5761 w/ A123 - Soft Start Needed??

[RichS]

Ok, after my Mag85 spanked my 5761 with 6 Eneloops (and it was supposed to be my step-up mod), I've got a Black & Decker VPX pack sitting here and I'm ready harvest two A123 18650s from it. Hopefully these batteries will allow me to see the true potential of this bulb.

Before I put these in the light, is anyone running A123s in their 5761 without any kind of soft start and no insta flashing? The reason I ask is I spent the time to solder in a UTC (failed miserably), and was told that other's had my setup without a soft start and had not had a problem with insta flashing.

I would like to not mess with the UTC soft start if I don't have to, but I don't know if the bulb can handle the current in-rush of the A123s without it.

Thanks for your help!

-Rich

 

[LuxLuthor]

As I told you in the PM, with the link to the A123 Data profile showing the voltage ???....so I don't see the point in using two A123 18650 cells here.

The Eneloops have 1.45V hot x 6 = 8.7V

2 x Emoli will be about 4.15V hot x 2 = 8.3V

A123 IMMEDIATELY drop under load to 3.3V x 2 = 6.6V
​

 

[Anglepoise]

Lux,
I am still a little confused. In your incan tests you report that a 5761 bulb 'flashed' at 7.9 volts. ( applied ) .

Then above we know that 6 freshly charged AA's will have a resting voltage of around 8.7 volts. So my thinking from this info is that the bulb should NOT flash as the applied voltage from the 6 AA's will sag immediately to ( in my case) to around 7 volts. I do not understand how RichS could 'flash' his 5761
with 6 AA's. Have I got this all wrong.
Your incan tests were done using a power supply that did not drop with load.
Batteries on the other hand sag immediately.

 

[LuxLuthor]

It depends on the NiMH cells you use, and how well they hold up at a certain Amp rating. You have to look at SilverFox battery testing at various Amp load rates. The Eneloops hold up their voltage at 5-6A, the Elite cells even better.

Remember also the initial startup is with a cold bulb filament. There is an initial millisecond voltage spike when you first turn on switch, then it reaches a stable voltage/current flow. It is that initial spike that often kills the bulb. That is why the NTC or AW's D Driver has the "soft starter" feature. They blunt the spike.

 

[Anglepoise]

Got it. Thanks.

I am using Sanyo 2700's and I think I will need something a little more robust before experimenting with a 5761.

 

[RichS]

Thanks again for the information LuxLuthor. I didn't mean to aggravate you, I sent you the PM after I posted this thread to get more specific information from you on your specific 5761 setup. The reason being that your "Most Powerful Mods" thread inspired me to do the 5761 since you had it listed as a brighter mod (two steps up from the Mag85). However, mine is noticeably dimmer regardless of what I try. I thought you might have some insight as to how you got yours to be noticeably brighter.

Thanks again for taking the time to provide helpful information. I'll try to be less impatient next time before sending out a PM for info.

As I told you in the PM, with the link to the A123 Data profile showing the voltage ???....so I don't see the point in using two A123 18650 cells here.

The Eneloops have 1.45V hot x 6 = 8.7V

2 x Emoli will be about 4.15V hot x 2 = 8.3V

A123 IMMEDIATELY drop under load to 3.3V x 2 = 6.6V
​

 

[richied]

Hi,

let me weigh in here too. Just to add to the confusion, I pm'd Lux also recently with a very similar question...and my name is Rich too so...two diff folks here doing the same things
I just built a 5761 based light in a mag 3c using 2xa123s.(lots of boring with a brake hone and removal of the cardboard around the cells needed) I am using AW's 3 stage pwm switch for a c body.
The light is ok, but doesn't match my 3D based mag85 with Lux's eneloop packs. The trade off is the much nicer size/feel of the 3C mag.

The a123's do drop to nominal 3.3v/cell almost immediately. I picked them for the same reasons as you stated, based on Lux's testing, the emoli's will likely flash the bulb hot off the charger.
I have been looking for a regulator to hard set the bulb voltage to 7.2-7.4 but have no joy as of yet for a c size solution. This would then allow the emoli's to be used.
I just found a 4C mag last week and am thinking about building a 3xemoli light with the 1185. This may be over the flash limit too, but using the low stage of the AW switch may be ok. The rub is the stages are set by quick taps of the on/off switch and it is easy to get to full power.
I'm trying to duplicate the performance of the 3D 1185 in a narrower host size like the C..and make it reliable enough to take camping/backpacking etc..and cheap to run(my m4 devastator surefire with an extended tailcap is 20% overdriven and very nice..but cost $10/hour to run with the CHEAPEST price for cr123s).

Rich D.

 

[richied]

oh, a bit more info,

The runtimes and guestimated lumens with the mag3c, 5761 and 2 x a123's and AW's 3 stage switch are approximately as follows:

low- 1hr, 200 lumens, yellowish but useable.
med- 40min, 400 lumens, nice white
high - 23min 700 lumens, nice white

I'm rounding up a bit on these numbers and guestimating the lumens from Lux's tables, my calibrated eyeball comparison to my mag85 and over driven m4 and the formulas found elsewhere in the forums for lumens estimates.
The medium setting is putting out a noticeably larger hotspot on the tightest focus than the overdriven M4(225 (non overdriven) rated lumen bulb) but just slightly dimmer/yellower color.
The high is as bright as the overdriven M4 with the larger hotspot. The lumen overdrive formula I think peg's the M4 at 475 or so lumens in this config.
By hotpot here ,I mean the apparent brightness seems the same as the M4 hotspot but maybe 25% larger, so more light.

For comparison, the 3D mag85 with Lux's eneloop pack does the following:

High- 43min, estimated 975 lumens( I round to 1000 )

It is significantly brighter than anything else mentioned here on high. I volunteer patrol some bike trails here at night a few times a week, and with the snow on the ground, high is way too bright with the mag85, low is just about right.

I didn't do runtimes on the lower settings.

RichD

 

[LuxLuthor]

RichS....LOL! I did have RichieD and RichS swimming in my head, and didn't realize you posted this thread before the PM....so my bad, sorry. Bottom line is the 2 x A123's will be too low of voltage to be worth doing for the 5761.

The problem that you guys are facing is the 5761 while bright when optimzed, has a very narrow band of tolerable voltage. The ONLY regulated voltage driver that is currently in existence (& I have about a dozen of them) is the AWR Hot Driver...and it works extremely well.

The Lithium battery supplier, AW (no connection to user AWR) has come out with a C & D Mag Driver which offer soft starting, and multi-level output, but is not a REGULATED driver....meaning that you supply a battery with say....9V, and it gets regulated so only 7.2V is delivered to the bulb.

AW is eventually planning on offering a regulated version that has been mentioned in this topic....but in the meantime, your only option is to try and get close to the optimal voltage for your bulb without flashing it....or try to scrounge up an AWR Hot Driver.

 

[richied]

Too many folks named rich

Yup, I have the AW D and C 3 stage drivers and I know they are pwm only and not regulating the output peak voltage, just the duty cycle. I think he is only coming out with a d size regulated one at some point but no c size. I really want to make a C size light with mag85 output levels. If you ever want to sell one of those AWR regulated drivers...send me a PM. If I was really motivated, I should dust off my EE degree and design one.

richD

 

[RichS]

Thanks Lux, this clears it up for me. So to make amends, how's about you sell me one of those regulated drivers!

Kidding, but seriously, what'll you take for one?? Everyone's got a price...



On another 5761 note, I just got back from my local BatteriesPlus store, and bought a couple packs of 2700Mah Powerex NiMH batteries. I read around a little, and from what I've read it seems this battery should be better at handling the high draw of the 5761 than my Eneloops. Hopefully they will have a little less sag and deliver a little more voltage to the bulb. Can anyone confirm this? If there really isn't any advantage to using the Powerex over my Eneloops, I'll just take them back and save $30. I would have gone with the Elite NiMH, but I'm going out of town and didn't have time to order online. This was the best I could find from a local store.

Also, I read in Lux's post that some batteries need to be conditioned. Does anyone know if the Powerex batteries need conditioned? Does that just entail running them all the way down first and then fully charging?

-Rich

RichS....LOL! I did have RichieD and RichS swimming in my head, and didn't realize you posted this thread before the PM....so my bad, sorry. Bottom line is the 2 x A123's will be too low of voltage to be worth doing for the 5761.

The problem that you guys are facing is the 5761 while bright when optimzed, has a very narrow band of tolerable voltage. The ONLY regulated voltage driver that is currently in existence (& I have about a dozen of them) is the AWR Hot Driver...and it works extremely well.

The Lithium battery supplier, AW (no connection to user AWR) has come out with a C & D Mag Driver which offer soft starting, and multi-level output, but is not a REGULATED driver....meaning that you supply a battery with say....9V, and it gets regulated so only 7.2V is delivered to the bulb.

AW is eventually planning on offering a regulated version that has been mentioned in this topic....but in the meantime, your only option is to try and get close to the optimal voltage for your bulb without flashing it....or try to scrounge up an AWR Hot Driver.

 

[LuxLuthor]

Remember when I said you have to look at SilverFox's battery testing thread? It is listed in the pinned topic in battery section. About 2/3 the way down that NiMH Shootout topic you will see the PowerEx 2500 compared to the Eneloop. The 2700 would not be as good regarding Amp rate as the 2500, so neither are as good as the Eneloops for this purpose. Look at the 5A and 6A curves to see how well the voltage holds up. Also go look at the Elite 1700 AA cells graph at cheapbatterypacks.com

I could spell it out in really easy links and copying all the graphs, but I believe it is useful to teach people how to fish, so they will remember next time.

After you look at the graphs, if you don't understand what they are showing, then let me know here and I will be happy to explain them....but go look at them first.

 

[RichS]

Remember when I said you have to look at SilverFox's battery testing thread? It is listed in the pinned topic in battery section. About 2/3 the way down that NiMH Shootout topic you will see the PowerEx 2500 compared to the Eneloop. The 2700 would not be as good regarding Amp rate as the 2500, so neither are as good as the Eneloops for this purpose. Look at the 5A and 6A curves to see how well the voltage holds up. Also go look at the Elite 1700 AA cells graph at cheapbatterypacks.com

I could spell it out in really easy links and copying all the graphs, but I believe it is useful to teach people how to fish, so they will remember next time.

After you look at the graphs, if you don't understand what they are showing, then let me know here and I will be happy to explain them....but go look at them first.

Thanks for saving me $30 Lux. To be honest, I did look at that battery thread earlier when you pointed me to it, and I did see the Powerex in there not performing so well. But I also saw that the thread was 3 years old. That, combined with the fact that I have seen major modders (NL) mentioning using Powerex in more recent posts made me think that maybe they have done something to improve their batteries in this application. So much for my logic..

It seems I also just made the connection that typically the higher the runtime (mah) a battery has, the worse the amp rate. It seems the best is the 1700 mah Elites, the 2000 mah Eneloops are pretty good too, and the 2700 mah Powerex have poor amp rates. Not sure why this is the case, but it's just another piece in the puzzle for me.

Thanks again Lux! I'll be ordering some Elites (if I can find them in stock somewhere - cheapbatterypacks.com is out) and see if these will tip the scales in favor of my 5761 over the Mag85. If not, then I may have to try the resistance fix on the stock switch.

 

[starburst]

If you really want to use your A123 cell's with the
5761 you will need to hunt down and strangle to
death any resistance you can find with copper,the
tailcap spring and switch fix should help alot. with
battery charge terminated at 3.6 volt's the 5761
should be safe from instaflasing. Not sure how long
the 18650 size will run.

 

[LuxLuthor]

RichS, good job checking that out. Even though some of the SilverFox charts are old, their performance is valid for that battery model. The higher mAh is a supposedly a measure of the total stored energy of the cell.

Most the time, when they put more energy capacity, it is at the expense of not holding voltage under higher Amp loads....which is fine for most consumer electronics (mp3 players, TV remote control, etc.). There are technical explanations (layer thicknesses, type of electrolyte mixtures, internal resistance, etc.) for why this happens....but many manufacturers bank on people not knowing how to find out the real performance of batteries (either the REAL mAh or voltage sag under loads), resulting in advertised mAh rating being falsely inflated.

PowerEx (MaHa) is one of the more highly respected brand names in terms of delivering the mAh capacity that is stated....but as you now have seen, holding voltage in higher mAh cells under load is an intentional tradeoff.

It is rare that you need to get the higher Amp output that the Elite cells provide, so their product is really a "niche" high current application battery....and I would be surprised if they continue making the Elite AA size for much longer. Most RC users have moved to the LiPo packs, and the newer Emoli/A123 Lithium cells.

Starburst, again the A123 cells drop to 3.3V under load immediately....so it doesn't matter if you charge them up to 4 Volts. Two of these cells will ONLY provide 6.6V, and there will be some decrease even if you remove most sources of resistance that you can in the light. Ohm's Law is Ohm's Law, and every part of the light has some resistance unless you direct solder thick wires from cells to bulb.

I would say on a practical basis, two A123 cells will end up delivering more like 6 (or less) volts to the bulb for most setups. Unfortunately, 3 x A123 will be too much voltage for the 5761. This is why 2 x Emoli at 4.15 to 4.2V (hot) is a better choice for this bulb.

 

[Anglepoise]

After reading all about this, I come to the conclusion that we need to get the resistance as low as possible in whatever switch and tail cap spring we use.

And the only way to measure this is with a low ohms meter.

These are very expensive to buy but after a little research I have found what looks to be a viable DIY solution for under $10.00.

I originally tried to build a more sophisticated one, but my electronic skills were not up to the task. This simple one ,powered by a 9 volt battery was easy and works very well indeed.

I finished it last night ,so have not yet started to mod my switch (End cap spring I did last week ) but will take measurements as I go along and see how much resistance I can remove via 'hard wiring' the Mag switch.
See link below for full instructions and pics

http://users.tpg.com.au/pschamb/lom.html

 

[richied]

Lux,

what do you think about these two setups: mag 3c, 2 emoli 26700a's, 5761 bulb and AW 3 stage vs mag4c (extension needed?),3 emoli 26700a's, 1185 bulb and AW 3 stage?
On paper I think the 1185 setup should run longer because of the lower current draw. I think both setups will flash the bulb at full power but am not sure and should be similar in output based on your tables. I am going to aquire some emolis soon to experiment.
Do you know if the emoli's are slightly narrower than the A123s? I had to do a lot of brake hone boring with the 3c and remove cardboard and retape the a123s to get them to fit. I'm not comfortable with the setup, I think the a123's could possible melt the mag body if they should ever short thru the tape!

richD

 

[LuxLuthor]

Anglepoise, thanks for that link. I remember seeing another post recently that I think linked over to RC Groups forum but seeing how involved that looks, sometimes I think of just buying one.

Realistically, you can start out assuming that a stock Maglite Switch has about 200 mOhms, and get close to the ideal voltage/current for a bulb, taking that into account...and get a couple extra bulbs....if you flashed one at a particular hot voltage, you need to either use a NTC to blunt the initial millisec voltage spike and/or use a bit of resistance in your setup.

It is not always best to reduce resistance as you suggest....as the scenario where your battery voltage is close to the flash point of the bulb, lowering the resistance will push it over the edge.

 

[Fulgeo]

Have a Mag85 with FM 9 cell battery holder with Sanyo 2700. I also have a Mag61 with 2 A123 cells. Both using FM bi-pin holders. Have not done resistance mods yet. Just got my Progold today BTW. Anyhow when I bounce the two lights off the ceiling the Mag85 seems to beat the Mag61 but outside on distant targets I might give the edge to the Mag61. The 5761 bulb generates a fatter more usable beam. Everything else is equal on these two flashlights including FM reflectors and such. Considering that my Mag61 is a 2D unlike the 3D Mag85 methinks I will give the nod to the Mag61. Heading outside right now with both flashlights to give this another look.


Your wattage may differ.

 

[LuxLuthor]

Have a Mag85 with FM 9 cell battery holder with Sanyo 2700. I also have a Mag61 with 2 A123 cells. Both using FM bi-pin holders. Have not done resistance mods yet. Just got my Progold today BTW. Anyhow when I bounce the two lights off the ceiling the Mag85 seems to beat the Mag61 but outside on distant targets I might give the edge to the Mag61. The 5761 bulb generates a fatter more usable beam. Everything else is equal on these two flashlights including FM reflectors and such. Considering that my Mag61 is a 2D unlike the 3D Mag85 methinks I will give the nod to the Mag61. Heading outside right now with both flashlights to give this another look.

Your wattage may differ.

Definitely the Philips 5761 is brighter than the 1185 light if overdriven enough. You can see my measurements of the two bulbs at various voltage levels here....the problem with using 2 x A123 cells is you won't get the voltage up high enough to maximize what the 5761 can really do.