mag 85 soft start Q.

[alpg88]

I'm building, actually rebuilding, my mag 85, i bought a bored out 4c tube, and i plan on using 3x26650 in it, before my mag 85 ran on 3x18650, ran great, no soft start, no instaflashes, granted i never used cells hot off charger, and cells were not new, now i'll be using higher capacity\curent cells.

has anyone used this combo?? 3x imr26650 and 1185???

now i understand soft start will fix instaflash, aw soft start is out of stock,
so i was thinking to install one in tailcap, i used inrush current limiter before, i have few of them, but the problem with ntc is when you shut the light off ntc is still hot, and if you turn it on right away again, ntc will offer no protection. (ntc will be hot so will the filament, is hot filament less sensitive to over voltage, or it is easier to burn hot bulb than cold??? and how fast the filament cools down, in 1185 for example)

i,ve heard inductor also works like current limiter, has anyone tried it?? what value i should look for?? any additional components necessary?? i'll have only as much room as tailcap spare bulb cavity offers.

i also found few treads on mosfet soft start, but if i understood correctly i couldn't use it in tailcap, it needs constant power,

 

[hoongern]

has anyone used this combo?? 3x imr26650 and 1185???

I have used 3xAW IMR26500 and 1185 w/ Kiu socket - with the cells only charged to 90% (4.10V)

I now only use it with 1909s.

Granted, I'm not sure if you're using a Kiu socket or the standard Mag socket - which has higher resistance. I also did a tailcap (copper braid) mod for less resistance.

 

[Linger]

with 3xIMR26500's you can only use mode 1 and mode 2 for significant period of time. This is no hinderance mind, as 1185 on medium with 3s fully charged C imr's is white hot and bright.
point being, even with NTC you may poof unless pack voltage bled down.

 

[Bob96]

Much safer to go up to a FM1909 bulb. I even use IMR26650 cells and have not yet flashed one with stock Mag switch or AW Soft-Start.

 

[mrartillery]

Alpg, I built a a NTC for one of my Mag85's (the one with 3 26500's) CPF member Justin Case hooked me up with the resistor I needed and it works great, no I used a Minimag lens with a copper disc on top of it with a short length of wire soldered to the resistor which is then soldered to the tailspring. Justin had a pic of his posted but i cant seem to find it at the time, any way, shoot him a PM and he will tell you which resistor it is you will need.

 

[alpg88]

Alpg, I built a a NTC for one of my Mag85's (the one with 3 26500's) CPF member Justin Case hooked me up with the resistor I needed and it works great, no I used a Minimag lens with a copper disc on top of it with a short length of wire soldered to the resistor which is then soldered to the tailspring. Justin had a pic of his posted but i cant seem to find it at the time, any way, shoot him a PM and he will tell you which resistor it is you will need.

thanks, i think i remember that pic. with square contact

 

[mrartillery]

Yup that's it.

 

[Justin Case]

My posts on using a dropping resistor for a 2D Mag85 running 3xIMR26500 are here. The resistor was an 0.15 ohm, 1%, 3W Vishay through hole resistor from Digikey. Post #17 of the linked thread has the ordering information.

A dropping resistor is not the most elegant solution. It wastes battery power and generates some amount of heat (probably about 1.5W or so for the 1185). But it works and is easy to implement. I"m happy with it. I've thought about other solutions such as MOSFETs, op amps, inductors, and the like, but for me they are just too much effort for what they return. However, if I really needed to eke out the last bit of lumens and run time for some notional application, then I would consider a true soft start solution vice a dropping resistor.

One thing with the 26650s (BatterySpace?) vs AW's IMR26500s. I measured almost 2X the internal resistance for the 26650s vs the IMR26500s using an iCharger 208B (about 55mohms vs about 33mohms). So you might be able to get away with using those bigger cells without instaflashing. The caveat is that I don't know how the iCharger does its internal resistance measurement -- DC, AC, joule heating, etc. So you also may choose to take this internal resistance measurement data with a large grain of salt.

 

[alpg88]

thanks guys, i do have few ntc 1 ohm (cold). i tried to use protected cells with hot wire before, but it would trip protection, after installing ntc problem was solved, partially, it would still trip after a while the light was off, but not if i turned it back on right away, now that i think of it i think i know why, filament cooled off, but ntc didn't so it was like i never installed ntc, but in this case i will install it to save the bulb, not prevent pcb from tripping.

 

[Justin Case]

Besides the hot start issue with an NTC thermistor, there is also the cold ambient condition issue. If you anticipate using your light in say sub-freezing weather, the NTC can have a high initial resistance of 3-4 ohms and the soft start could take longer than at room temp.

 

[alpg88]

update.
i build this little thing and 1185 didn't flash yet,

 

[mrartillery]

Nice, neater job than I done. :duh2:

 

[sterr11]

I cant tell from the pic, but how is the red lead in the spring isolated from the base?

 

[alpg88]

I cant tell from the pic, but how is the red lead in the spring isolated from the base?

by a red insulator

 

[Hack On Wheels]

update.
i build this little thing and 1185 didn't flash yet,

I'm trying to get a sense of scale... does the thermistor sit inside the stock spring, or is that assembly in place of the stock spring? Either way, that looks really nice. I may have to copy you!

 

[sterr11]

by a red insulator

Sorry, I should have explained better, I see the red insulator on the wire, what i am asking about it how you insulated the thermistors red lead from the spring that is soldered to the copper washer?

 

[Roland Gama]

+1 Sterr11.
Looks like red wire is soldered to spring with the thermistor sleeping.

 

[Justin Case]

It looks to me that one leg of the thermistor is soldered to the underside of the round board. The other thermistor leg connects to the red wire, which runs through a hole in the board to a solder point at the top of the spring. The spring is also soldered to the top of the board. As long as the top and bottom of the board are not connected electrically, the thermistor will be in series and all will be well.

 

[sterr11]

It looks to me that one leg of the thermistor is soldered to the underside of the round board. The other thermistor leg connects to the red wire, which runs through a hole in the board to a solder point at the top of the spring. The spring is also soldered to the top of the board. As long as the top and bottom of the board are not connected electrically, the thermistor will be in series and all will be well.

Yes, that's why I am confused, because the board looks like a copper washer in the photo, which would mean the thermistor could not be in series with the filament

 

[alpg88]

It looks to me that one leg of the thermistor is soldered to the underside of the round board. The other thermistor leg connects to the red wire, which runs through a hole in the board to a solder point at the top of the spring. The spring is also soldered to the top of the board. As long as the top and bottom of the board are not connected electrically, the thermistor will be in series and all will be well.

yep, that is how it is.