TR-J18 upgrade options?

116 posts / 0 new
Last post
Tom E
Tom E's picture
Offline
Last seen: 17 min 9 sec ago
Joined: 08/19/2012 - 08:23
Posts: 14625
Location: LI NY

Pokasaha wrote:
Tom E wrote:

Agree - I measured 5,500 lumens with the IOS driver using 3 KK batteries - swapped my mod'ed one for a stock, now ordered another IOS driver - will do the same.

Didn't quite get the last sentence, did you take out the IOS driver and put back the original or what? Did you mean you're also building a dive light by that "will do the same"?

Ooops, sorry. No, I swapped my mod'ed J18 with another BLF member for his stock J18, so now I'm back to the stock J18, which means I'm upgrading to the IOS driver again Smile.

jaws_one
Offline
Last seen: 8 years 4 months ago
Joined: 04/09/2013 - 10:28
Posts: 34
Location: Australia

I just can’t get the driver out… the copper loop on the inside sits so that it won’t fall out no matter how I try to manipulate it… is there a trick?

Pokasaha
Offline
Last seen: 8 years 1 week ago
Joined: 11/22/2012 - 12:04
Posts: 29
Location: Finland

I’ll put this here also

jaws_one wrote:
Hi Guys,

My TR J18 has suffered the same fate. After about 4 months of use it has dimmed noticeably. I’ve only run 18650’s in it and it’s had a gentle life. Something wrong with that board huh?

I have managed to pop out the driver board but there is that wound copper wire thing on the inside stopping it from coming all the way out out. Doesn’t seem like it will come out without breaking it. Thinking they wired in the driver board first then glued in the reflector? It’s so annoying.

Any ideas?

James

Didn’t understand that wound copper wire thing, mine had nothing else keeping the driver than the power wires. The reflector is not glued, it’s attached with two screws which you can loose from the inside after moving the driver away. The leds are glued to the frame. Silly way since you can’t take out a “drop in”..

Few days ago I figured out that the frame is separateable to two pieces, if one would like to do some additional plotting…

Tom E
Tom E's picture
Offline
Last seen: 17 min 9 sec ago
Joined: 08/19/2012 - 08:23
Posts: 14625
Location: LI NY

jaws - didn't have that problem with mine - driver came right out. Got the pic in the other J18 thread to show it dangling. Not sure what you can do...

Pokasaha
Offline
Last seen: 8 years 1 week ago
Joined: 11/22/2012 - 12:04
Posts: 29
Location: Finland

A pic would help a bit? Maybe jammed threads if you mean the copper ring that keeps the spring against the driver?

Dexter Tang
Offline
Last seen: 8 years 4 months ago
Joined: 03/02/2013 - 14:46
Posts: 7
Location: Jurong West, Singapore

Is there a way to mod in a variable control switch in the tr j18 and if so, where to get one that fits that monster?

jaws_one
Offline
Last seen: 8 years 4 months ago
Joined: 04/09/2013 - 10:28
Posts: 34
Location: Australia

Managed to get it out – now to determine the fault. It took a lot of force to remove. The copper coil on the board had bent out a little so I needed to pry with a lot of force on an angle to remove it. Pushed the copper coil back down now it slides in and out easily. Still works too, albiet the same at only up to about 50% power…

Pokasaha wrote:
A pic would help a bit? Maybe jammed threads if you mean the copper ring that keeps the spring against the driver?
jaws_one
Offline
Last seen: 8 years 4 months ago
Joined: 04/09/2013 - 10:28
Posts: 34
Location: Australia
Tom E
Tom E's picture
Offline
Last seen: 17 min 9 sec ago
Joined: 08/19/2012 - 08:23
Posts: 14625
Location: LI NY

Dexter Tang wrote:
Is there a way to mod in a variable control switch in the tr j18 and if so, where to get one that fits that monster?

Anything is possible, but no simple way for sure, that sounds like some decent amount of engineering (EE/ME) work... 

Phirebird
Offline
Last seen: 8 years 4 months ago
Joined: 04/30/2012 - 22:11
Posts: 13
Location: Sydney, Australia

I looked around for good tailcap switches after I blew the original after attempting a resistor mod on my stock driver.
First I added a resister in parallel to the current sensing resistor to get to 50ohm from 150, then I added some extra mini cylindrical magnets between all the nano lithium A123 cells I use and the tail cap until the tailcap blew from the extra power!

I tried replacing it with a larger switch from a broken vacuum cleaner however it wouldn’t fit in the tailcap.
I thought all hope was lost until I was fiddling with one of those budget 5 dollar bulb torches that use those 6v super heavy duty alkaline batteries – I noticed a perfectly sized switch for the button. A quick swap and it works!

Also, put the vacuum cleaner switch back in the 5dollar cheapo light and modded a spent 6v alkaline battery to store a 2s2p 18650 lithium battery pack with balance charger connector Big Smile and now that little beast is always ready to shine since those alkaline batteries are notorious for dieing horribly when your not watching.

Phirebird
Offline
Last seen: 8 years 4 months ago
Joined: 04/30/2012 - 22:11
Posts: 13
Location: Sydney, Australia

The resistor in question is the flat one directly adjacent the large yellow component on the circular pcb in this image:

http://i.imgur.com/orXD3yB.jpg

It is quite noticeable thanks to its wider and flatter design. Mine measured in at 150 ohms. I did a parallel resistor calculation and found a 75 ohm resistor added in parallel would give me the 50 stated in the OP. I wanted something easy to remove if things went wrong (such as flashing/flickering/low beam/etc) so I went with parallel resistors and Surface Mount Component Fix from Fixit products.

I found a 75 ohm resistor using a multimeter and an old graphics card board and just popped it off with one of the multimeter probes (surface mounted) and prepped the current sensing resistor using the Surface Mount Component Fix. It is basically a conductive glue like wire glue intended from surface mounted components. I applied some component fix to either half of the original resistor as the 75 ohm resistor was significantly smaller than the original and made sure to cover the two tinned ends of the original.

After letting it dry I applied the 75 ohm resistor directly above the 150 ohm resistor and put a generous coating of surface mount on either end. I checked with a magnifying lens and my T6 torch to ensure there wasn’t any shorting with nearby components, scraping off excess component fix with a pin.

Initially I thought I had failed as the flashlight was very dim after a short burst and it took me a while to realize that the batteries were drained (3.3v across 3×3.3v batteries!) and so I charged them up (ah the beauty of LiFe nano lithium) and it blazed away for a good few seconds before the tail cap blew. I used some angled long-neck tweezers to open the tailcap (alternatively use the multimeter probes) and it smelt acrid. The rest was pretty easy once I had the correct switch, just heat your soldering iron for a good few minutes and apply to the circular tail cap board on whatever solder you can to remove the old and replace with the new.

I had a few issues with the torch not lighting the first few times due to a poor connection, so I re-coated the circular board with extra solder wherever there was exposed trace to increase conductivity and reduce heat. This thing blazes!

gr8mikey
Offline
Last seen: 8 years 4 months ago
Joined: 05/18/2013 - 20:43
Posts: 2
Location: Alabama, USA

I recently won an ebay auction for one of these flashlights for $43. I am so looking forward to trying this mod. I have ordered the I/O driver board and have already replaced the tailcap switch with the 14V 10A Judco switch that was mentioned earlier in this thread. In order to make the Judco switch fit, you have to pry it apart and remove the copper components from it so that you can cut the switch housing down to a square. Then put it back together and bend to copper leads so that they will fit. Once you do that, you can file/sand the original switch keeper (the round plastic piece with the rectangular cutout) so that it will fit around the Judco switch. Then just solder it to the pcb that the old switch was mounted to and then beef up the current carrying ability of the spring by soldering a piece of braided copper wire.

As far as batteries, I had a 36V dewalt power tool battery that had a bad cell. So I disassembled the pack and tested the cells individually. I ended up with 9/10 of the cells being good, measuring 3.2 to 3.3V each. These are the A123 Systems 26650 Lithium Iron Nanophosphate cells that the poster above is using. Topoff charge on these is 3.6V. The trustfire TR006 charger has a switch on it for 4.2V and 3.0V. I think the 3V is in error because it does in fact bring these cells to 3.6 to 3.65V on this setting. Too bad these cells are only 2300mah, but I guess that is the tradeoff for being able to sustain 70 amps per cell.

So far the light is working great and is very impressive with the stock driver. I guess I need to get a lux meter so that I can do some before and after measurements to see how much increased output I get from the new driver. Any good recommendations on a decent lux meter?

Phirebird
Offline
Last seen: 8 years 4 months ago
Joined: 04/30/2012 - 22:11
Posts: 13
Location: Sydney, Australia

I am sat here with my tr-j18, and those nano phosphate cells sitting comfortably inside, and I can’t help but be irritated to the fact that those cells drain so quickly at high mode.

Now I generally use the torch on low mode and love the excellent feature of being able to remember the previous setting, and even the fast charge rate of 3.5-4.5A, however I think a torch or mod needs to be tailored specifically to these cells.

What I was thinking would be to have from dual to quad cells in parallel such as in the sky ray king and nitecore tm26, or a mod that uses a DC step down module to power a tr-j18 from a 33v battery pack.

gr8mikey
Offline
Last seen: 8 years 4 months ago
Joined: 05/18/2013 - 20:43
Posts: 2
Location: Alabama, USA

Are there any issues with using nickel/copper plated neodymium magnets as spacers? I just did a test run by leaving the light on high until the heatsink was uncomfortably hot and immediately pulled the batteries out. The batteries and magnetic spacers were barely even warm. So this shouldn’t be an issue as far as I can tell. If you have heard of any negative consequences of doing this, please let me know.

Phirebird
Offline
Last seen: 8 years 4 months ago
Joined: 04/30/2012 - 22:11
Posts: 13
Location: Sydney, Australia
gr8mikey wrote:
Are there any issues with using nickel/copper plated neodymium magnets as spacers? I just did a test run by leaving the light on high until the heatsink was uncomfortably hot and immediately pulled the batteries out. The batteries and magnetic spacers were barely even warm. So this shouldn’t be an issue as far as I can tell. If you have heard of any negative consequences of doing this, please let me know.

I fried one nickel plated neodymium disc out of 100 in 12 months. they were the 3mm*3mm cylinder type, way too small to be used as spacers for the 26650 a123s individually when modded so I now use three in parallel between each cell for better contact as they do have a bit of resistance. I may electroplate some copper onto them to see if that can’t be improved.

zeremefico
zeremefico's picture
Offline
Last seen: 1 hour 48 min ago
Joined: 03/27/2012 - 02:44
Posts: 1389
Location: Greece

Hi, got today my TR-J18 from lightmalls, along with the well known upgrade driver.
Before I start the driver change, I measured the old driver (TR-169A), 5 amps tailcap and 12 volts from driver to leds. They are in series.
This means 12/7=???
The new driver shows 23 volts, as it should be.
Can anyone solve the mystery?

₪₪₪₪ ΟΥΔΕΝ ΚΡΥΠΤΟΝ ΥΠΟ ΤΟΝ ΗΛΙΟ ₪₪₪₪

My YouTube channel

Flashlights & edc gear

K40M F16

Phirebird
Offline
Last seen: 8 years 4 months ago
Joined: 04/30/2012 - 22:11
Posts: 13
Location: Sydney, Australia
zeremefico wrote:
Hi, got today my TR-J18 from lightmalls, along with the well known upgrade driver. Before I start the driver change, I measured the old driver (TR-169A), 5 amps tailcap and 12 volts from driver to leds. They are in series. This means 12/7=??? The new driver shows 23 volts, as it should be. Can anyone solve the mystery?

Maybe it is inefficient, made of lower rated components or designed to output lower voltage at higher current (i.e. giving the same power with different voltage/current values – you should place your ammeter in series with the LEDS to see).

Remember; P=V*I and so you were pulling 60W or so from the cells you were using (3x Li-ion cells at 5A).

Nightbird95
Nightbird95's picture
Offline
Last seen: 1 year 7 months ago
Joined: 01/03/2013 - 01:16
Posts: 1153
Location: N 11º 08', E 124º 37'

Jonney_boy wrote:
Ok, one more update. The light seems to be pulling less and less amps. On 2 cells I am now only able to pull 2.3 – 2.4 amps and on 3 cells, about 2 amps. Seeing how all the led’s seem to be working, wiring seems to be “in tact”.. and batteries are able to (at least for a short while) pull 6A without the protection kicking in.. I’m thinking it has to be the driver..

so… one last try…. I orderd one more driver from IO this time (the last one, which looked the same was the customer programed one from KD.. which looks identical to the IO one).

fingers crossed. I will let you guys know when I get the new driver delivered and installed Smile

thanks.

Any update on your new driver Jonney-boy? I’m interested to know what the problem is as I’m also looking forward to do mods too. :~

Jonney_boy
Offline
Last seen: 5 years 8 months ago
Joined: 12/13/2012 - 15:23
Posts: 38
Location: Canada

ok.. one more update…… I messed up the order (my fault) and shipping took a LOT longer than expected….. anyhow, new driver came in today and..

the KD driver I ordered was a dud… I re-ordered a IO driver (looks the SAME as the KD driver) and the light works.

I didn’t get a chance to put a meter on it yet.. but it is bright… crazy crazy bright.. wow!!

The light does get very hot on high, I mean I had it on high for about a min and it was fairly warm to the touch.

I don’t think I can do too many wall shots of it either, the flood is so wide, and it has so much power that it washes out all my other lights.. haha…..

This light was a bit more trouble and expense than expected.. but it’s soooo worth it… if anything is considering this mod… do it.. (or go the cheaper route and mod the j12).

In summary….

Ordered J18, full set (4) of TF flames 18650…. one marginal cell that would trip protection early and shut down the light… eliminated the flaky cell and light worked great… for a week…….

Factory driver then blew… light was very dim.. it would work.. but would be very dim.. (driver toast?)

Replaced with KD driver… no luck… light would not turn on.. only a quick flash… (batteries not good enough? tripping protection?)

Ordered full set (4) TF flames 26650 + xstar charger for the new bigger batteries… (no luck.. DOH….)

Re-installed factory (blown) driver… same as above.. really dim….

Modded factory driver (resistor mod…put a pot in it).. light was better.. but only 20-30% of factory….

Finally replaced driver again with IO driver…. bingo… it works Smile : )

somewhere in there I also upgraded the contacts a bit… Driver spring was doubled up (I just added a second spring and wound them together…. tail cap has copper braid in it.

Nightbird95
Nightbird95's picture
Offline
Last seen: 1 year 7 months ago
Joined: 01/03/2013 - 01:16
Posts: 1153
Location: N 11º 08', E 124º 37'

Oh I see! Shocked

So I guess I’ll have to stay away from KD then if I decide to do the mod.

Thanx for the update Jonney-boy! Smile

Jonney_boy
Offline
Last seen: 5 years 8 months ago
Joined: 12/13/2012 - 15:23
Posts: 38
Location: Canada

Nightbird,

I would not say to stay away from KD. A lot of people have used their driver and it has worked very well. Also, they do have a “custom programed” version of the driver that is very nice.

There are a bunch of reasons why I would have gotten a bad driver. Including me messing up the driver trying to solder on the B+ spring the first time (I initially used a cheap iron that didn’t tin properly). It is also possible that I had received a defective driver, or a wrong driver all together.

Depending on what you want to do with the light, I would say that the upgrade is worth it. The 30% mode is similar to what the stock light puts out (usable and “enough” light for most people’ needs). The 5% mode is similar to what my cheap single T6 light puts out (ultrafire KFt60). At full 100% output there is just a LOT of light… and a lot of heat. The head will get very warm in just over a min. I would not want to run this mode for any real length of time…. but it’s great for the “ooo.. wow” factor.. and for times where you just want to check something (walking in the park @ night.. hear a noise…. turn around, flip on flashlight and light up the ENTIRE parking lot + field behind you).

Few other notes. with the IO driver, I don’t see any pwm flicker (at least to my eyes). But I can hear it… loud and clear….. This may annoy some people more than others.

Besides the driver, the springs inside will all need to be upgraded too, as now you are pulling LOTS more current than stock. And even in stock mode, the light is at the limits of the springs inside. I ended up doubling up the B+ spring and adding copper braid to the tail cap spring. Batteries do make a diff also. I tried with both 18650 TF flames and 26650 TF flames. The light works with both but is a fair bit brighter with the 26650’s, telling me the 18650’s just arn’t discharging fast enough.

Finally, I have yet to find a easy drop in switch that is cheap and rated for the current. The stock switch while works, is not rated for the current we are putting through it. It will die given time.

Tom E
Tom E's picture
Offline
Last seen: 17 min 9 sec ago
Joined: 08/19/2012 - 08:23
Posts: 14625
Location: LI NY

Nice to hear it finally worked out for you. I was getting 5,000 lumens at 30 secs with the TF 26650's. With KK's it does 5,500. Was thinking of ordering the KeepPower 26650's which use KK cells from doingoutdoor.com, but $60 for 2 pair, yikes - more than what the light cost. Wish KumaBear was around because those protected KK's w/raised top for $10.48 is a great deal, for 3 maybe about $36 w/shipping, and I think those were the batteries that delivered the 5,500 lumens.

Nightbird95
Nightbird95's picture
Offline
Last seen: 1 year 7 months ago
Joined: 01/03/2013 - 01:16
Posts: 1153
Location: N 11º 08', E 124º 37'

Jonney_boy wrote:
Nightbird,

I would not say to stay away from KD. A lot of people have used their driver and it has worked very well. Also, they do have a “custom programed” version of the driver that is very nice.

Yeah you’re right. Sometimes I get items with defects (which I can repair) from FastTech but still I keep buying from them. Smile

Pokasaha
Offline
Last seen: 8 years 1 week ago
Joined: 11/22/2012 - 12:04
Posts: 29
Location: Finland

I was so close to finish my diving light with these components, but seems that another IO driver broke up (!!). One already burned two of the xm-ls and then just didn’t do anything anymore. I ordered another IO driver and new leds (Q2’s this time) and put in one piece. Got it working nicely. Today I tried to do a battery run time test and suddenly the light stopped working. I took it apart and saw that the two metal rod holding the drivers plates together had snapped apart. I welded them together but the lamp doesn’t work still…

One question. What should the voltage be in the supply side if theres no connection? With mine it’s few millivolts..

some pics..
https://www.dropbox.com/sh/w1go0n30u40gw0d/9sK-j7jMjv

Pokasaha
Offline
Last seen: 8 years 1 week ago
Joined: 11/22/2012 - 12:04
Posts: 29
Location: Finland

Hahaa, got a new driver (third one) from IO for free since there was obvious manufacturing defect. Now got it working and man this torch is bright! I got it underwater and it was a piece of sun there. :bigsmile: Crown Only minus is the spill that tends to reflect a bit too much from the small particles in the water.

Pages