1st MTG2 Build

Nice work, good idea with the penny and how you trimmed it…nicely done!

MT-G2, look left at my avatar. Comfy and Richard and a couple of others have been working on something intriquing. They are removing all the chips from a Qlite or 105c driver and putting an FET on the board. Just one. It lets the emitter pull almost direct drive current from the cells. Wanna know how this works with the MT-G2? I just put one in my M8 today (Thanks Richard!), tested a lot of different cells. It works FABULOUSLY! :slight_smile: Want to control how hot you run it? Use a lesser cell, want to run it to the max, use a 20R or Efest 35A. I’m pulling 10.51A and making 4244 OTF at start-up, 3547 OTF at 30 seconds. :slight_smile: And oh yeah, moon mode on an MT-G2, how’s 7 lumens sound? It’s insane, loving it!

Looking good, keep up the good work and keep the pictures coming!

For the record, $1.08 apiece Sanyo UR18650FJ lap pulls do 5.82A and 2853 Out the Front at 30 seconds. Not shabby at all! With 5 modes I have 7, 55, 414, 1998, and 3547 lumens on demand. (with a pr. of 20R’s)

Edit: Even in a chopped AA MiniMag running from 2 IMR14250 cells on a “stock” Qlite, the MT-G2 will do 1346 lumens in mule mode (no reflector) In a small light, beware that it does indeed make a lot of heat, over 3A it grows exponentially.

Looking good Scotty. These MTG-2 leds are addictive. looking forward to the rest of your build.

Interesting phenomenon with the MT-G2's,pressing the power switch also activates the muscles responsible for producing a silly grin. No way around it.

Very cool and nice job with the pennies. Should be really cool when your all done. Thanks for showing your work in such a nice way.

Thanks guys. I’m really enjoying this. In fact I just worked on it some as soon as I got home. Still in scrubs, stethoscope in my pocket drilling into the copper I added to the pill. Something must be wrong… it appears to fit perfectly. Mind you the driver has not had the thick silicone wires soldered or been tacked to the pill with solder, just tested to see that it fits together nicely. Hopefully the reflector won’t ground out as I have no kapton tape. I dread having to go to Fry’s for kapton but I’ll cross that bridge if I have to.

DBCstm, your talking my language. Are you able to get high medium and low modes? The last that I heard there were issues with PWM and whining. I gotta get more copper. My next project is a triple XML and I was hoping to use that driver. I’m stoked it’s working for you.

I will update with more pics tomorrow. Thanks guys.

I posted the lumens for each level on the M8 with the FET driver in the thread up above, the Qlite in the chopped AA is running the standard Low-Med-High. No whining, no flickering, just works.

But if you want to save yourself some work next time around pick up some of these.

Itinifni, I have been looking at copper slugs, washers, wire, sheets and shaped punch outs on that Etsy… I was just impatient and wanted to see what I could do now. They have a great selection of stuff like these would be cool on top of a braided spring. I think all my discretionary money will be going to Richard’s Mountain Electronics and Etsy.

Great idea Scotty with the penny. I filed a few down the other night, along with a couple of old aluminum stars.

Now that I’ve seen your idea for shaping them down I sure feel like I’ve wasted some time. Thanks for posting and look forward to seeing your completed build.

Um, y’all do realize don’t you that destroying pennies is a Federal offense? If ya just gotta do it, perhaps you shouldn’t proclaim it on a public forum.

Copper sheet is readily available. There are punched copper discs to be found on ebay, even the cheap copper stars that have a mask under the thermal pad are useful for projects such as these. Government monies are wasted badly enough already…

And I’d only drill a hole to mount the disc if I was planning a triple that had a hole in the center of the star. Otherwise, you’ve removed precious heat sink material from directly under the source…the emitter. There was an article about the path of the heat under the emitter, accompanied by spectrograph charts, that showed the heat tends to take a similar angle as the beam of light itself. So the area directly under the emitter is crucial.

Would love to see these tests. We’re the emitters on stars with big heatsinks sucking the heat down? Is there any way to direct this heat outward towards the body?

If you draw a V that represents the radiant angle of the light coming out the front, extend the arms of the V down into an X and that’s kind of what the heat dispersal looks like under the emitter. The only real way to get the heat to disperse is to bottleneck it, making it spread out, much like blocking the light with a reflector to guide it. BUT, in so doing the path of heat dispersal is also slowed, causing the emitter to run hotter and thereby less efficiently. Heat is going to be radiant by it’s very nature, so giving it good contact will automatically result in it filling up all the sink material available, with ambient temperture being a key factor in removal from the surface. This is where the fins come into play. Are the fins adjacent to, above, or below the heat source? Would seem to me that fins below the heat source would be more of a sponge, wicking heat from the source to the outside air. Fins above the source are only going to come into play when everything is saturated and effectively backing up to the fins. Adjacent would be a compromise between those two factors.

I’m no engineer, just someone that’s been burned. lol I’ve worked outdoors in the Texas sun most of my life, radiant heat is a way of life. You DO NOT lean your forearms on a dark rusty metal gate in full Texas sun in August! Or the blue hood of a car. Or the dark green plastic slide some idiot put in the playground for my kid to play on. For me, it’s like running a water hose over a slanted sheet of plywood, the water will run straight down and leave the board if nothing stops it, put a deterrent in the way and the board will get wet on a wider path. Similarly, let the heat escape and it will take a straight path out. Block it, and it will back up and saturate everything. Right? So, getting a direct thermal path under the emitter and enough sink mass to radiate the heat out to the fins for the air to take away, and there’s your efficiency. Simple enough on the face value, more difficult than that in application. So now I can be shot down with all my ideas. :stuck_out_tongue:

Run an MT-G2 at 10.5 - 12 A and see where the heat goes…

Speaking of heat… #%$&, I destroyed my driver in a pathetic series of soldering whoopsies and and hurried shoddy repairs. I was so excited and so close I just knew I could make the repairs and get it to work. There goes $20. Will be placing an order with Richard tomorrow. I can salvage the chips to use on another build using reflow. I can do reflow but the soldering is still a work in progress. So a little break until more parts arrive.

What happened Scotty?

Sorry to hear about the bad luck. Great to see you have a stick-to-it attitude. Says way more that a easy success does.

Holy Hell it's nothing but LED!

Wow...that thing has GOT to get warm really quick (there just ain't much meat on a Convoy S* series), but most awesome!

What kind of power are you going to push thru it?

Hope it goes better round 2! :slight_smile:

RMM, I ordered the driver with a bunch of chips on the bottom, so I could not use the securing ring and had to solder the driver to the pill. Hind sight vs fore sight. That took a while to heat up to take the solder, I overloaded the tip and pill/ driver - didn’t secure the pill well and I got a huge blob of solder on the positive lead and the (sense resistor) chip and some of the chips you stacked really pretty. First fix I tore the plate or base thingy off of the board for the + contact. During the process of scraping, wicking the solder off I busted a resister off the other side. Fixed that after I found it. Ugh. Put it together and the positive just snapped off and I can’t see the trace or contact pad to re-attach.

Good news is I did not burn myself this time. That’s a first in a long time. :slight_smile:

Thanks Ima4wheeler.

Any driver using the 9.4kHz firmwares are prone to audible whine, no getting around it, it's not caused by any of the hardware but just the PWM frequency. They do it even with 7135s. It's most noticeable in the modes around 50%, it goes away completely at 100%. If you have one that doesn't squeal it doesn't mean anything except that you got lucky and nothing in your light happens to resonate in the 9-9.5kHz range.

9400Hz is a really annoying tone, but most humans can't hear 19,000Hz. The wiggly parts in the light (springs, switches, even the cell rubbing the I.D. of the tube) that make noise at 9.4kHz are probably also vibrating with the 19kHz firmwares, we just can't hear it.

Scotty, sounds like you need a bigger soldering iron or bigger tip. Using a higher temperature and lots of flux is also best when soldering to the pill. Generally if I have to hold the iron on the pill for more than 5-10 seconds before getting a good bond then something is wrong, the less time you are in contact with the pill the less overall heat you will transfer. It may seem counterintuitive but sometimes a higher tip temperature can end up causing less component damage.

There is a lot of thermal mass in the brass pills so in order to solder to them without heating up the whole assembly too much you need to be able to transfer enough heat fast enough into that particular section that it heats up to soldering temperature before the heat can be pulled away. That's not a very clear explanation is it? Pretty bad actually. I'll try another poor analogy. :)

Let's say we have a kitchen sink. In order for the solder to stick correctly the sink must be full of water (sufficient local temperature), but we want to minimize the amount of water going down the drain which will spill all over the floor if it overflows the small drain bucket it's connected to (amount of heat spread over the entire pill assembly, water spilling is sufficient heat to melt other components). The drain size is fixed (thermal conductivity of heat away from soldering area).

With a small tip and low temperature, it's like the water is barely outpacing the flow of the drain. It may eventually fill the sink (provide enough local heat for solder to flow and stick) but we will have drained a lot of water out of the sink before that happens, possibly creating a big mess on the floor (overheated driver). What we need is a larger, faster water spout (thicker tip / higher temperature / higher transfer). This way we can fill the sink very quickly before very much water has a chance to drain out and spill everywhere.

That's a very rough (and slightly inaccurate) analogy but it makes some sense.