Prototype off-road/boating spotlight - King COB!

Yeah, I pulled out a pile of old 18650s that I have last night and realized I could probably fit 12s2p inside it with that little LEDiL reflector if I wanted to. Though, I’m actually thinking about using 3s 18650 battery holders for maintenance reasons. (3s are the largest I’ve found to date) I’d like to put a balance connector and charge connectors as well on it but I’m not quite sure how to do that safely.

Digikey has 4s but it’s pin mount… Though you could probably just grind or snip off the pins and solder on leada.

Found models for the COB and Ideal mounting components.

So… Do you have machining capabilities? This looks pretty cool. I should be getting those TIRs in this week so I won’t draw anything up until I have them on-hand - because if I can use something existing as a host I will.

Speaking of which, I got in the CMT2890 so I get to play with that a little…

I have access to a Tormach PCNC1100. It’s a piece of junk but I can use it whenever I want. :wink:

Soon, I may have a running manual lathe and a Bridgeport as well.

LOL… I have an import mini-mill, of which I purchased only the castings, machined for the motion components, and installed precision LM rails and ground ballscrews. Though I’m growing out of the open-loop steppers I have installed and looking to install either servos or closed-loop steppers. I also have a manual mini lathe that needs to be tuned up.

What are you using for CAD/CAM?

I have a copy of FeatureCAM that I use… mostly because I know how to use it. I need to install a copy of Fusion 360 and see what that is like. I’ve heard it is very capable but can be a pain.

We just had all the stepper drivers on the Tormach fail. I managed to repair two and then switched the A-axis drive to Z. I looked at upgrading to polyphase and to closed loop. Closed loop seemed really expensive and upgrading to polyphase would have been my choice if the $5 12v regulators didn’t fix the drivers.

The biggest failings of the Tormach are the ways and the spindle. The ways just have so much slop in them. Even with fairly tight gibs, you can still easily move the table 0.003”. If it was my personal machine, I’d consider tearing it down and scraping them or going to linear guides. The whole machine is just kind of junk. The new MXs are crazy overpriced for what they are even though they now have linear rails and servos. If the new 1100MX was ~$15k and had a 3HP spindle, it would be about right for a home shop.

LOL you can get a used HAAS for the ctazy prices Tormach wants.

You can get closed loop Leadshine steppers for a reasonable price. A lot cheaper than buying them from Tormach. Then gut their control and use a UCCNC control. Even a SmoothStepper with Mach3. PathPilot is nice but it’s all proprietary crap.

I’ve run a Haas before but yeah, I was looking at a DT-2. For $50k new, it’s a hell of a machine. But I doubt I’ll ever have the budget or space for something like that. Haas controls have some really irritating software bugs and I’ve heard that their newest control is REALLY bad.

Anyway, I’ll just use the Tormach for one offs and prototypes and then outsource any production.

I really don’t mind PathPilot all that much. Their UI stinks and there are some dumb quirks like not being able to easily drive the machine off a limit switch. PP is just LinuxCNC with a Tormach sticker slapped on it. It’s actually funny but I got them to sell me the upgrade even though they shouldn’t have sold it to me given that our machine is “too old” to run PP.

There’s a guy who designed a lever operated hybrid hydraulic drawbar system for the Tormach that I’d love to have on the one I have access to but I doubt that will ever happen.

Doing some thermal modelling on the battle lantern tonight and it looks like it will be able to handle 35-40w continuous or there about. It could probably do a little bit more if I used copper here and there but copper is expensive.

Yeah they have some quirks but for the money they’re pretty good machines. And for most stuff they’re more than serviceable. If I had the space and money I’d like a Brother Speedio, but still waiting to hit that Mega Millions.

I’m surprised how long it took Tormach to come out with a machine with LM guides and servos. But you could get an SYIL for about half what Tormach is asking.

On my mini mill, I had a 3HP 24krpm electric spindle installed, but I basically work with aluminum about 99% of the time, and I use high-speed (constant engagement) toolpaths almost exclusively. Only thing about the spindle is that it uses ER collets, so no ATC…

I’ve seen a few hydraulic drawbar assemblies, and it really is not a big deal to build. You could probably find a good hydraulic cylinder on eBay, fabricate a couple simple plates, and get some Belleville washers - the concept is pretty simple.

From what I’ve heard, the only problem with the Syil is that it requires 3 phase even though it should be able to run just fine off single phase 220vac.

I saw another machine not too long ago which looked a little nicer than the Syil but I didn’t look into it all that much. YCM FV50T It kind of looks like a Brother Speedio.

I’d love to have a machine with HSK tool holding but that’s basically not a thing here in the US it seems even though it beats the hell out of CAT30/40/50.

The control software bugs drove me up the wall because I was doing a lot of “screwy” heavy hand coding of parts. The other machine I used to run, ran FlashCutCNC which almost has more in common with a proper programming language than g-code. You could do some pretty cool stuff.

FlashCut CNC if I remember correctly was one of the first “prosumer/light industrial” controls that ran off a USB port. That was actually a pretty big thing back then, maybe about 12-15 years ago?

I wouldn’t say g-code is not a “proper” programming language; it’s been around forever. I used to hand-code stuff, but unless it’s something simple I do everything in CAM. While I like the look of efficient code, it doesn’t matter anymore today because there’s no worries about system memory. Heck even Mach3 is capable of over 100 million lines of code, which even a lowly Pentium I and a parallel port could parse, no problem. I have Cut2D for simple 2.5D stuff and a child could generate toolpaths with it, yet it’s capable of doing some complicated stuff if you use enough imagination. I grew out of that however and got VCarve for its signage capabilities. Finally for full-on 3D programming I got a seat of OneCNC; and I’ve used it since XR2 Mill Advantage which is at least 10 years ago, and have upgraded over the years (I’m at XR7 Mill Expert right now.)

At my old job the FlashCut machine made a family of micro-machined parts so I wrote a pile of subroutines that calculated tool paths, feed rates and spindle speeds on the fly. I could even do constant surface speed on ball mills to compensate for the contact angle. Anyway, with the set of subroutines, I could code new parts MUCH faster than any CAM operator could by piecing together the required subroutines.

When I say “proper language”, I mean, it could handle named variables, structured programming and could do a LOT of math. Sadly, last I checked, FlashCut still hasn’t updated it at all. The only problem I had with it was that it had a memory leak so you would have to restart the control software a few times each day if you were running really complicated code.

Well, I have a LDH-45B-1050W coming to play with on the battle lantern since I don’t think it will be able to handle more power.

It would be fun to have a turbo button though…

I used one on a 50W outdoor flood light given to me with a bad druver. Works like a charm, dimmable, and I was able to work on the car at night for about 2 hours with my 3s LiPo only dropping to 3.6-3.7V according to the LVP I had connected to it.

Looks like I got those TIRs and housings in… hopefully tomorrow I ca look at it, take some pics, and set up a test rig.

I’m hoping that MW LDH arrives tomorrow so I can ghetto rig something and take a walk around the neighborhood.

I wonder if an aspheric zoomie COB light could be built. Maybe use one of the ~9mm LES COBs? Since it’s a COB and not something like an XHP70, it wouldn’t have that goofy square throw beam.