SST-90 Mag 2D aspheric build, low resistance, direct drive

Candle Power Forums

Help Support Candle Power:

Re: SST-90 Mag 2D build in progress - Update #3.5

This time, with a new twist, I'm revisiting this extremely low resistance 4.8V battery configuration again:
4s2pexperimentcollage.jpg


I've ordered five of the 8X7135 2.8A boards from KD. If they continue to be back ordered by the end of the week, I'll just get the 10 pack of 4X 1.4A boards from DX and stack them myself. I plan to use 3 of the 8x boards (or 6 of the 4x boards) in parallel for a total of 24 AMC7135 chips and 8.4A of regulated current. (The other boards are for building an SST-50 torch the same way.)

I've run the 2.8A boards on 2X LiFePO4 batteries which is 6Vin+ without them overheating, but they were somewhat heatsinked the by way they were mounted. I'm hoping that if I can't figure out a good way to heatsink them that on ~4.6V input they wont overheat and fail on me.

If this works, I should have 30+ minutes of runtime, most of it regulated at 8.4A, and still keep the formfactor down to 2D. Mostly I'll be proud to have assembled the ultimate poor-man's SST-90 max current regulator/regulated torch. I wish I had chosen "The_Poor_Man" as my handle on CPF so I could write "The_Poor_Man's guide to...."

Question: Should I be concerned about abusing the Duraloops this way? In this two-parallel (four serial) configuration, each battery is splitting the 8.4A load and so constantly running at 4.2A current draw. Vout is sustained pretty well in these batteries (link to chart), about 1.15V at that draw, but will I be damaging them (quickly)? That could be a pretty expensive prospect for the poor man. (Oh, that poor man...)

I suppose I could move everything to a 3D and squeeze in 16AA as I did in the LED USL, but this time configured as 4s4p. That would double runtime to an hour and reduce battery load to 2.1A each...
 
Re: SST-90 Mag 2D build in progress - Updated

Techjunkie said:
2) The torch is able to maintain up to 12A current draw without the tint shifting to blue. I'd say that based on that, the heatsinking is doing it's job.
Click to expand...

+1

I realize these SST emitters are already a bit tougher than MCE an P7s, but IMO to say the heatsinking is "doing its job" is an understatement.

I dont have any of the SST emitters, but I can tell you when the layer of epoxy between a bare emitter and heatisnk is even a little too thick, or insufficient in any way a quick:poof: is commonly the evidence.

I am actualy surprised at how well this led has stood up to the torcher.

When I first started using bare MCE emitters, I put one on a mag C heatsink with AA epoxy about .5mm thick ( I know I know:eek: ) but it didnt seem like much at the time. I put 3 rested Duraloops in and in less then 30 seconds got a very blue color and then death:confused:

This being said, if something is "wrong" (not saying there is) I dont think that your trouble lies there:D
 
Last edited:
Re: SST-90 Mag 2D build in progress - Updated

Thanks. I'm glad you agree. When I do use thermal epoxy instead of solder to mount bare emitters, I usually slide the emitter back and forth across the heatsink to spread the epoxy as thin as possible. This heatsink doesn't allow you to wiggle it at all. I had to push down firmly from the top while pulling the wires from the bottom to make sure as much epoxy as possible squished out from the wire turrets at the corners. I'm confident that I did the best job possible with these materials, and the proof is in the puddding - no :poof:

vestureofblood said:
+1

I realize these SST emitters are already a bit tougher than MCE an P7s, but IMO to say the heatsinking is "doing its job" is an understatement.

I dont have any of the SST emitters, but I can tell you when the layer of epoxy between a bare emitter and heatisnk is even a little too thick, or insufficient in any way a quick:poof: is commonly the evidence.

I am actualy surprised at how well this led has stood up to the torcher.

When I first started using bare MCE emitters, I put one on a mag C heatsink with AA epoxy about .5mm thick ( I know I know:eek: ) but it didnt seem like much at the time. I put 3 rested Duraloops in and in less then 30 seconds got a very blue color and then death:confused:

This being said, if something is "wrong" (not saying there is) I dont think that your trouble lies there:D
Click to expand...
 
Re: SST-90 Mag 2D build in progress - Update #3

Techjunkie said:
I rebuilt one of my Dewalt 18V tool packs with Tenergy 3500mAH sub-Cs and it now beats the pants off of both the OEM NiCd pack and an aftermarket 3000mAH NiMH pack that I bought a year before. Torque is stronger and rotational speed is faster. (Unfortunately the torque is so high that the "hi-amp" solder tabs the batteries came with failed under the load when using a 2.5" hole saw bit. I had to shorten and resolder the one that popped.)

If the ROV cells compare at all to the Tenergy cells I used, then I'm sure you'll be very pleased with their high current capability and low sag under load. I'm purely guessing here, but I assume that sub-c's are primarily marketed for high current applications like tool packs.
Click to expand...


hey techjunkie, very impressive builds. i am trying to make a thrower out of a sst-90. im new at this so i figured doing a mag mod would be easiest. do you think i should go with a C sized body or a D? (i dont mind the extra size, im concered with brgihtness of the light) im going for the brightest farthest throw, not long run time, so morer like a show light. just looking for some pointers. i was gonna get the britelumens sst-90 heatsink, the dx aspheric lense u recomended for the 90. so i was still curious about the batteries to use for maximum output and reflector, if u think one is needed with the aspheric lense. and possibly a new tailcap spring and switch. i would also need 3 drivers wouldnt i? thanks for your time and help.
 
Re: SST-90 Mag 2D build in progress - Update #3

Hi Tj,

How long did it take for your emitter to arrive from Avnetexpress?
 
Re: SST-90 Mag 2D build in progress - Update #3

Fulgeo said:
When I first ordered a SST-50 and a SST-90 from them it only took about 5 days.
Click to expand...

Techjunkie said:
I checked to make sure they were in stock first, so it only took about a week by FedEx Ground. (I'm in NY)
Click to expand...

Awesome thanks guys. I am thinking about the top bin 5700k myself. I have found the true neutral whites to be just a bit on the yellow side for me. I of course dislike the blue/purple of the 6500+ even worse.

I just wanted to know for sure before I ordered. There was something about a "13 week factory lead time" written there that scared me, but I guess that's just if you order more that what is in stock.
 
Vestureofblood when I ordered my initial emitters I gave them a call to make sure that they actually had them in stock. They were very friendly and at the time they had the emitters in stock and ready to ship even thou their web page did not reflect this. Also for some reason my 5700K SST-50 WJ at 4.2 amps is noticeably brighter than my 6500K SST-50 WJ at 4.2 amps. I had an early conversation with a Luminus engineer on this subject and his opinion was that the 5700K might have a slight edge over the 6500K. I got myself a 6500K emitter and when I put it together and tested it I thought to myself "I'll be damned he was right!". The 5700K SST-50 emitter is defiantly a bit brighter. It also is a bit on the vanilla white side. The 6500K is a nice cold white. I prefer the "colder" shade emitters over the warm as long as there is no blue or green preservable. So I prefer the 6500K tint, but this is personnel preference. I have 3 of the 6500K SST-50 emitters. I have sent one of them to jtr1962. Hopefully in a few more weeks he will have the testing done and post the results in his "White LED lumen testing" thread.
 
Great thread and good work.
Saw your 3 videos on Youtube today.

Hope 10A drivers are available soon!
 
Nice thread. Can you address the question about why a reflector is needed with the aspheric?

Also, have you ever had the chance to compare your SST-90 with the Olight SSR-90? I'd say you just saved yourself $350 or so!
 
Last edited:
Walterk said:
Great thread and good work.
Saw your 3 videos on Youtube today.

Hope 10A drivers are available soon!
Click to expand...

Thanks, Walter. I just realized that I never went back to post #41 to add a link to my Big Red build where I think I've found the best possible reflector option (Mag Rebel reflector with emitter mounted down in the neck).

Both of my SST-90 Mag 2D builds make me very happy. I get maximum output from the amazing SST-90 emitter, very repspectable runtime with consistant output, and didn't have to use any exotic/expensive/unproven drivers to achieve either of them. I should quit while I'm ahead. (So I started experimenting with SST-50 emitters instead :naughty:)
 
gcbryan said:
Nice thread. Can you address the question about why a reflector is needed with the aspheric?

Also, have you ever had the chance to compare your SST-90 with the Olight SSR-90? I'd say you just saved yourself $350 or so!
Click to expand...

No reflector is needed with the aspheric. It's an either/or kind of a thing. I was experimenting with reflectors with this one, but only liked the Mag Rebel reflector which wouldn't fit with the BriteLumens heatsink, which is why I built another SST-90 torch where I sank the heatsink into the neck. That one is regulated and runs on AAs. This aspheric is direct drive and runs really, really damn well on Cs.

I'm not one for expensive name brands, but I certainly would enjoy comparing them. I'd rather mod a bunch of Maglites or cheap Chinese hosts than buy a single high priced light. I modded a Streamlight once, but it was a beater that I only paid $15 for.
 
I made a 4D nimh cells direct drive on a sst 90. Tail cap current was measuread at about 4 A, using the thin DMM standard probes. I was running before the emiter from 4 alkalines, no problem at all. Now, the emitter just :poof: after some seconds when running at 4 NIMH´s D size.

I will rebuild this flashlight using two of these resistors, in paralel, to get 1 ohm resistance, to run this in 4 D NIMHs:

http://www.britelumens.com/images/930.jpg

Do you think guys, I will not fry again my expensive SST 90 :crazy: ?
 
gadagain said:
I made a 4D nimh cells direct drive on a sst 90. Tail cap current was measuread at about 4 A, using the thin DMM standard probes. I was running before the emiter from 4 alkalines, no problem at all. Now, the emitter just :poof: after some seconds when running at 4 NIMH´s D size.

I will rebuild this flashlight using two of these resistors, in paralel, to get 1 ohm resistance, to run this in 4 D NIMHs:

http://www.britelumens.com/images/930.jpg

Do you think guys, I will not fry again my expensive SST 90 :crazy: ?
Click to expand...

I think that because the SST-90 current draw scales so incredibly quickly with the smallest changes in voltage, and because proper heatsinking is so incredibly important to keep from destroying it at high currents, that SST-90 builds are not for novices. Four D-cell NiMH certainly produce enough power to kill an unregulated SST-90, but if it were heatsinked properly, you should have had some advance warning in the form of blue tint shift for at least two seconds before the emitter completely died.

If you decide to try again, a resistored setup is a possible way to go, but I do not recommend it. For starters, if you used two of the resistors that you linked to in parallel, you would likely kill the second SST-90. Those resistors are 0.1 ohm each. Two in parallel would yield 0.05 ohm of resistance. Your application calls for at least 0.15 ohms of resistance and the resistors would have to be able to handle at least 12W. (Your power source is 4.8V, the emitter calls for about 3.5V to run at 9A, so you would need the voltage drop across the resistor to be 1.3V. To cause a 1.3V drop at 9A, 0.1444 ohms resistance is needed using the formula 1.3/9. The resistors would have to disspate 9*1.3 or ~12W of heat.)

You could use 3 of the 5watt 0.47ohm resistors that RadioShack sells in parallel to achieve the right combination, but an even better solution would be three 2.8A 8*7135 regulators wired in parallel. With the regulators, you would have a constant 8.4A current over a very large voltage range and as Vbatt drops below 3.5A, the regulators would maintain a higher current than the resistored setup would.
 
Techjunkie said:
I think that because the SST-90 current draw scales so incredibly quickly with the smallest changes in voltage, and because proper heatsinking is so incredibly important to keep from destroying it at high currents, that SST-90 builds are not for novices. Four D-cell NiMH certainly produce enough power to kill an unregulated SST-90, but if it were heatsinked properly, you should have had some advance warning in the form of blue tint shift for at least two seconds before the emitter completely died.

If you decide to try again, a resistored setup is a possible way to go, but I do not recommend it. For starters, if you used two of the resistors that you linked to in parallel, you would likely kill the second SST-90. Those resistors are 0.1 ohm each. Two in parallel would yield 0.05 ohm of resistance. Your application calls for at least 0.15 ohms of resistance and the resistors would have to be able to handle at least 12W. (Your power source is 4.8V, the emitter calls for about 3.5V to run at 9A, so you would need the voltage drop across the resistor to be 1.3V. To cause a 1.3V drop at 9A, 0.1444 ohms resistance is needed using the formula 1.3/9. The resistors would have to disspate 9*1.3 or ~12W of heat.)

You could use 3 of the 5watt 0.47ohm resistors that RadioShack sells in parallel to achieve the right combination, but an even better solution would be three 2.8A 8*7135 regulators wired in parallel. With the regulators, you would have a constant 8.4A current over a very large voltage range and as Vbatt drops below 3.5A, the regulators would maintain a higher current than the resistored setup would.
Click to expand...

Thanks for the tip. Are you refering the regulators from dealextreme or kaidomain ?
 
gadagain said:
Thanks for the tip. Are you refering the regulators from dealextreme or kaidomain ?
Click to expand...

I used the KD regulators becuase it was less work to wire up 3 of the double 1.4A boards than 6 individual 1.4A boards (DX doesn't have 2.8A doubled boards). You could go either way though. I think DX is sold out on the 10 pack of 1400mA 4xAMC8135 boads, so if you don't need 20, then KD is your next best option (they still sell them 10 at a time).
 
Techjunkie! Thanks so much for all your effort put into this thread!

I have finally decided to begin playing with the SST-90 and found your experiences in this thread to be most valuable! You have definitely inspired me to work towards an aspheric mod.

So, if you were to summarize the battery configuration you found preferential in direct driving these SST-90s, would it be three C NiMHs?

If you decided to start again from scratch (like I am) how would you build your SST-90 mag now?

Thanks again for so much good info!
 
NeSSuS-GTE said:
Techjunkie! Thanks so much for all your effort put into this thread!

I have finally decided to begin playing with the SST-90 and found your experiences in this thread to be most valuable! You have definitely inspired me to work towards an aspheric mod.

So, if you were to summarize the battery configuration you found preferential in direct driving these SST-90s, would it be three C NiMHs?

If you decided to start again from scratch (like I am) how would you build your SST-90 mag now?

Thanks again for so much good info!
Click to expand...

NeSSuS-GTE, If I were to do another SST-90, I'd probably go the same host (Mag 2D) and battery (3x NiMH C direct drive) route again, with the same resistance mods, and only change the heatsink to a pedistal-less one recessed into the neck and use an SSR-90 on star, mounted not with epoxy but with thermal grease and screws. Basically, design-wise, I'd combine this torch with my Big Red torch that I described in post #41. I think it would provide the best bang for the buck and would be the least risky design for many reasons.
 
You must log in or register to reply here.

Latest posts

Back
Top