drew2001
Newly Enlightened
All in favor of the Mag623, give me a "Harumph" ......
....... Hey I didn't get a Harumph out of You !
(uh, part borrowed from the movie Blazing Saddles)
I am happy to report that my recently purchased JimmyM's JM-SST circuit board has been installed in my Mag623. I present a simple but rather complete tutorial on an assembly procedure of the Soft_Start_Trim. This board is made to fit inside a Kiu socket adapter's aluminum base, and is now fully wired to operate high amperage switching duty within a Maglite switch housing. It features an adjustable time elapse softstart electronic switch with a variable pulse width modulation (PWM Duty Cycle) voltage control.
The adjustment is accomplished through turning the trimpot's screw. Turning it an anti-clockwise direction from it's full right stop point, the trim pot allows rising rate adjustments of voltage from a 50% DC and slow-start upwards to the 100% DC fast-start maximum voltage when fully turned. A slower-start up is also available (at 100% DC) by adjusting it back from full left aproximately a 1/3 of a turn.
I'll just say it took me several hours to complete the installation. I still managed to finish in one long evening.
Hopefully this guide helps to cut the time down a bit for readers planning on tackling an installation.
My subject hotwire starts off with perhaps a year old MaxBlaster host. To refresh, it's a tri-bored out 3 D cell Maglite body, using a MOP (I think it's one of FiveMega's) aluminum reflector, and a borofloat window with the Osram 64623 100w bulb. Made famous for its higher voltage overdriven "sunlight" lighting capablilty. Casting a retina burning 3000-4000+ lumens out a torch window.
The typical run time is short, hitting high power demands hard for maybe a conservative 6-7 minutes total run time, that's before an obvious light output sag/dimming. Using a 15.6v pack of 13 x 2/3a 1500mah NiMH cells and two dummy cells make the pack a symetrical flat ended cylinder at 6-7/8" long.
Since it really burns power fast, these lights should be planned for short periods of intermittant useage.
When time comes for batt packs.... I highly recommend getting ahold of LuxLuthor for suberb cell pack fabrication needs.
The addition of Jimmy's circuit board into mag623 and other hi-performance hotwire flashlights minimizes wear on mechanical switch contacts otherwise stressed from constant high wattage and 10+ amp operating conditions. As well as allowing a softstart option protection for the bulb when using a slightly higher start voltage and first hitting the switch.
Picture of my MaxBlaster and Torch.
CPF Forum member cljn3, recently posted a good hint worth sharing. Jimmy's sale thread mentions an external trimpot access hole. Done by drilling a hole into the Kiu adapter. By cutting a pattern template to fit inside the round adapter and then transferring the hole locations using an old mag plastic lens for the template material. A clear trace helps one "see" the hole location to transfer the holes to the inside of the adapter. I happen to use a piece of Zip-Loc baggy for the template, it worked ok.
Be sure to clean out ( I brushed it )... all the aluminum dust out of the newly drilled adapter.
Check the fit of the circuit board to Kiu's adapter for alignment with the predrilled mount holes. As I continued assembly I found that I preferred to have the board's mounting through-holes cut open at the outsides to allow for slight shifting for screw alignment later when screwing it pinched together on top of the mag switch's housing.
Before I attempted this modification, I determined where I wanted the bulb and socket to seat, height-wise in the mag tube. I also knew that I preferred the Mag head turning adjustments to be nearly bottomed out. Screwing the Mag head to almost full tight. I can back it off a few turns to adjust for the best beam, a nice clean throw to spill mix. Other height adjustments found more noticeable artifacts in the beam with the MOP. A better overall smoother beam reflector choice may be a heavy stipple, or a Heavy Orange Peel (HOP) like I believe the "Torch" has.
Due to the above adjustability wants for my reflector and this 64623 bulb, I did not use stand off spacers between the ceramic socket to aluminum base. I also used a round file to open up the aluminum base's center hole into a wider arc just in-between the screw holes for the flush mounted socket's wires to flow direct and straight through the socket's center hole. One can use a cylindrical dremel cutter bit, a wide flute preferred made for soft metals like "flute clogging" aluminum.
Think it all through, be patient and keep yourself steady handed, it's not too difficult a project to complete cleanly.
Next, the wiring.
Jimmy's board diagram :
I cut ~3/4" pieces of heat shrink tubing of 1/8" id size, sliding them over the socket's wires to fit right up against the ceramic base. To further protect against excess heat insulation damage. While the HS tubing is still hot you can prebend the wires to fit inside the base in cooled precurved shapes. Later it's going to be an exacting tight fit inside the base as they will be laying next to hot aluminum and board components.
Reason for the heat shrink tubing. Upon taking apart the MaxBlaster I found that it's socket wiring had ended up heat cracked. Creating spiderweb like splits in the red rubber insulation right where the wires had fed through the aluminum base. I felt that I wanted to try to avoid a potential high amperage heat related direct short later on.
First advice prior to begin soldering, be sure the kids are tucked in and you are not disturbed, easy on the coffee and make sure you have a clean tip on your solder gun or iron. I actually used my old Weller 100/140 watt gun on "low" (is 100w low?
) ...for short bursts creating a fast heat soak on the pre-tinned heavy gauge wires.
This quick heat application yeilds a glossy wet looking solder puddle at point of contact using pre-tinned wire ends, and is a good sign of a through saturation and if done carefully less heat is transfered to nearby components.
I used a 24 gauge wire for the B+ switching task, from a coil of radio shack "hook up and lead" wire as there is just low ua current for the electronic switching task in Jimmy's board design. A very low signal current passing through makes the mag button switch last a long life.
Below are pictures after the tower was cut off. Pics of the trimming, done using a mini-fluted triangular shaped dremel bit and the drilling of the mag switch housing, planning for the route and fitment of wires through the plastic housing. Hold the board and housing right next to each other to figure for alignment of the holes needed for internal wire routes.
I was not sure what drill size to recommend, (my maxblaster light's magswitch had those holes already in it) but fellow forumer DonShock wrote that he uses a 3/32" bit for the 2 screw hole sizes which then are tackling the self-threading by carefull alignment and screwing them firmly and straight into the black base's drilled holes. The wire's holes can use a 1/8" drill bit that obvously fits into a dremel's 1/8th tool collet. The planned board-to-switch housing will end up a flush fit later on when finally screwed down.
Note that the housing plastic is a dry semi-hard plastic almost like an electrical outlet bakelite composition. This creates a lot of dusty crumbs when grinding, just a heads up for the messy dust to come.
Then thread the wires through, pulling base and board together and screw it to the housing. This provides for the correct fit lenght of the power wires, before cutting them and soldering takes place.
Below pic shows when I found that the board having been first trimmed to open the mount holes in U shaped cut outs realized an easier alignment. (pic showing it almost snug not tight)
Solder point of the mag's blue inner sw + wire...
Note Check the direction the blue switch fit into the housing as it only slides back in and seats correctly through one side of the black housing. Be sure to route the 24g sw+ wire out the correct side housing opening to be able to solder it to the switch tab and slide the switch back in easy. Then the wire can be tucked nestled inside the housing.
Threaded wire routes in the below pic.
I used a solid copper carpet tack and cut the shafts tip off. Next took my leatherman pliers and kinda "screwed" it into the coil spring.
I left a little gap between the head of the tack and the coil spring to be able to wrap the threaded positive wire to loop around to a snug fit around the copper nail. It's then easy to solder the wrapped positive wire and clip the extra off flush making it a secure springy copper positive contact for the batt pack's positive end. The spring base's contact pad remains as the positive contact for the switch.
The soldering of the negative wire to the grub nut/plate was a bit of an experiment. It worked out well and came out clean.
Note: put the GRUB screw pointed end out, BACK into it's hole now...
For the negative wire connection I took the thin metal strip of the nut and bent a crimp-able section near the nut and using this "crimp" as the method to hold the pre-tinned neg wire end in place. Then soldered the neg wire onto the grub nut strip. First check how the neg wire lenght fits before the soldering.
I assembled it putting the strip's side out and pushed the nut back into it's locating plastic keyhole. Then turned the allen wrench to start the threads of the grub screw, in preparation for later installation of the completed switch to Mag tube.
All done.
It sure makes for a nice clean soft start variable duty cycle output low current switching assembly ready for any number of Mag hosts well into the future.
Final fitment of the switch assembly back into the mag body seemed to be easier if you leave the base securing screws a couple of turns backed off so that the base to swith housing can be wiggled a bit thus creating room to slide it in. It made it a bit smoother to push it down into the tube. You've got to remove the circlip inside the mag body - if one is stll in there - and go push it in from the top head end.
Then you can re-snug the holder screws back just snug. Realize that the Mag's plastic base is not metal and so the screw's holes made of plastic can easily be stripped out if you try to tighten too forcefully. It's not going anywhere so plan for error on the light side of torque applied. After alignment by eye of the Mag's button hole to Mag switch take the 5/64 or 2 mm allen and snug the grub screw down and just put the rubber button cover back on.
The rest of the adjustment choices are yours. I figured opon just finishing it up, to start out with the trim pot turned to about a quarter turn back away from a full left anti-clockwise (use easy light pressure on the pot) a very thin bladed long jewelers screwdriver worked well. This position gave a semi-fast start at about (.3) second to saturate the filament slower and yet have 100% duty cycle at the max voltage output.
FOLLOW UP report:
I'm using stock tail spring for neg contact, I've since wrapped a piece of copper wire from top spring loop to the base loop. I charged the pack, let it rest until it was at 19.00 volts inserted it and hit the switch, the bulb blew after a quick 2 second bright light. I then let the batt pack rest a little bit more until it showed 18.80 volts and also turned the adjustment screw clockwise a bit more making it about a 1/3 turn from the full left stop in the pot. The bulb that I used for replacement had at least an hour of prior overdriven ~14 v Torch use.
This time it did not poof and the v. sag happened perhaps in the short time it's heating up to allow a somewhat rested pack to be used. LuxLuthor's "Destructive Bulb Tests" results, has reports indicating that the 623 bulb's max voltage (until it popped) was found through applying an accurate volt measure from bench power supply source in incremental increases up to ~16.9 volts. So now my having set the trim adjustment at this slight additional turn increment, perhaps a 18.5 volts open circuit vbat will not flash. I'm taking into consideration that the vbulb measurement will also show a drop down to/or a little under 16.5 volts. (my power is being supplied by this particular NiMH battery pack) ... This 18.5v, as my light is set up, could be considered a safe slow starting level at 100% Duty Cycle voltage. Your mileage may vary, due to taking into consideration any slight additional resistance fixes (tail spring ohms tests)... So go try and enjoy testing!
It has that interesting PWM buzz, definately louder when soft start is just loading up power.
Thanks for reading ... Blast-away!
....... Hey I didn't get a Harumph out of You !
(uh, part borrowed from the movie Blazing Saddles)
I am happy to report that my recently purchased JimmyM's JM-SST circuit board has been installed in my Mag623. I present a simple but rather complete tutorial on an assembly procedure of the Soft_Start_Trim. This board is made to fit inside a Kiu socket adapter's aluminum base, and is now fully wired to operate high amperage switching duty within a Maglite switch housing. It features an adjustable time elapse softstart electronic switch with a variable pulse width modulation (PWM Duty Cycle) voltage control.
The adjustment is accomplished through turning the trimpot's screw. Turning it an anti-clockwise direction from it's full right stop point, the trim pot allows rising rate adjustments of voltage from a 50% DC and slow-start upwards to the 100% DC fast-start maximum voltage when fully turned. A slower-start up is also available (at 100% DC) by adjusting it back from full left aproximately a 1/3 of a turn.
I'll just say it took me several hours to complete the installation. I still managed to finish in one long evening.
Hopefully this guide helps to cut the time down a bit for readers planning on tackling an installation.
My subject hotwire starts off with perhaps a year old MaxBlaster host. To refresh, it's a tri-bored out 3 D cell Maglite body, using a MOP (I think it's one of FiveMega's) aluminum reflector, and a borofloat window with the Osram 64623 100w bulb. Made famous for its higher voltage overdriven "sunlight" lighting capablilty. Casting a retina burning 3000-4000+ lumens out a torch window.
The typical run time is short, hitting high power demands hard for maybe a conservative 6-7 minutes total run time, that's before an obvious light output sag/dimming. Using a 15.6v pack of 13 x 2/3a 1500mah NiMH cells and two dummy cells make the pack a symetrical flat ended cylinder at 6-7/8" long.
Since it really burns power fast, these lights should be planned for short periods of intermittant useage.
When time comes for batt packs.... I highly recommend getting ahold of LuxLuthor for suberb cell pack fabrication needs.
The addition of Jimmy's circuit board into mag623 and other hi-performance hotwire flashlights minimizes wear on mechanical switch contacts otherwise stressed from constant high wattage and 10+ amp operating conditions. As well as allowing a softstart option protection for the bulb when using a slightly higher start voltage and first hitting the switch.
Picture of my MaxBlaster and Torch.
CPF Forum member cljn3, recently posted a good hint worth sharing. Jimmy's sale thread mentions an external trimpot access hole. Done by drilling a hole into the Kiu adapter. By cutting a pattern template to fit inside the round adapter and then transferring the hole locations using an old mag plastic lens for the template material. A clear trace helps one "see" the hole location to transfer the holes to the inside of the adapter. I happen to use a piece of Zip-Loc baggy for the template, it worked ok.
Be sure to clean out ( I brushed it )... all the aluminum dust out of the newly drilled adapter.
Check the fit of the circuit board to Kiu's adapter for alignment with the predrilled mount holes. As I continued assembly I found that I preferred to have the board's mounting through-holes cut open at the outsides to allow for slight shifting for screw alignment later when screwing it pinched together on top of the mag switch's housing.
Before I attempted this modification, I determined where I wanted the bulb and socket to seat, height-wise in the mag tube. I also knew that I preferred the Mag head turning adjustments to be nearly bottomed out. Screwing the Mag head to almost full tight. I can back it off a few turns to adjust for the best beam, a nice clean throw to spill mix. Other height adjustments found more noticeable artifacts in the beam with the MOP. A better overall smoother beam reflector choice may be a heavy stipple, or a Heavy Orange Peel (HOP) like I believe the "Torch" has.
Due to the above adjustability wants for my reflector and this 64623 bulb, I did not use stand off spacers between the ceramic socket to aluminum base. I also used a round file to open up the aluminum base's center hole into a wider arc just in-between the screw holes for the flush mounted socket's wires to flow direct and straight through the socket's center hole. One can use a cylindrical dremel cutter bit, a wide flute preferred made for soft metals like "flute clogging" aluminum.
Think it all through, be patient and keep yourself steady handed, it's not too difficult a project to complete cleanly.
Next, the wiring.
Jimmy's board diagram :
I cut ~3/4" pieces of heat shrink tubing of 1/8" id size, sliding them over the socket's wires to fit right up against the ceramic base. To further protect against excess heat insulation damage. While the HS tubing is still hot you can prebend the wires to fit inside the base in cooled precurved shapes. Later it's going to be an exacting tight fit inside the base as they will be laying next to hot aluminum and board components.
Reason for the heat shrink tubing. Upon taking apart the MaxBlaster I found that it's socket wiring had ended up heat cracked. Creating spiderweb like splits in the red rubber insulation right where the wires had fed through the aluminum base. I felt that I wanted to try to avoid a potential high amperage heat related direct short later on.
First advice prior to begin soldering, be sure the kids are tucked in and you are not disturbed, easy on the coffee and make sure you have a clean tip on your solder gun or iron. I actually used my old Weller 100/140 watt gun on "low" (is 100w low?
) ...for short bursts creating a fast heat soak on the pre-tinned heavy gauge wires. This quick heat application yeilds a glossy wet looking solder puddle at point of contact using pre-tinned wire ends, and is a good sign of a through saturation and if done carefully less heat is transfered to nearby components.
I used a 24 gauge wire for the B+ switching task, from a coil of radio shack "hook up and lead" wire as there is just low ua current for the electronic switching task in Jimmy's board design. A very low signal current passing through makes the mag button switch last a long life.
Below are pictures after the tower was cut off. Pics of the trimming, done using a mini-fluted triangular shaped dremel bit and the drilling of the mag switch housing, planning for the route and fitment of wires through the plastic housing. Hold the board and housing right next to each other to figure for alignment of the holes needed for internal wire routes.
I was not sure what drill size to recommend, (my maxblaster light's magswitch had those holes already in it) but fellow forumer DonShock wrote that he uses a 3/32" bit for the 2 screw hole sizes which then are tackling the self-threading by carefull alignment and screwing them firmly and straight into the black base's drilled holes. The wire's holes can use a 1/8" drill bit that obvously fits into a dremel's 1/8th tool collet. The planned board-to-switch housing will end up a flush fit later on when finally screwed down.
Note that the housing plastic is a dry semi-hard plastic almost like an electrical outlet bakelite composition. This creates a lot of dusty crumbs when grinding, just a heads up for the messy dust to come.
Then thread the wires through, pulling base and board together and screw it to the housing. This provides for the correct fit lenght of the power wires, before cutting them and soldering takes place.
Below pic shows when I found that the board having been first trimmed to open the mount holes in U shaped cut outs realized an easier alignment. (pic showing it almost snug not tight)
Solder point of the mag's blue inner sw + wire...
Note Check the direction the blue switch fit into the housing as it only slides back in and seats correctly through one side of the black housing. Be sure to route the 24g sw+ wire out the correct side housing opening to be able to solder it to the switch tab and slide the switch back in easy. Then the wire can be tucked nestled inside the housing.
Threaded wire routes in the below pic.
I used a solid copper carpet tack and cut the shafts tip off. Next took my leatherman pliers and kinda "screwed" it into the coil spring.
I left a little gap between the head of the tack and the coil spring to be able to wrap the threaded positive wire to loop around to a snug fit around the copper nail. It's then easy to solder the wrapped positive wire and clip the extra off flush making it a secure springy copper positive contact for the batt pack's positive end. The spring base's contact pad remains as the positive contact for the switch.
The soldering of the negative wire to the grub nut/plate was a bit of an experiment. It worked out well and came out clean.
Note: put the GRUB screw pointed end out, BACK into it's hole now...
For the negative wire connection I took the thin metal strip of the nut and bent a crimp-able section near the nut and using this "crimp" as the method to hold the pre-tinned neg wire end in place. Then soldered the neg wire onto the grub nut strip. First check how the neg wire lenght fits before the soldering.
I assembled it putting the strip's side out and pushed the nut back into it's locating plastic keyhole. Then turned the allen wrench to start the threads of the grub screw, in preparation for later installation of the completed switch to Mag tube.
All done.
It sure makes for a nice clean soft start variable duty cycle output low current switching assembly ready for any number of Mag hosts well into the future.
Final fitment of the switch assembly back into the mag body seemed to be easier if you leave the base securing screws a couple of turns backed off so that the base to swith housing can be wiggled a bit thus creating room to slide it in. It made it a bit smoother to push it down into the tube. You've got to remove the circlip inside the mag body - if one is stll in there - and go push it in from the top head end.
Then you can re-snug the holder screws back just snug. Realize that the Mag's plastic base is not metal and so the screw's holes made of plastic can easily be stripped out if you try to tighten too forcefully. It's not going anywhere so plan for error on the light side of torque applied. After alignment by eye of the Mag's button hole to Mag switch take the 5/64 or 2 mm allen and snug the grub screw down and just put the rubber button cover back on.
The rest of the adjustment choices are yours. I figured opon just finishing it up, to start out with the trim pot turned to about a quarter turn back away from a full left anti-clockwise (use easy light pressure on the pot) a very thin bladed long jewelers screwdriver worked well. This position gave a semi-fast start at about (.3) second to saturate the filament slower and yet have 100% duty cycle at the max voltage output.
FOLLOW UP report:
I'm using stock tail spring for neg contact, I've since wrapped a piece of copper wire from top spring loop to the base loop. I charged the pack, let it rest until it was at 19.00 volts inserted it and hit the switch, the bulb blew after a quick 2 second bright light. I then let the batt pack rest a little bit more until it showed 18.80 volts and also turned the adjustment screw clockwise a bit more making it about a 1/3 turn from the full left stop in the pot. The bulb that I used for replacement had at least an hour of prior overdriven ~14 v Torch use.
This time it did not poof and the v. sag happened perhaps in the short time it's heating up to allow a somewhat rested pack to be used. LuxLuthor's "Destructive Bulb Tests" results, has reports indicating that the 623 bulb's max voltage (until it popped) was found through applying an accurate volt measure from bench power supply source in incremental increases up to ~16.9 volts. So now my having set the trim adjustment at this slight additional turn increment, perhaps a 18.5 volts open circuit vbat will not flash. I'm taking into consideration that the vbulb measurement will also show a drop down to/or a little under 16.5 volts. (my power is being supplied by this particular NiMH battery pack) ... This 18.5v, as my light is set up, could be considered a safe slow starting level at 100% Duty Cycle voltage. Your mileage may vary, due to taking into consideration any slight additional resistance fixes (tail spring ohms tests)... So go try and enjoy testing!
It has that interesting PWM buzz, definately louder when soft start is just loading up power.
Thanks for reading ... Blast-away!
Last edited:
