands there with his jaw drooping ....
Me: "Have a good night" :: walk away ::




This amount of light is superb, I have to give a huge amount of thanks to Andrew Wynn Rouse because without his hotdriver, this just wouldnt be possible or safe. 3000Lumes REGULATED! FLAT LINE OUTPUT OF 3000L .... just impeckable! the hotdriver sales thread is up HERE for anyone who is thinking about regulating their magmods. ill keep looking for opportunites that will be able to showcase this lights true power. i will post here if i find any goodies =)

Questions/ Comments/ Drooling are all welcome! Tell me what you think.

 

[Josey]

Re: Mag625R-4d (full write up PICS)

WOW!!! That is impressive.

 

[rscanady]

Re: Mag625R-4d (full write up PICS)

Holy $h!T that is freegin BRIGHT


Ryan

 

[Delvance]

Re: Mag625R-4d (full write up PICS)

Goodwork Flakey, i see you've ditched the 64623 bulb . Absolutely nuts (in a good way!) with the 4 D size Lions in that light...I've got a similar light to your setup and yes, it does make the ROP look very inadequate heh.
One thing actually (just noticed). What bulb are you using ? You've named it "Mag623R-4d" but the 623 bulb can take upto 16 volts ? And you're saying you blew one at 12.98V ? My 623 is getting over 14V and still re-rates to 300+ hours ? Did you mean the 64625 bulb ?

Actually, i'm thinking about getting a 1/2 or 1D extender for my light, and putting one more GP 2000 cell in there, which will bump the output up by about 20% :devil: (maybe then i will try one of the regulators too!)

 

[leukos]

Re: Mag625R-4d (full write up PICS)

Flakey,

I guess I've been out of the loop, where can one get D sized Li-ions?

 

[Flakey]

Re: Mag625R-4d (full write up PICS)

Goodwork Flakey, i see you've ditched the 64623 bulb . Absolutely nuts (in a good way!) with the 4 D size Lions in that light...I've got a similar light to your setup and yes, it does make the ROP look very inadequate heh.
One thing actually (just noticed). What bulb are you using ? You've named it "Mag623R-4d" but the 623 bulb can take upto 16 volts ? And you're saying you blew one at 12.98V ? My 623 is getting over 14V and still re-rates to 300+ hours ? Did you mean the 64625 bulb ?

Actually, i'm thinking about getting a 1/2 or 1D extender for my light, and putting one more GP 2000 cell in there, which will bump the output up by about 20% :devil: (maybe then i will try one of the regulators too!)

huh? the light is named Mag625R-4D not Mag623R-4D. The 625 bulb is very close to its melting point at 12V so overdriving it is a tedious idea. I would be carefull with the regulator with too much overhead Voltage. The fet is very close to blowing on the start up of this light. (in fact a tailcap resistor is used to keep that extra V off the fet) the 623 bulb became too much of a pain in the back for AWR. so we just settled on the 625 wich is INSANELY bright and white.

Leukos, d sized lithium ions were sold in a group buy by cpf'er emilion. Im pretty sure that the group buy is over at this point. Keep in mind that these cells are UNprotected and if not properly treated can become quite volitile. This should not be your first Lithium ion project.

 

[wtraymond]

Re: Mag625R-4d (full write up PICS)

What Delvance was saying is that in the first line of your first post you accidentally have "Mag623R-4D." The title is correct, but you have a typo in the post.

Awesome power, Flakey! Good job.

 

[Delvance]

Re: Mag625R-4d (full write up PICS)

What Wtraymond said (thankyou).

I've been using the 623 for awhile now...and i absolutely love it. I will try the 625 sometime though, must see how white thing thing is.

 

[tattoou2]

Re: Mag625R-4d (full write up PICS)

Holy crap! That's incredible.

 

[DUQ]

Re: Mag625R-4d (full write up PICS)

Flakey,

I guess I've been out of the loop, where can one get D sized Li-ions?

I think the cells came from Emilion.

Man nice work flakey. Thats a killer light.

 

[jwl]

Re: Mag625R-4d (full write up PICS)

AWESOME!!!:rock: I thought my 6D ROP was pretty good, now I think it's time to move up to more power (I have been trying to resist).

Flakey - if it wouldn't be too much trouble, would you mind listing the 'recipe' for building this light? Would it have to be run on Lion batts or could quality Nimh batts do the job?

Thanks and GREAT JOB!!!

 

[AlexGT]

Re: Mag625R-4d (full write up PICS)

Dang! thats bright!!!! Nice work!

 

[zehnmm]

Re: Mag625R-4d (full write up PICS)

Thanks for sharing the great post Flakey! And, congratulations!

I am awaiting one of AWR's hot driver kits for a 64610 50W setup. Now I feel like an 8 yr. old waiting for Christmas.

Regards.

 

[Icebreak]

Re: Mag625R-4d (full write up PICS)

Excellent report.

I've got several questions but I'll try to narrow them down to a few.

I'm particularly interested in the tailcap resistance. How was it calculated and how was it cobbled?

I'm thinking the "first to possibly fail component" of the circuit is the fet. Is this correct? Does the tailcap resistor not only give the fet relief but also capture "first to possibly fail component" designation?

To your eyes, how does this light compare in color temp to Mag35-8 ?

Sidebar comment on ckt design and dial in adjustment. The approach of finding a lamp and battery set-up that are close to being in agreement, then introducing regulation makes sense.

Thanks one more time for building this. We don't really know what can be done unless adventurous fellows do it.

- Jeff

 

[wtraymond]

Re: Mag625R-4d (full write up PICS)

I've got several questions but I'll try to narrow them down to a few.

I'm particularly interested in the tailcap resistance. How was it calculated and how was it cobbled?

I'm thinking the "first to possibly fail component" of the circuit is the fet. Is this correct? Does the tailcap resistor not only give the fet relief but also capture "first to possibly fail component" designation?- Jeff

The hotdriver works by bleeding off the excess voltage as heat when the Vbatt is great than your setting for Vbulb. The FET is capable of bleeding off only so much heat before it burns up. The limit is about 10watts max (AWR can confirm this but I think this is close).

As you can see Flakey is running his bulb at 12.79v and the four li-ions are providing about 14.4v under load when fully charged. That's a difference of 1.61v. The bulb draws about 9.2 amps at 12.79v and that calculates to over 14watts going to the FET with fully charged batts. That's too much heat.

By introducing a little resistance (about 100 milli-ohms?) Flakey has lowered the overhead voltage to about 1v. At 9.2 amps that's about 9 watts being bled off by the FET. As the batteries discharge and their voltage drops the watts on the FET continue to decrease.

I'm not sure of his setup, but Flakey has simply added a resistor to the circuit to relieve the load on the FET. I wouldn't say the resistor is the "first to possibly fail component." I'd say that designation deservedly belongs to the bulb!

 

[Flakey]

Re: Mag625R-4d (full write up PICS)

hehe you guys are giving me way to much credit here, The solution was 'dialed in' by my main man, AWR. Yes this is the kind of service you get when you buy one of these drivers. he MADE it work. the tailcap resistor is 75 miliohms. so yeah im putting around 11W on the fet. THAT is not the problem. it can handle 11W eisily, it is the start up surge that will kill it. AWR has recorded bursts of MORE THAN 40 AMPS! during start up, were talking miliseconds here but none the less, this is serious stuff. It took ALOT of fine tuning, and andrew has a pile of over 20 blown FET's sitting on his work bench to so for it. but such is the price of progress. The color temp on this light is the whitest i have ever seen from a non filtered incandescent. it simply blows me away every damn time i turn it on. It is so white because it is litteraly being driven only a vew hundreths of a volt away from the melting point of the bulb. And POWER, man tonight i burnt through, an Icee cup (styrophome cup for frosty drinks) a LOT of paper, I lit my friends cigerette (took some practice but we got it to work) .... its truely insane what this kind of wattage can do. ... oh dear the power ..... all the time on this forum i hear "It turns night into day" but rarely do i believe it. However, in 12' o clock sun, if i hold this light as high as i can (around seven feet) and point it at the ground, i can see a readily noticable hotspot. so at NIGHT with dark adapted vision, whell you get what i mean ... anyone in the southern california area, id be happy to arrange for a night hike sometime, =)

as far as a parts list .... whell in alot of ways i think this is a one off. The tailcap resistor, is a big part in that. and the D cell lithium ions are going to be hard to find untill we can get them in mass quantities. However the thing is acutally a easy build. Alu reflector, Boro lense, 64625 osram bulb, High power kiu W/AWR driver, and ... the modded tailcap with .0075 ohm resistor. at that point all you have to do is dial in the Bulb V and light it up.


Whoops sorry about the 623 typo, yeah in the end the 623 wasent capable of being overdriven enough by only 4 cells, maybe with 5 it would be a different story! YIKES though good luck trying to fit in an fet with that kind of absorptive power ..... for the V that we can make with 4 cells, the 625 is near perfect.

 

[andrewwynn]

Re: Mag625R-4d (full write up PICS)

Q&A..

delvance.. i've been trying to get the 623 to work with the hotdriver.. but it wants over 35A to start it up and with the small FET i use it has a 35A package limit... the 625 will start ok with 25A (compared to 15A for the 138).. but the 623 just pops FETs like popcorn.

The 625 lamp has something goofy with the specs because it has the same life rating and voltage as the 138, yet it can't handle nearly the voltage, not even close.

I've blown every 625 lamp i've tried to run even with soft-start at 12.9V.. i was able to get it to run with 25A current limiting startup at 12.88V (number chosen purely for 'show and tell' that is the value that mathematically works out to 3000.0L (torch).

I had flakey dial down to about 12.72V to reduce the odds of flashing the lamp since it's the only one he has 'til he gets a re supply shipped.

What host are you using for your 623?

JWL.. My prototypical "Mag625" is a 3D 3-bore host with 11x GP2000 cells.. the voltage is properly matched and it doesn't need the tailcap resistor (more on that later).. I will be selling the battery adapters soon.

recipie:

(1) tri-bore 3D light
(1) finned head
(1) alum reflector.. bored out to 12mm
(1) hotdriver high power with KIU (you can also buy the KIU separately and build it in
(1) battery holder.. you can build your own i have a DIY online
(11) GP2000 cells
(2-3) 64625 osram lamps.

i think that's it..

zehnmm.. wait won't be long, christmas is coming.. the 610 lamp is incredible.. with about 1/4 the size of filament.. it is much nicer beam than the 625 (easier to control).

icebreak:

Well,. since the 625 really wants 11 cells.. and it's just really darn hard to cut 1/3 of a cell off a D LiION w/o flames shootin out.. I did the next best thing.

I calculated the resistance i wanted by this formula:

(14.4V - 12.7V)/9.5A = 178mohm.. my first resistor i built with (4) 150mohm resistors (3W power type).. series parallel (when you put resistors series-parallel with 4 you end up with the resistance you started but 4x the power capacity.. that makes as 12W resistor.. now.. at 9.5A x 150mohm.. it overdrives the resistor pack a bit at 13.5W... but it is heatsinked direclty to the tailcap.

I found out quickly i didn't need that much voltage drop. (i build in a voltage probe under the rubber switch cover where i (and now flakey) can measure the vBat under load.. and found out that with battery resistances and such i didn't need that much resistance so i cut half off and now it's 75mohm (flakey had an extra 0 in there saying 7.5mohm)..

Now.. at 9.5A it's 6.7W on the tailcap which makes the lamp over 95% efficient in spite of throwing away 1/3 of a cell on a resistor.

The resistor pack is rated for 6W and i'm pushing it just slightly over, but it's not that likely to fail, it's a tough little monkey.

It gives the FET 6.7W lower power demand, and is a big of a 'hack' compared to a more elegant solutinon like winny';s PIR but i still like and prefer the 'plug and play' nature of the hotdriver even with it's limitations,.

Bill..

The 625 actually pulls closer to 9.5A at these voltages.. the FET i use actually is rated for 120W.. but you'd need some active cooling to keep it that high.. the problem is.. that in the confined space in a mag head.. and even with heat sinking to the body very well.. it works out.. that with about 4W you can run 30 minutes w/o over heating the driver to shutdown.. but at 10W.. you get about 2 minutes.. so.. i aim to have less than 10W initial that tapers off to 4W within a few minutes..

with this solution.. actually there is usually only about .4V overhead or about 4-5W of heat on the FET average.. and when the batteries are 1/2 spent you can swap the tailcap for a standard one to get full regulation back.

It was actually me of course that developed using the tailcap resistor.. simply because i had 'moved on' from the 62138 lamp to the 64625 and just had to get it running in my 4D light with LiONs.

Flakey.. cool.. you actually got a cigarette to light.. seen a picture of a USL doing that but i've had tough luck trying to ignite white items.. my favorite though is shooting at my heel of my black leather shoes.. takes 2.1 sec for smoke to wisp off.

This same mod should work without the tailcap resistor with the 138 lamp.. and if i can ever figure out the right current startup and a suitable FET it'll work with the 623 lamp as well..

The reason i gave up on the 623.. is that the 625 kicks it butt.. i did a side-by-side with a direct drive of the 623 from 4D LiON and my Mag625 running on 11xGP2000s at 12.5V.. and the ceiling bounce had about 10% more light from the 625 and not even in the same class of CCT, the 623 just looked yeeLOW.. the 625 is brilliant unbelievable white!

So.. flakey actually has the one and only mag625 driver right now.. i dropped my Mag625 and knocked a wire loose (it was one of the hand-made models).. and my other Mag625 was using the tester driver that now flakey is using.

The current limiting is fascinating to see in action.. the voltage rises using a soft-start curve, but if and when the current gets over the set value the current is 'clamped' to that level and the voltage actually makes an 's' curve as the filament warms and rises in resistance.

The 138 works great with a 15A limit.. it'll start in about 200 miliseconds.. if i try to start the 625 that low of current.. it takes maybe 500-600ms.. but the small FET i'm using can't take the heat (quite literally) and poof.. there goes another FET to the FET boneyard.

I have used the light w/o the tailcap resistor at all.. just won't run very long before the driver shuts itself off from over heating.

-awr

 

[Aaron1100us]

Re: Mag625R-4d (full write up PICS)

Stupid question. How are you getting 12.79 volts out of 4 1.2 volt batteries? Am I missing something? Does it have to do with the regulator or are the batteries more than 1.2 volts? What kind of batteries are you using? Thanks

 

[andrewwynn]

Re: Mag625R-4d (full write up PICS)

yeah you missed something.. those batteries are LiON 3.6V batteries.

 

[Icebreak]

Re: Mag625R-4d (full write up PICS)

wtraymond -

Thank you. As usual, I learned much from your post.

Flakey -

Nice follow up. I enjoyed every word of it. The description of color temp was perfectly explained.

Andrew -

Thanks for the detailed reply. I actually understood your answers which is almost as amazing as the work you've done here. On the tailcap resistor module...to me, sometimes un-elegant is elegant. It's kind of like a separate master valve on the whole current system. For a "hack" it has value if one understands what it does, meaning that it helps in understanding ckt design. Of course it has functional value as well.

Thanks for the calculation.

This project is just plain interesting, staight up and straight down.