Looks great! So, does it now have pad size for 85’s without bending pins?
Sadly no, the pins are SO long they want to cross over the ground ring which could/would cause a short.
It is possible I could rearrange things to fit the 85 pads but it would sacrifice something else.
15mm is simply not much room to work with while keeping enough clearance between parts to hopefully be able to use 0805 components in a pinch along with on-board programming.
OK, thanks! Still a very nice driver. Thanks for making it. Still waiting for your complete/built drivers sale thread to pop up. 
It will be a little while till that happens. People take for granted how long it takes to actually build these. Just don’t have time for that right now.
Plus if I was going to do it I would prefer do a run of ~100 of them but I can’t find any production houses that do small quantity’s for a good price. I have not had time to look very hard though.
Oh boy, been away a bit on this one. Gonna order both designs now 
, both are look'n good! I could use both, but of course they are sooo expensive, I'll have to split the order over two credit cards.
Let me see, the cost of 30 of these boards is the same as 3 SRK drivers... I think we need to go small more often. Actually, it would be cheaper to use the stock SRK driver and piggyback in a few 15 mm drivers. I don't get the 10:1 cost ratio, but guess it works out to 3x3 15's are the same size as one 45x45, so 46x46 is about equal to 10.
Edit: Ok, did some side by sides, and decided to order 3 of each in thinner substrate, and 3 of the "normals" in the thicker substrate for a whopping $5.25 for 9 boards.
K, ordered!
Nice, can’t wait to see how they work out, I usually like to be the first to try out a new design but frankly I don’t have any lights these could be used in.
You have also discovered why the Q8 PCB’s are a little impractical to buy for modding, you could end up spending as much on the PCB’s as the flashlight unless you will be modding several.
I ran into this same issue when trying to get some drivers for the SRK’s I modded. Just so much more costly compared to ~17mm drivers.
I sure hope that TF accepts the tailcap PCB at least, that cuts the cost in half to convert to XHP’s. Unless we went with a buck driver I don’t think using a single driver for all the desired setups is real possible.
That said a 3 channel driver for the Q8 would still be a very cool addition IMHO.
You would not believe how many blank OSHPark driver boards I have laying around here - like 50+? I've been buying them since day 1 we discovered the service and came out with the first generation. It seems like every time things would settle down, I'd stock up, then a new design/approach would merge. Actually the first OSHPark boards I bought were a flashlight driver that was Z8 based, and after buying all the dev tools and parts, I gave up on it. I also have 8 SRK unused drivers as well, mostly from the BLF Q8 iterations. Probably got about $150 worth of wasted driver parts.
I thought Richard would be the settling influence - it's great he offers the drivers in all 3 forms - assembled/tested, bare board, part kits. But at this point, I don't care for his zener approach to the 25/45/85 issues, basically killing practical use for e-switch lights with the parasitic drain, which he says is a very small percent of his sales - I'm sure that's true. I wish he'd get on board with DEL's design of using the 4.7 ohm resistor, but he appears to be uninterested. No sign he's even monitoring the threads anymore. He just came out with a 20 and 22mm version of the zener design now, so I guess he's committed to it for the next stretch of time.
I still use his BLF17DDm v1.1 driver boards though, but I do some creative chip mounting to add the 4.7 resistor prior to the Schottky diode, and yank or not populate the zener. At least he's got everything coming in thru the Schottky diode, so it's a good point to add in the 4.7, but the negative thing is everything is after the diode, effecting the voltage divider and values used for voltage monitoring.
Yeah, I also have a few dozen blank boards laying around here as well that will most likely never see a light. On the cheap boards that is not a big deal but if I am spending the money for a SRK sized driver I better be using it lol.
That is why I decided to come out with the Avenger drivers, I wanted one driver to do it all so I never again had to worry about it being outdated or not working for a particular setup.
I think they have turned out pretty good, they have all the hardware options I can think of (except LDO, still have plans to give that a shot on the larger boards but won’t fit on the smaller ones, well maybe it will….) and the firmware is top notch.
Bistro is coming along nicely, when TK gets done compacting it for more space, and hopefully adding the internal voltage reference code as an optional define, I should be able to fit all the features on it that 95% of people would need. Thus negating the need for people to have to do anything more then flash the hex file.
I love this as well because I can build drivers in batches, flash them and know that they all have the same firmware / options making it easy to keep track of exactly what the mode group order is, hidden modes ect.
Now that you have narsil working on a triple channel driver we have the ultimate e-switch firmware as well, (I still want to play with narsil but just don’t have a light to use it in).
The combo makes for a potent setup IMHO. I honestly don’t see linear / 7135 based drivers being able to get much better then a PWM controlled triple channel design. A quad channel would be interesting and I have considered it, it is possible if the internal voltage ref is used.
The issue here is it means that a zener can’t be used and honestly after playing with the drivers I don’t think a 4th channel would add anything to the performance. The spacing between 1x 7135 and the bank is already about perfect IMHO. You can simply leave off a few 7135’s to adjust it if desired. From the bank to turbo is if anything too close to start with, so adding another channel in there would be pointless.
So in the end it is exactly like the title, one driver for all your flashlights. Standardization FTW.
Well at least it has standardized my collected.
If either of you guys wants to get rid of some of your ‘extra’ PCB collection, I’d be glad to take some off your hands!
I have so many old Illuminated Tailcap boards it isn’t even funny. At $0.30 to ~$0.80 a board, I hate to think how much cash I have sitting there with no future.
You pay the postage and there are a bunch of outdated, almost worthless boards coming your way
Actually for me, a favorite fixed mode setup is 5 modes including moon, and for a FET+1, that means only mode #4 is PWM'ed on the FET in normal 1 LED lights. Only for 7+ modes would I see more than 1 mode using PWM's on the FET. For smooth ramping, it's a different story.
So for a triple, only 1 of the 5 modes actually uses the bank of 7135's - yes, it makes that mode more efficient, and could be argued that might very well be the mode most likely used extensively - full FET power typically won't be used long, while dropping back one level is the mode to use long term.
Actually for e-switch lights, I prefer 5 modes, while on clicky lights, I prefer 4 modes w/moon, but same issue for 4 modes - next to last mode is for the 7135 bank.
For me it’s
Moon
Low (50ma)
Med (3*7135 for single emitter, 4*7135 for Triple)
Turbo
It’s purposefully a bigger jump from Low to Med. But I don’t expect that to be an option, I’ll still mess with my own settings
Sounds like you like the same modes I do.
Mine are generally setup as moon, ~10-15 lumens, 1x 7135, all 7135’s (adjusting the number to reach the output I want), then Turbo.
I find this provides a good balance of modes with good spacing. Although I am more and more swapping to only having turbo in the hidden modes, I just don’t use it very much and like to know when I am. I like this because only turbo is not regulated.
The 3rd channel could indeed be argued but thats what I like about the drivers, if you don’t want to use the 3rd channel, then simply leave the 7135’s off the driver and use as a normal 1 sided FET+1. Personally I feel that the bank of 7135’s is the most useful mode and gives me the best of all worlds. But we each have our own oppinions.
The goal with these drivers is to let everyone have their own opinion and setup without having to change the hardware. Ideally the firmware will have all the modes built in to not need changes to that either.
Which gives me an idea for another mode group, one based on a FET+1 setup that skips over the bank of 7135’s modes.
Thats actually not a bad option for a mode group, basically just skip the 1x 7135 channel and go right to the bank then let people adjust the current by adjusting the number of 7135’s. Simple enough.
I think I kinda settled on ~38ma as my personal preference low mode. usually nets around 10-15 lumens depending on LED used and I like it since it would last ~100 hours on a fully charged battery. Not blinding in pitch black but also useable in low light.
Actually the more I think bout it, the bank being used for your 2nd highest, or Hi vs. turbo, whatever you want to call it, suits it well - it should be often used. No matter what light you have, a full FET just won't last long because of heat, battery drain, etc., so the 2nd highest could be the most popular. When I'm walking trails at night, I like just enough light to see where I'm going, with some side spill, but want quick access to hi/turbo if I hear something or spot something in the distance. Someonein the BLF Q8 thread said the same thing, and that's why I added the double-click to turbo from ramping, then click OFF, click ON returns you to your previous setting.
I’m still not super fond of the table design in Bistro. I like the set-up or the earlier firmwares better, though I know it’s more difficult with so many mode groups. For instance if I don’t want moon enabled on a light, I want my low mode to be more like 20ma than 50ma. Also when I’m using a special LED (like a red auxillary), I want to use that sometimes for moon but not always. Also I like to have a red moon and a red “high”, but where do I put that on the ramp table? I don’t want it to suddenly pop in if the light is in LVP ramp down or something.
Sorry, this is really a discussion for the FW thread.
Yep, I love the high regulated mode with the 7135’s. I find I hardly ever need to go to turbo except for a few seconds at a time. it stays WAY cooler as well since it is pulling less then half the amps but yet gives me most of the light output. All while still having turbo at the ready should I desire it.
Like I said, the idea is flexability, anyone can use them to do anything they want (well almost).
It is really not that bad once you get the hang of it. I honestly don’t really care if it used “modes” or raw PWM values, they are the same thing, modes are just a shortcut to the PWM.
Far as your single LED complaints, they are actually quite simple to deal with, simply setup multiple mode groups with all the options you want. Thats why I started the mode group thread. For example I already have groups setup like you describe both with moon mode and a mid mode and 2 low modes without moon, all 3 ect ect. Looking for more options as well.
The nice thing about having a bunch of mode group options is that they don’t do anything if you are not using them but they are there if you do want them. The key is having a “standard” firmware package so you can keep track of them.
For the dual LED light you will have to use workaround since no firmware is made specifically for dual LED’s. to keep LVP from ramping through a set of modes (not sure how it would be any different in the A6 firmware?), simply put the LED you don’t want it to use at the end of the ramp table. It will start LVP there if you are in it but if not it will not go up there.
For the single LED, yeah it’s easy. I’ll just move the “moon” toggle to the end of the toggle list for mine if I feel like it and make more groups.
For the dual-led, It’s like that already in the A6 firmware, I was just hoping I could make it better in that regard with bistro.
For your goal of “standardizing” though, bistro is ideal.
I added both of ya’lls mode ideas to my ongoing mode group table. Up to 23 modes and climbing. I think I will fill up any leftovers with some slight variations on popular groups for those picky folks among us 