Feedback/impressions on the PhD-M6 custom battery pack

[archimedes]

Wow, I just stumbled across this thread - this is so cool!

I did have one question (and apologize if this has been asked and answered, I tried scanning through the 15+ pages of discussion, but might have missed it), namely, will this work with the AW soft-start three-level M6 tail-switch?

I am constantly amazed at the things that are created here on CPF....

[oldways]

No. It will NOT work with the AW soft start switch.

The PhD-M6 pack has built in soft start.

[archimedes]

No. It will NOT work with the AW soft start switch.

The PhD-M6 pack has built in soft start.

Yeah, that's what I had guessed. I might have to look for a second M6 tailcap, as I'd rather not have to uninstall and reinstall the AW switch. Those don't seem to pop up too often on CPFMP, though.

Thanks for the info.

[wquiles]

Thanks much oldways

[LuxLuthor]

Just wanted to give you ongoing praise. This has been the best thing for my M6's since sliced bread. They tend to squash the sliced bread though.

Thanks for making and supporting these regulators.

[DM51]

Good to see this bumped. I agree with LL and others - this is easily the best solution for the M6.

[wquiles]

Thank you guys

[Tim W]

Good to see this bumped. I agree with LL and others - this is easily the best solution for the M6.

+1!!!!

I've not posted much (if anything) about mine, but it does what it does fantastically well!

While I KIND OF miss running an 1185 with a FM 3x17670 holder, I LOVE THIS THING WITH the MN61! New cells give easily 30-35 minutes continuous beautiful WHITE light.

[oldways]

+1!!!!

I've not posted much (if anything) about mine, but it does what it does fantastically well!

While I KIND OF miss running an 1185 with a FM 3x17670 holder, I LOVE THIS THING WITH the MN61! New cells give easily 30-35 minutes continuous beautiful WHITE light.

It will run the 1185 well at 10.8v

[DM51]

All we really need to make it totally perfect are some AW IMR17670s, so we can run the MN21 without overstraining the cells...

[Tim W]

It will run the 1185 well at 10.8v

Yeah, I know, but not with any kind of regulation which was the biggest selling point for me.

[Flea Bag]

Hi guys, I've been away for a while and now I come back to find that the SF incans are discontinued... I quickly ordered another M6 for keeping and then stumbled upon the completion of this thing! I actually contributed quite a few thoughts back in the early feedback/design phase and I'm really happy it's turned out the way it has!

Right now, I'm considering which voltages I should go for. I have a preference for bulbs that run at least an hour on the original 6xCR123 configs but at around the 200 lumen level. As such, my favourite bulb is the MN20 (newer version with oval hotspot). Also have one of each: MN15, MN16, MN21, MN60, MN61, MN10.

Decided so far on:
Level 1: 6.8v
Level 2: 7.4/7.3v (slightly lower than 7.5 for better durability of MN20, MN15 bulbs?)

It's level 3 and 4 that I'm undecided about. I like that 10.8v allows usage of some WA lamps as well as Lumens Factory LAs but I don't own any of those yet. What I'm more interested in are the MN60 and MN61. I've read impressions about the MN20 vs MN60 and though the 60 has better endurance, it also seems to yellow sooner than an MN20 on primaries. I can always test this myself but I trust your impressions.

But with the new PhD and its regulation, how does the MN60 at 9.8 or 10v compare to the MN20 at 7.4/7.3V? Does it still yellow sooner? I like that the MN60 is comparable to the MN20 but has the potential to run quite a bit longer.

[wquiles]

The best guide as to what to expect was created by Eric. His table compares many bulbs, in regards with their output, current draw, recommended drive level, etc.. We are not recommending using all of those bulbs with the PhD-M6, since many of them really go past the safe 2C load on the cells, but it is still a very good summary, and a good way to compare bulbs, so that you can have some options. Updated bulb charts/tables from Eric:
PhDM6_Bulb_Chart_8_21_10.xls

Thanks to forum member and contributor Justin_Case, we also have available the same information as a graph/photo (PNG format):
http://m3coupe.com/electronics/PhD/p...rt_8_21_10.png


I run the MN61 in my personal pack, and from memory it is set at 9.2 volts or so, and I love it. Nice, bright, white team, great output level, is not too hard on the cells, and the host does not get too hot on very long runs. It is my favorite bulb on the PhD-M6 for "actual" use

Will

[Flea Bag]

Thanks Will... I did use Justin_Case's list to arrive at what I'm at so far. (With great appreciation)

I'm still not certain about how underdriven the respective bulbs are when at the lower end of the 'recommended range' though. For example, the MN20's range of 7.2 to 8.1v -Are the bulbs quite yellow at 7.2 or still quite white, similar to running on 60-70% charge CR123s?

At this time, I'm probably going for two packs, the first of which is:

1. 6.7v (MN16, MN21)
2. 7.3v (MN15, MN20, Lumen's Factory bulbs)
3. 9.2v (MN61)
4. 10.0/9.8 (MN60)

You'll notice how I'm about 0.1 to 0.2v lower than the common 7.4/7.5 and 6.8v. Would you think this would be of minimal impact to whiteness while helping a bit in physical durability (vibration/accidental light impact?)

[wquiles]

I think that the range Eric listed is a good guide as to what is practical/doable from under-driving the bulb to over-driving the bulb, but perhaps is best to look at the Recommended Drive Level for the MN20 being at 7.5volts, MN21 at 6.8volts, etc.

Most folks have used those recommended values and have reported great success, but a 0.1/0.2 volts difference is not very significant, so if set a tad lower than the "recommended" value, the impact will be minimal (I doubt you will be able to notice with your bare eyes, plus you of course have bulb-to-bulb variations, etc.), but of course I will program your pack(s) as you request.

Will

[Flea Bag]

The PhD uses a buck converter correct?

If so, what is the max voltage it will take and would it work just as well with 6 x CR123 primay cells, assuming I stick to within the 1C to 2C discharge range?

[Justin Case]

Part of selecting the desired voltage is the bulb's rated filament life at the rated voltage. When you have a bulb with a very long life like 4000 hrs, you often have a fair amount of leeway for overdriving it without undue risk of instaflashing the filament. But if you have a lamp that is rated at say 6V and 20 hrs filament life, then things get dicey. There are equations available to re-rate incan lamps, but they get iffy as you get into the "margins". When you are already at a rated 20 hr filament life, it doesn't take much overvoltage to get to a calculated filament life in the low single digits. Then the question is what is the reliability of that calculation? Is it really 8 hrs life or is it 2 hrs or even negative hours (an instaflash)? How robust is that estimate relative to lamp manufacturing variances, i.e., will you be fooled into thinking the setting is reliable because one lamp works fine but with statistics of larger numbers it turns out that X% instaflash, where X is a non-trivial number?

The problem with the MN21 seems to be that it hasn't been entirely characterized in terms of its V-I behavior and its rated filament life. I tried Googling for the MN21 filament life, but couldn't find anything authoritative. Is it 5 hrs, 10 hrs, 20 hrs, 50 hrs...? I don't know. What is the rated voltage? Is it 6.0V, 6.1V, ...? I don't know. But I do know that the specific value for these variables can have a large effect on filament life for overdriving.

mdocod has examined a lot of incan lamps and is very careful in his recommendations. So 6.8V is probably a good one. But your requirements may differ from his. Presumably he had a particular balance of output, current/power consumption, heat generation, and lamp life in mind. Since SF has phased out its incan line, your weighting for lamp life may be different from when mdocod's table was constructed in 2010.

If I assume a rated voltage of 6.4V, a rated bulb life of 20 hrs (this is a typical value for SF's P60 lamp), and a bulb life re-rating exponent of 16, 6.8V gives a calculated lamp life of 8 hrs. 6.7V gives 10 hrs. Perhaps not much difference. 6.6V gives a calculated life of 12 hrs. The calculated color temp and bulb lumens also don't really change that much.

The way I see it, you could stick with whatever is closest to the "factory" drive level (is that 6.4V?) to use SF's choice of output/life balance. If you want a bit more output while maintaining some reasonable assurance of filament life reliability, then perhaps go with 6.6V, keeping you in double digits for calculated life (it seems that when the calculated filament life goes into the single digits, that's when you start to get into the margins and sometimes bulbs instaflash unexpectedly because of variations in manufacturing tolerances. You get a slightly weak bulb, and it flashes). I personally would not go beyond 6.8V unless I knew that SF rates the MN21 life at say 50 hrs. It all depends on how much incremental increases in lumens are worth to you vs decreased lamp life. I personally don't think you'll get anything noticeable between 6.4V and 6.8V to compensate for the decreased life. Similarly, I don't think you are going to get an appreciable change in color temp that outweighs the decreased life But you will great some increase in lumens and some whiter output.

[wquiles]

Justin - great reply. Thanks much for your great insights

The PhD uses a buck converter correct?

If so, what is the max voltage it will take and would it work just as well with 6 x CR123 primary cells, assuming I stick to within the 1C to 2C discharge range?

Yes, a buck down converter, but the hardware/board (voltage dividers) and software (taking into account discharge rates) were programmed to work with 3x fully-charged LiIon cells plus a little margin (around 14 volts or so, if my memory is right). I have not done tests nor approved using anything other than 3x LiIon cells, so I would not recommend using 6x CR123 primary cells.

Will

[Flea Bag]

Great post Justin... Thanks for taking the time. I did some hard work and read the entire thread too and stumbled upon mdocod's recommendations for a member who may have wanted a little more safety. You're right in stating that I want to be more careful with my voltages now that the M6 is discontinued and we're not sure of how may lamp assemblies are available out there.

Will, I know it's not tested but is there any reason why the PhD-M6 can't use 4 x CR123 with two dummy cells?

Also, will the PhD-M6 fit 2 x 18650 cells and a thin dummy? I'm interested in the new 2600 and 2900 mAh capacities AW/Panasonic has available and plan to use them with the MN21.

[wquiles]

Will, I know it's not tested but is there any reason why the PhD-M6 can't use xyz cell(s)?

Yes - the pack was designed, tested, and calibrated to work only with the 3x 17670 LiIon cells. Preferably the ones from AW.

Myself and Eric (and the Alfa/Beta testers) took great care and spent a lot of time testing the pack with the approved 3xLiIon cells - ZERO time spent on any other combination. No other combination is approved for the PhD-M6 pack.

Will

[Flea Bag]

My apologies for sounding tiresome about it...

I think I've figured out what voltage combinations would work best for me. Just need to work out other bulbs, parts and accessories I'll need before I place an order... Thanks again Will and Justin!

[wquiles]

My apologies for sounding tiresome about it...

I think I've figured out what voltage combinations would work best for me. Just need to work out other bulbs, parts and accessories I'll need before I place an order... Thanks again Will and Justin!

If you were to try other combinations/cells/etc. that we have not tested, the pack might be damaged, the bulbs might blow up, the regulation points could be off, the battery low level would certainly don't work as well, etc., etc., etc. - we just don't know. No product is perfect, and I am not claiming that my PhD-M6 is perfect, but based on the 40-50+ packs in use all other the world, the pack seems to work really well with the 3xLiIon cells, and the bulbs/voltages Eric came up with, so that is why I recommend staying within that "envelope"

Will

[Flea Bag]

If you were to try other combinations/cells/etc. that we have not tested, the pack might be damaged, the bulbs might blow up, the regulation points could be off, the battery low level would certainly don't work as well, etc., etc., etc. - we just don't know. No product is perfect, and I am not claiming that my PhD-M6 is perfect, but based on the 40-50+ packs in use all other the world, the pack seems to work really well with the 3xLiIon cells, and the bulbs/voltages Eric came up with, so that is why I recommend staying within that "envelope"

Will

I get you Will... No worries! I've got some engineering background too so I know where you're coming from! I'm still working on the final voltage at the moment...

[mdocod]

Hello Flea Bag... err... that doesn't sound very nice,

(now why are you making me call you that? just doesn't seem right!)

hehe.... anyways. I was just checking email and noticed that Will had sent me a heads up that there was some activity that may well warrant another opinion on the matter.

As I recall, I think most SF bulbs were rated by SF to be 30 hour lamps (or maybe it was 50? can't recall). Keep in mind that, bulb life ratings for bulbs intended to be direct-driven by batteries are always going to be rough estimates because there are so many variables. CR123s when used to drive bulbs direct drive, can be somewhat abusive because they tend to deliver higher inital voltage and then settle down within a few seconds (or fraction of a second depending on the load and temp and state of charge etc etc). Matching bulbs to CR123 cells requires leaving enough overhead in the bulb design to handle higher voltage peaks than when matching bulbs to a constant voltage source. Temperature and state of charge have a huge effect on drive levels.

When selecting "recommended" voltages, my goal was to try to make a reasonable estimate of where a constant voltage drive level with soft start, would roughly match the bulbs originally intended life on CR123s. I can't guarantee anything because there is no concrete published data on SF bulbs, however, I suspect I have come respectably close to that goal. The idea being that, we can get away with a slightly higher average drive level than the bulb would have seen on CR123s, because we have traded out the abusive characteristics of direct drive.

That said, if you want to get the MN21 running closer to it's average drive level on CR123s, 6.6V would technically be closer, for the MN15/20, 7.5V is probably pretty darn close but dropping a tenth wouldn't hurt either. The MN15/20 bulbs, IMO, are likely to have more overhead in their design because they would normally have been subject to higher voltage spikes as a result of the way CR123s behave into those loads.

As far as the CCT is concerned.. Assuming that a MN20 is running ~3325K at 7.5V (a complete and total guess; please keep this within context as it is for comparison purposes only), then at 7.3V, it would drop to ~3300K. If they were side by side, you might be able to notice a difference, but if you had any time between comparing them, I think it would be pretty hard bordering on impossible to identify which was which.

One last consideration: I wouldn't personally bother with a 10.8V setting. I have it on my packs, but never have any desire to use it because it really does not provide any meaningful regulation, we just don't have enough voltage overhead to work with. Also, if I had to choose between a ~7.5V>MN20 vs a ~10V>MN60. I think I would be tempted to lean towards the MN60 if beam pattern differences were not an issue. For reasons which I do not have the gumption to explain at this time, a higher voltage, lower current, similar wattage, albeit still primarily regulated setting, is going to make better use of available stored energy

Eric

[Flea Bag]

Thanks for popping in Eric and to Will for the heads-up... It certainly seems like all the bulbs I'll be using will be 10v or lower.

After consideration and the your kind inputs, I'd like the following levels:

1: 6.7v
2: 7.3v
3: 9.2v
4: 10.0v

That and international shipping please...

Total $195 + $15 = $210. Let me know when I can send the payment!

Thanks again Eric, Will and Justin... I know I'm late to the party, but in case anyone wants to know, my voltages above are prioritised purely for the SureFire HOLAs and LOLAs (LOLAs a bit more for durability and a little less 'wow' since they're meant more for endurance anyway) with levels 3 and 4 specifically for the MN60 and MN61. My figures were derived largely from Eric (mdocod), Justin_Case and Will's posts throughout this thread.

[wquiles]

Thanks for popping in Eric and to Will for the heads-up... It certainly seems like all the bulbs I'll be using will be 10v or lower.

After consideration and the your kind inputs, I'd like the following levels:

1: 6.7v
2: 7.3v
3: 9.2v
4: 10.0v

That and international shipping please...

Total $195 + $15 = $210. Let me know when I can send the payment!

Thanks again Eric, Will and Justin... I know I'm late to the party, but in case anyone wants to know, my voltages above are prioritised purely for the SureFire HOLAs and LOLAs (LOLAs a bit more for durability and a little less 'wow' since they're meant more for endurance anyway) with levels 3 and 4 specifically for the MN60 and MN61. My figures were derived largely from Eric (mdocod), Justin_Case and Will's posts throughout this thread.

No problem - glad we were able to help

The ordering information is in the Sales Thread:
http://www.candlepowerforums.com/vb/...m-battery-pack

Basically post your order in that thread, claiming the next # (19, I believe), and send Paypal to the email address listed there.

I should be able to work work on your pack and get it done this weekend, and then notify you when it ships.

Will

[Flea Bag]

No problem - glad we were able to help

The ordering information is in the Sales Thread:
http://www.candlepowerforums.com/vb/...m-battery-pack

Basically post your order in that thread, claiming the next # (19, I believe), and send Paypal to the email address listed there.

I should be able to work work on your pack and get it done this weekend, and then notify you when it ships.

Will

Sigh... I got carried away... Will post in the sales thread!

[wquiles]

No worries

[Flea Bag]

Okay! My pack has arrived and I had a nice 47 minute run with the MN60! But first the following words:

From the French movie Taxi:

"A tourist who isn't scammed isn't a tourist" or something to that effect.

Here's mine:

"A flashaholic who hasn't blown a bulb isn't a flashaholic".

Up till now, I've yet to blow a bulb. So...
Perhaps you can tell where this is leading...

I thought I had it all worked out. I thought I read Will's instructions and seen the pictures so many times that I knew by heart which switch positions meant which levels. So for the test run, I placed three newly arrived, freshly charged and properly rested AW 17670 cells into the PhD-M6, selected what I thought was level 2 (7.3V) on the dip switches for the MN20 bulb and turned the light on... What I didn't realise was that I was actually running level 3 (9.2V)!!!

Long story short... I got a very very white MN20 which ran safely for over 5 minutes continuously until I decided to turn the M6 off. No bulb failure!!! MN20 still perfectly functional!



Okay... Now really I was very lucky with that misadventure not to have blown my faithful MN20. The guy had over 30 hours on it, probably much more. To be honest, when I first turned it on thinking it was 7.3V, I was actually a bit disappointed. It wasn't the brightness or whiteness -that was very impressive. Rather, I was thinking: I should have gone with an even more conservative voltage than 7.3v because I quite like the slightly warmer tints (even for an incan) and want to make my bulbs last a little longer and be physically tougher against little bumps and impacts. But anyway, now that I’ve run the MN20 at 7.3v proper, I’m very happy with the beam tint and would be happy even if it were 0.1 or a little more lower.

I then proceeded to testing one of my main purposes for using the PhD-M6 pack: The MN60, which should give me close to 70 minutes, an improved efficiency to the MN20 at around the same output. I was then planning to test the MN61 and finally, the MN21 where by then, the cells should have been through a few cycles.

Unfortunately, progress was halted with the MN60. I managed only 48 minutes total runtime on my new AW 17670 cells. This was the set of cells that did 9.2V on my MN20 for 5 minutes, then 7.3V on the MN20 for 8 minutes and the remaining 35 minutes on the MN60 at 10.0V. So I thought perhaps the MN20 at 9.2 must have pulled some serious current from my cells during that 5 minute run. I did notice the extra heat of the output at 9.2, but it couldn’t have drawn so much more as to lose around 20 minutes of MN60 runtime!

So I proceeded to use the second set of new 17670 cells on the MN60 and got only 51 minutes until Moon mode kicked in and around 2.5 minutes until protection shut the light off completely. It was at this point that I thought perhaps there was too much resistance somewhere or that the pack was malfunctioning.

The next day, after the first set of 17670 cells had their charge and a rest, I placed the cells in an FM 3x17670 holder and ran the MN60 direct drive. Clearly overdriven, it was a very white tint, for almost all its runtime until around the 47.5 minute mark where tint had dimmed noticeably and relatively quickly to become equivalent to my MN20 2x18650 reference in a Leef 2x18650 body. Then at 49 minutes, protection cut in. I proceeded to test the second set of 17670 cells and it too produced a 49 minute runtime.

Two more runs using the PhD-M6 with both sets of 17670 cells produced 53 and 52 minutes total runtime respectively.

As I type this, I’m using an MN10 in the adorable looking ‘sea-cucumber’ config. to test the health of my cells on an easier load.

[wquiles]

Awesome - looks like you are having tons of fun :-)

Will