17mm/22mm MTN-MAXlp - Low-Profile 1A-4A Programmable Buck Driver + 22mm MTN-MAX

Your first mistake is believing that even I have any idea what I'm going to be working or releasing.

For 20mm/22mm there's enough room for a decent sized inductor, so there's no need for two boards with the 20mm and 22mm. My early attempts were to "have it all" in a single 17mm design, and those didn't work out very well.

I wasn’t trying to put words in your mouth. The OP is about a low profile 17mm buck driver and you mentioned a taller 6a+ Version. I apologize if I wrongly guessed you would release a 20mm version once you were able to get the 17mm worked out. Ultimately, we are all just sitting back in anticipation of what you are able to come up with.

Just to clarify, here's my current roadmap:

17mm, 2 boards:

i. MTN-MAXlp: 4A max, low-profile, small SMD inductor. [RELEASED]
ii. MTN-MAX: 6A+ max, larger external inductor. [IN TESTING]

20mm, 1 board: MTN-MAX: 6A+ max, either a larger SMD inductor or external inductor. [IN TESTING]

22mm, 1 board: MTN-MAX: 6A+ max, single sided, either larger SMD inductor or external inductor. [IN TESTING]

Plus some other larger boards down the line, with the same basic layout but improved efficiency due to the availability of space for a larger inductor.

Wow - this is great to have a true programmable buck driver! Thanx again Richard!

Question: I'm sure I missed this somewhere, but how do you program one of those MCU's in that form factor? Can it be programmed on-board?

I think I saw these listed on your site but didn't know what they went with. Can't navigate Mtn-E now to explore...

there’s a mmu programming board in the oshparks thread:

https://oshpark.com/shared_projects/uyjO7GgP

Does this mean no option for doing on-board re-programming with this MCU?

Yes. You're gonna have to pick a firmware and stick with it for a while, which I know will be hard for you to do!

I played around with adding some breakout traces to solder wires to, but that's more of a pain than just reflowing the dang thing and it makes a mess on the battery side.

Edit: If you were really determined, you could get away with just two breakout traces for RST and MOSI, then solder the other wires to the existing components that are used for the other signals, but that I think that may be more trouble than it's worth.

Ok - may be a challenge when tweaking LVP, timing checks, etc - those are more of the reasons for me vs. changing mode selections later on. May be a good bench setup could be done for doing those kind of tests/tweaks?

Couple of reflows of the MCU would probably be dicey for me... Dunno...

Just use the standard 130 off time value. No need to tweak. All that I've built have been pretty consistent.

Ahhh - so one value for any # of cells (of course parts are different for different # of cells as stated)? That would be real nice... I'm think'n my need is 2 cells right now.

Oh, no. Read that on my phone and missed the LVP part. Right now it's set it for one set of cells and go, but I was going to get around to adding some logic that would let it work with 2S-3S without changing anything in the firmware.

^Cool stuff. No wonder you’re such a busy guy :wink:

From what little research I have done the zener mods are inefficent at low power, thus I have been waiting on a (less then 4 amp is fine) 6V buck to run the 6V emitters but with out spilling the difference into heat. I am no expert and I might have gotten somting reversed but it seems that for other then DD way hot short run lights a more sophisticated driver is needed.

How are you controlling the amperage of these drivers? I see 4 amp and 6 amp listed. Is it a function of what the battery has available? If it is then other than the components used, isn’t it DD?

Look up “buck converter”.

Yea, well that is beyond me. I don’t get it. I’m reading more…
I really wish I could understand what exactly different size inductors do. I’m going to start studying it…
BTW, I need another M6, when are you opening back up? :bigsmile:

Richard. Be a nice boy now and get back in that nursery and look after the good wife before I come around and spank you. This forum will survive without you for a few days. Saying that, will you survive without us for a few days?

This may be less efficient running an MT-G2 off of two 18350s, depending on what current level you set it at. I suspect that at 3A or more the 7135 driver will be on average more efficient over the full discharge of the batteries. If you wanted to run an MT-G2 at something silly like 1.5A or 2A the buck driver would probably be more efficient.

Linear regulators, like the AMC7135, are pretty efficient where the input voltage is close to where the output voltage is needed to maintain a certain output, with all extra being turned into heat. With 18350s at 3A and above there is enough voltage sag that you're not staying very far above what is needed for very long.

With a buck regulator, you generally have more switching losses, and you are also always running all of the LED current through a sense resistor, which burns off power, and through an inductor, which also burns off some power (especially smaller inductors). So, when input and output drop to the even, or almost even point, the efficiency of the small buck driver will be worse than what you would have gotten out of a 7135 based driver.

Now, remember that this may not hold true for 2x26650 or 2x18650 running an MT-G2 at lower amps, but an MT-G2 with 2x18350s at 3A+ I think that the 7135s are actually doing a pretty good job efficiency-wise.

Has anyone received their boards yet? Anyone built one yet? I've built several of these now in different configurations and have them in different lights, and I feel very confident that they work and work well.

A few more thoughts and tips:

1. Absolutely do not try to solder this by hand. You need to reflow to get these done right, especially since the thermal pad under the buck IC is critical to proper performance.

2. Make sure you get enough solder underneath the Buck IC's thermal pad. The thermal vias tend to wick up some of the solder, so you need a bit more than you would think.

3. Space your low and critical levels a bit farther apart than you normally would. For example, if you normally use 160 and 145 for your low and critical values, use 170 and 145 instead. The way that the buck driver loads the batteries is different than with a DD or linear driver single-cell setup, because the current draw actually increases on each cell as they drain instead of decreasing, so having values that are too close together leads to an immediate shutoff after the LVP is activated with no or very little "in between" time.

OshPark expects my boards to be back around 4-19. The OP was 4-5 and I ordered 4-7. That being said I don’t even have all the stuff to attempt this build yet, nor am I confident enough that I won’t screw it up.