BTU Shocker Triple MT-G2 with a twist -- Aiming for >100Watt ~9000Lumens -- With external 2S power pack, handle etc...

Thanks for posting, very cool light you have there.

That seems to be the sensible way to use a reed switch, with an actual permanent magnet for activation. I came to the conclusion that trying to do an electro-magnet activated reed over the distances I would need wasn’t going to be a great idea or very efficient so I decided to go back to just doing a spring loaded contact to get the trigger switch line into the head.

Is there any progress in this project?
Can’t wait to see beamshots :open_mouth:

No progress till last night! :stuck_out_tongue:

Wow it’s been a while since I’ve been on here but it’s definitely time now to get this project finished.
I’m sure you guys know the feeling when you have 10 or so aspects of a project that all need attention and non are obvious or simple to knock off the list first. It’s easy to get sidetracked.

In any case I’ve now got two of those critical parts working. Namely the clicky switch circuit that passes power into the driver cavity through the contact board and the layout of the main power circuitry on the driver assembly itself.

First is the turning contact for the clicky switch. I just needed a raised contact patch that would complete the circuit between the switch wiring in the handle and the MCU in the head. That was easier said than done since it needed to be isolated from the grounded contact board and make contact only with the pin on the battery carrier when the body of the light is fully tightened down. Was fiddly to get the height and location of the copper patch tuned in but thankfully once I had it in the sweet spot it worked as planned.

It’s just a small piece of copper center punched to create the raised contact area, then insulated from the plate below with some kapton tape. Behind is a small hole through the aluminium with the wire to the MCU passing through it, the whole thing is then epoxied into place.

…and here is an old photo showing the spring loaded pin on the “battery carrier” that makes up the other half of the system.

Next I had to tackle getting the main power to the driver board. The positive will run directly to the LEDs so I cut a slot out of the driver assembly to accomodate the wire and for the ground I mounted a copper standoff to one of the driver posts. The whole assembly is grounded and each drivers outer ground ring soldered directly to the copper posts to make a solid contact. It’s really messy stuff so don’t look too closely, but hopefully it will be functional. The small red wires are PWM signal wires for each driver.

So that’s it for now but hopefully with the ball rolling again I’ll be making plenty of progress over the next few weeks. Next up is torture testing the whole power system wiring at full power (~18Amps) to make sure nothing melts or shorts out, fingers crossed. :wink:

Looking good. What are you going to do with the excess heat? May I suggest running a small power plant to supply electricity to a large town. Whats the eta for night shots? No pressure or anything.

Not a bad idea, though I’m hoping to get a few extra lumens out of the red hot glowing carcass of the BTU myself :wink:
I’m focusing entirely on this light at the moment so I should have it up and running well before Christmas, that is if I don’t spend too much time drooling over in your BTU monstrosity thread! :slight_smile:

This forum is incredibly intoxicating, so many cool and interesting projects going on!

Quick update on the driver, i’ve been building the final version of the control board and programming it with the STAR firmware.

Of course that means I spent the first 2 hours dicking about with the SOIC clip and building a whole second board because I wasn’t able to power up the old MCU for flashing…thought it was because the pins where a bit too angled and dirty. Of course once I built the second board I realized that star 2 being soldered for moon mode doesn’t allow the chip to be flashed…doh! That might be something to add to the Firmware annotations actually as I’m probably not the only fool to be caught out by this :stuck_out_tongue:

Anyway I ended up with a cleaner board than the first attempt and decided to investigate adding low voltage cutoff capability back to the zener modded driver.

By my calculations I needed to replace the 19kOhm resistor on the voltage divider circuit with a 47kOhm version to get the correct Vout range for the MCU to measure. I also needed to get a clean battery plus+ (>6v) connection for the voltage divider that wasn’t connected to the zener diode controller power circuit. Since otherwise the measured voltage would simple be at a stable 4.3v just like what the MCU sees.

Ultimately I’ve ended up with this setup. Bear in mind I haven’t yet tested the LVC using this board but theoretically it should work just fine, at least once I’ve dialed in the correct settings in the driver firmware.


Front of the control board showing the 47k resistor connected to the unused Led - traces which I’ve reappointed as a clean battery + supply. 200Ohm resistor as usual for Zener mod on the diode pad.


And the back showing the spaghetti wiring. From the battery + pad to the led - trace is a wire which I’m using to feed clean 6v+ battery power to the voltage divider circuit on the other side. The small long wire going from ground to star 2 is simply a workaround for not being able to flash the MCU when that connection is soldered. So now I just unsolder this wire before flashing but I’ll probably adjust the firmware so that moon mode is enabled by default.

Again the long thing wire is the PWM signal that will be driving the 7135 slaves and the bigger wires are power for this board.

Finally there’s the usual 4.3v Zener diode piggybacked on the the capacitor.

Big thanks goes to JonnyC and everyone who worked on the STAR firmware, the documentation is fantastic and the driver source is a joy to work with. Just fantastic stuff all round. Not to mention I have some 1uF caps on order to make this an off time memory driver just how I like it. :bigsmile:

Linus

RBD still comes to mind. I really am going to have to learn Chinese to understand whats going on. I can only wish you all the best on this.

It is a familiar look but with all the new board options available it’s possible now to get the Zener mod with lvp included. Something about a hacked up
105C still makes me chuckle though.

Haha yeah, this build is properly vintage style! :face_with_monocle:
For other retro BLF enthusiasts, you’ll notice those KaiDomain 7135v2 boards used as slaves. Those used to be all the rage before programmable Nanjg 105s came along.
Good riddance to that rubbish though :wink:

Planning happened this time last year so the approach is more than a little outdated I’m sure.
Things advance so fast around here, it’s awesome :slight_smile:

You might consider to change it to an electronic switch, that would reduce wire resistance and switchresistance….

I definitely considered changing the setup for an electronic switch when I started back on this project, especially with all the cool new driver firmware that’s around!

But it would have no effect on resistance since non of the main power is actually going across a switch. Power circuit to the LEDs is solely controlled by the 7135 slaves which are in turn activated or deactivated by the MCU board, the only thing actually getting switched on and off by the clicky switch in the trigger is this MCU board.
Hehe, if I was sending 18Amps across that sprial switch wire it would look something like this…

8)

Here’s a little update on the front end of the light.

Made up a cardboard template to make sure the emitters would be aligned nicely in the reflector, no centering rings will work and the emitters will be bolted down so getting things accurately aligned and in place is proving to be a fair challenge.

Looks about right :slight_smile:

Making a nice little nest for those Mt-G2s :slight_smile:

More to come soon.
Cheers Linus

…and all the emitters are nailed down! :bigsmile:

Fairly big milestone this, glad to get it out of the way as I can now make serious progress assembling and wiring up the front part of the light. Connecting Leds with the drivers etc.

-

Next stop is drilling the LED wire holes, and I also need to mount a registration pin that will stop the reflector from spinning when the front of the light is tightened down. I really don’t want to have any worries about accidentally dedoming these leds or anything like that.

And for those wondering, the copper squares are intended to sink some heat into the massive reflector. I have some more to cut to shape an add.
It’s a true “brushing the excess heat under the carpet” type of solution but should hopefully keep heat sag a little lower during the first few minutes…or rather seconds… :stuck_out_tongue:
They’re sanded very flat and just tall enough to make contact with the reflector once it is tightened down and pushes onto the led’s silicon skirt.

W - O- W !

:open_mouth: great job!

Thanks M4D M4X!

Just been sent this pic by MRsDNF and I’m pretty excited! Thanks again Steve!

It’s a BTU battery tube machined out to accept 4 18650 cells and it means I’ll also be able to configure this light in a 2s2p 18650 battery configuration for those times when a bulky battery pack isn’t ideal. Awesome :slight_smile:

Small update:

Received the 1uF micro capacitors yesterday so proceeded to slap one onto my frankenstein MCU board and flash it with STAR off time firmware.
Tested the driver last night with the default cap timeout values and the smile on my face was massive, this is what I call a reverse clicky UI! :bigsmile:

Just brilliant! Any thoughts of going for a momentary switch are gone completely.

I also tested the voltage cutoff with default values (128/118) and with my 47k resistor mod the step down seemed to kick in at around 5.35v. Bit low but not bad without any adjustments.
Of course I have to factor in any voltage losses in the cable before the MCU in order to monitor the true battery pack voltage (which can be quite significant at around 0.6v@18A) so this figure should give me an actual battery cutoff voltage closer to 6v.
I’ll probably tweak this to give me a safe cutoff near 6v for using internal 18650 batteries (which don’t see the voltage drop from the cable) and then I’ll just have to accept a higher cutoff when using the battery pack.
Fine by me since I don’t intend to run the lipo packs down that far anyway.

So the driver is ready to be installed. LED wires holes are also drilled so there’s not much stopping me from slapping the whole thing together now. Yay! Expect beam shots soon!

we expect…

Just made this sticker, have a feeling it will be appropriate here… 8)

Steady progress on the internals tonight.

Redesigned the main Battery negative connection on the driver board. Two nuts and a threaded bar now make removing the main connection possible without disturbing the copper post/slave driver at the same time. Want to make sure I can remove these main connections easily to make maintenance and adjustments easier.

I also added some AA and one of those compressible thermal putty pads to each 7135 board before clamping them down. Can’t hurt to keep those 7135s cool. Main heatsinking will be through ground ring soldered to the copper posts however.

Next was building the positive wiring beneath the led shelf, had to make sure this would align properly with the holes in the shelf and fit neatly between the driver assembly and the shelf. Keeping the wiring as thick as possible here to lower resistance loses.

And here is the driver assembly fully sunk into the head. Leaving just enough space to pack a couple coils of 14Awg cabling for the main power and of course the MCU board and it’s own wiring. The BTU driver cavity may seem generous but I’ll be using every last bit of it with this setup. :stuck_out_tongue:

Paramount is that I can easily get to the MCU board and reprogram it even after everything is assembled. Should only be a case of unscrewing the contact board and pulling it out.

And here is the front showing the wires coming through the led shelf.

Looking forward to seeing this running!

Did you plan all this or is it just happening? My head is spinning.