Help me with my Mag P7 build

What would the best Vf (using a P7) be for the SB driver here? I'll more than likely be running 3 x 1.5v = 4.5v input....right in the middle of the specs.

Also, if you can tell me why that's the case, that would be helpful.

Re: what Vf P7 for SB driver?

Sprinkles said:

What would the best Vf (using a P7) be for the SB driver here? I'll more than likely be running 3 x 1.5v = 4.5v input....right in the middle of the specs.

Also, if you can tell me why that's the case, that would be helpful.

Click to expand...

I = 3.25v ~ 3.5v
J = 3.5v ~ 3.75v

I would use the first suggestion DSWOI .The lower the VF the better as terms of brightness and runtime. Also I wouldnt suggest using Akaline batteries as they will not provide enough current for the P7. Three NIMH D batteries will provide more than enough current to your P7. When I built my P7, I tried to cut some corners by buying some cheap Tenergy batteries. Hot off the charger, I only pull about 1.50 instead of 2.5-2.8 amps. So save yourself the trouble and buy some Lsd accuevolution batteries.

Richard

Re: what Vf P7 for SB driver?

They would only register about 3.6v - which is enough to run on, so given what that driver is doing, as low a vf as possible would really benefit me here. Can I reasonably expect 500-600 lumens from a D-bin like this?

Thanks!!

Re: what Vf P7 for SB driver?

Sprinkles said:

They would only register about 3.6v - which is enough to run on, so given what that driver is doing, as low a vf as possible would really benefit me here. Can I reasonably expect 500-600 lumens from a D-bin like this?

Thanks!!

Click to expand...

Fully charged NIMH will be at 1.4 each totaling 4.2 volts. Realistically should expect 500ish OTF from a D-bin. Akaline batteries will not provide the same brightness nor runtime.

Hi Sprinkels,

I think using the DxxxI bin would be a better choice for the SB driver.

The reason I say that is you will likely drive the emitter at max for a longer period of time. J bin emitters have a VF of 3.75 ish while I bin has 3.5. So when the battery voltage gets below 3.75 after some run time the light will go direct drive and your MA will fall from 2800, and continue to go down until the cells are dead. The I bin will run at 2800 ma until 3.0-3.5 ish volts.

I also recommend using quality rechargeable NIMH cells for this. In the long run they pay off, plus you will have fun with the light more often if its not costing you anything to run it.

5-600 lumens is a reasonable expectation from a maglite using a D bin P7, provided your work is good. What I mean by that is low resistance. Use thick wire (for this I would use 22G silver coated teflon wire). Make sure your heat sinking is good. Use quality cells to power it. I use deoxit gold on all the metal to metal contacts like the tail threads and on the spring.

Aren't the D binned P7 emitters rated around 700~900Lm @ 2.8A? :thinking:

M@elstrom said:

Aren't the D binned P7 emitters rated around 700~900Lm @ 2.8A? :thinking:

Click to expand...

Well...yes, but....that's not OTF (out the front), heat loss, resistance...etc.


Thanks to all for the info.


I'm actually contemplating using this pack to power it - HERE. I think it would be a good little setup, with minimal complication. It wouldn't have great run times, but it would be simpler than using three cells that only deliver 3.6v. (The measurements add up - the pack is 183.5mm, and three D cells would be about 183-185mm) The only concern is, can the pack handle the current draw??

If I were to use some AccuNiMH's, I know they would give much more run time, but would they be able to hold their voltage high enough for that driver for very long? They're only 1.2v, and that doesn't give much room for any sag or voltage loss to stay within the driver specs. That's why I'm considering just using that pack instead.

(Also....I will probably give this to my dad as an "after surgery" gift...he could use the awesomeness to cheer him up. That means - less is more. I don't want him to have to worry about multiple cells complicating the usage. I can just buy the charger for the pack and call it good, simplifying the build and usage.)

Accupower D NiMH cells can hold over 1.2V at 5A draw. At 2.8A draw, you should be close to 1.25V per cell, or a Vbatt of about 3.75V for a 3D Mag.

The main issue is reducing parasitic resistance in your Mag so that at 2.8A, you don't suffer excessive voltage drop.

Certainly, using 4xNiMH would give you plenty of voltage headroom, with lower but perhaps (barely) acceptable driver efficiency. The NiMH battery stick that you linked, however, is too much. It uses 5x1/2D. Five NiMH cells are going to start at say 7.5V, and run at about 6V quasi steady state. If the high initial voltage doesn't kill the driver, the high steady state voltage will result in poor driver efficiency and a lot of heat generation.

The ShiningBeam driver is the usual AMC7135-based driver and those AMC chips have a max input voltage rating of 7V and a recommended max of 6V. Any voltage over Vload is just wasted as heat. Thus, the goal is to provide enough Vbatt to satisfy Vload (i.e., Vf) plus some overhead to address parasitic losses and to stay in full regulation for most of the run time.

When you use 3xNiMH, you want a low Vf LED to maximize the voltage overhead that you have available.

If you run an 2.8A AMC-based driver at 6V steady state driving a P7 having a Vf of say 3.3V at 2.8A drive, your driver efficiency is going to be essentially 3.3/6, or about 55%. That's not very good. You are generating about 7.5W of waste heat.

The closest match is 3xNiMH when using an AMC-based driver to run a P7 at 2.8A. That is probably the preferred battery configuration for a D Mag P7/AMC7135 mod, with the caveat at the top of my post that you reduce sources parasitic resistance/voltage drop.

I was looking for more information about the pack, given the 6v rating, I figured it would charge to about 7v. I threw it out there as an option, hoping someone would be able to comment while I researched. THANKS!!!!

I'm leaning toward the 3xNiMH option, (Accu cells seem to be the standard option.) I just need a "reliable" and cheap charger for those cells. I can't have anything fancy for it.


So a J bin (3.5-3.75v) would actually give optimum driver efficiency provided the Vi stayed good at the 1.2v+/-. But since there's some loss through resistance, the I bin gives a little headroom - am I getting that right??

Build an XP-G D Mag mod, drive it at 1050mA with a 3xAMC7135 driver, allowing you to use D alkalines. No charger, no charging hassles, go anywhere and get D alkalines, which will last a long time anyway at a mere 1A draw. Later, you can migrate to NiMH if you wish. You can use Eneloops in D shell adapters. That will allow you to use a standard AA NiMH charger, which are ubiquitous. The 2000mAh Eneloops will still run for ~2 hours in regulation.

At a Vf of probably 3V, you should have plenty of voltage headroom to run in regulation on 3 cells, alkaline or NiMH.

Using 3 Accupower Nimh LSD cells at a 2.8A load the voltage under load of those cells would be about 1.25v for the C cells and about 1.275 to 1.3v for the D cells. I linked both of the data sheets as you didn't state what size Mag host you were going to use (e.g. 3C in 2D host, or 3D in 3D host). So you're looking at 3.75-3.9v before you run into switch/spring/mag body resistance. AMC7135 linear regulator drivers need a input (Vin, battery) voltage of 0.12v higher than the Vf of the LED to stay in regulation. So ideally you should get a D***I emitter. The excess voltage input is shed off as heat in terms of watts, which is calculated by the voltage difference X amps. So if you had a Vf of 3.5v and a Vin of 3.75v, the heat generated would be 0.7 watts (hardly anything at all).

I would recommend heatsinking the amc7135 chips just in case for longetivity of the driver. I did a recent build utilizing the same SB driver and pictures of the heatsinking can be found here.

The problem with a case of Vf~3.5V and Vbatt~3.75V is the voltage overhead to run in regulation for AMC7135 LDO regulators. As you say, it is 120mV according to the datasheet. Thus, any parasitic resistance that drops the input voltage by more than 130mV knocks the light out of full regulation. At 2.8A draw, you need less than 50 mohms of parasitic resistance in the Mag flashlight circuit path to lose that 130mV.

I would either somehow get a sorted P7 that has a low Vf at 2.8A drive (say, closer to 3.25V), run the P7 at a lower drive current to reduce the Vf, or use a different LED with a lower Vf (like the XP-G example I gave earlier).

Since this is a gift for a person who sounds like he is not a flashaholic and is not interested in dealing with recharging three 10Ah NiMH cells, my choice would be the XP-G build I described earlier. I'd go for a neutral white 5B tint XP-G, which is available in R4 flux bin.

LilKevin715 said:

I would recommend heatsinking the amc7135 chips just in case for longetivity of the driver. I did a recent build utilizing the same SB driver and pictures of the heatsinking can be found here.

Click to expand...

That's a nice build but the heat sinking IMO appears to be questionable. The AMC7135 is heat sinked through the mounting tab. However, it looks like you thermally glued heat sinks to the top faces of the chip cases. It can't hurt, but the thermal resistance through the case is most likely very high and thus the benefit derived may be less than expected.

The person, although not a flashaholic, is certainly someone who would appreciate the complexity and effort put into this build. The charging of the cells isn't a big deal since he and his wife (my parents) charge other batteries all the time for cameras, flashlights, remotes....etc. He wouldn't mind the effort. He might like it more actually. It would be something different and offer a more involved experience. I might even be doing this build with him, or at least sharing what little I know of the process. He can't do much right now, so sitting around and watching something INTERESTING is better than
tv....

Besides, I want something a little different than the xpg. I've messed around with them enough, and I would like to do something with a P7 now.


I was considering potting the driver to reduce any heat, but I would think it would be negligible given the low overhead. That's what I'm hoping anyway. Say...anyone have one of those DxxxH emitters lying around??? hehehe

The "complexity and effort" of an XP-G Mag mod vs P7 mag mod seem basically identical to me. The main differences are lumens output, throw, and cost.

My estimate is that the P7 will deliver more OTF lumens but less throw.

The XP-G mod will probably be less expensive.

The XP-G probably can use the stock reflector and give you an acceptable beam. IMO although the P7 also can get away with using the stock reflector, I think the beam quality is low enough to be noticeable and bothersome. YMMV. Thus, you may opt for a $15 KD OP metal reflector.

To keep cost low for either build, you could get a pack of KD glass windows, or go for the big time and get a flashlightlens.com boro window.

The XP-G build saves money on the upfront cost of D NiMH cells and a charger.

For the P7 build, if you can't get a pre-sorted LED of known low Vf, I'd probably opt to drive it at something like 2.1A or 2.4A max instead of 2.8A, and possibly gain an additional 0.05V to 0.1V of headroom.

Hi Sprinkles,


Let me start by saying that Justin is right about that pack the voltage is way to high for the SB driver. I have used many many of these drivers and IMO 3 nimh cells is the best way to go. I use 3 accupower C cells in a 2 D mag, but for simplicity of build and long run time I think 3 D may be a better way to go with this project.

If you go any higher with the voltage from the battery to this driver you will need robust heat sinking, otherwise the heat will cause the current to the emitter to plummet like a rock off a bridge.

Here are a couple of links for the emitter you asked for.

I recommend these they are a CSWNI bin and the color is great!
http://www.cpfmarketplace.com/mp/showthread.php?t=218293

If you are just dead set on getting the max output these will give a slight gain over the C bin. (it takes a few weeks to get the KD ones)
http://kaidomain.com/SearchResult.aspx?SearchKey=p7+emitter&CategoryId=-1&SiteId=1

I personally just cut the cam off of the factory reflector and use it for mods like this, the finish on them is very nice so light output is good.

Later you can always try a MOP for a smoother beam, but there is some loss of light.
http://kaidomain.com/ProductDetails.aspx?ProductId=5632

As for a lens I don't recommend getting the cheap glass ones from KD or DX. I have found that even the stock Mag lens give better light transmission than those. I use Borofloat in all my mods now. I think it is far better than UCL because the coating on the UCL is very easy to damage, and the boro has almost exactly the same light transmission.

http://www.flashlightlens.com/str/index.php?app=ccp0&ns=prodshow&ref=boro_lens

The 52x2mm is the one you would want.

Hope this helps.

I agree with vesture that a boro glass window is the way to go. A plastic window can have decent light transmission, but only as long as it remains clear and unscratched. How long is that going to last? I probably should have suggested the KD glass window pack (IIRC, five glass windows for less than $1 per window) only if you were planning multiple builds. Then you start to save money. I'd prefer a KD glass window over the stock plastic window for the scratch resistance of glass. For a one-shot build, I'd spend the money on a boro window. StefanFS has built many 3D P7 Mag mods using 8xAMC7135 drivers. Based on his experiences, 3xD NiMH combined with an I Vf bin P7 should work fine. IIRC, he recommends a 4C P7 Mag mod over a 3C mod when using C NiMH cells.

If you don't want to hassle with this degree of Vbatt to Vf matching, you can use a PVC sleeve, install say 3xIMR26650, and do a tail spring mod to fit the cells. Use a quality buck driver like a Shark Buck 3A or H6CC (use a Remora or d2Flex, respectively, for multimodes). Use a SharkSink to sink the driver to the Mag tube wall.