I’am with flashpilot on this. I know you can’t possible implement every feature everyone wants but this is a feature that would add to user configurable operation of the driver. If you where using a newer Panasonic NCR18650a or b the cut off voltage could be set at 2.5v. With the amount of current draw this driver will be asking for from a single li-ion cell, most cells will hit 3v within a short amount of time (battery sag). Lowering that cut off point would allow for a longer run time without harming the cell. The voltage should jump back up quite a bit at that current, when measured at rest once the cut off has been tripped.
My 2 cents would still go to the attiny. The sense resistor mod also makes this driver user configurable, big +. Flashaholics cant leave things alone you know. 
Sounds like this driver could change the single cell driver world forever. No other driver would be so user configurable to meet so many applications.
Thanks relic38 for doing this.
Yes… gotcha…
thanks for explaining!
You guys have pretty much nailed what I want to do. I’m also hashing out the UI, specifically, how to do this all from one reverse-clicky without becoming overly complicated.
Also, the four modes is an example. The number of modes will be customizable too; from one to say, six, seven? Not everyone wants four modes. I personally need three (100, 30, 0.1), so having to go through a fourth duplicate mode would be frustrating. Single mode users would be frustrated programming four modes to the same value.
Also, in case it wasn’t clear, the modes can be reprogrammed whenever you like. They are not permanent after the first config you do. Now you want a lower medium? Just reprogram it.
I will post some more UI details once I refine it a bit.
I need to touch base with Bill too so he’s kept in the loop. I don’t want to design something he cannot build (and I will not have the time to do myself).
1 Thing I thought I might mention, if it’s not to difficult would be to make it compatible with an electronic switch.
This custom driver have the potential to become really awesome!
I love the idea of having the ability to change the amount of modes.
I just hope its not going to be noisy like a typical nanjg 105c is on medium mode (30%). Full of pwm whine (I believe that is what it is called).
So, what is the status of this board? When will it be available for purchase? I will be needing one shortly, maybe several if this one becomes available :bigsmile:
Edit: can we please put the status of the project in the first post?
I’d like to be on the list as well. Thanks for all the hard work.
No response from Bill last night. I have several hardware and firmware questions for him. I hope to reach him soon to discuss the schedule as well.
Audible whine from the PWM is due to the high frequency current fluctuation in components that have the potential to vibrate. This includes the switch, pill-to-head contact, and driver-to-pill contact. Picking a PWM frequency that is high enough to not be annoying (4kHz happens to be a particularly annoying frequency for humans; it’s right in our most sensitive range), but not so high that circuit efficiency is compromised.
Strangely, I notice that NANJG 105C has some whine even on High. This surprised me because at 100% ouptut no PWM is needed. There must be a tiny amount of switching going on still.
Regarding the electronic switch, this will depend on available I/O on the MCU if I can get Bill to agree to a 10-pin MCU then this is feasible. On an 8-pin MCU it may be possible as well. I had thought about adding this in, but I didn’t want to make promises on something that might not fit.
What a great idea Relic! The flexibility of this driver is starting to seem endless and should appeal to everyone.
Did you happen to read post 91? I’ll be happy to help out anywhere I can & discuss with you via PM/email/phone.
Hi FlashPilot, I did see that, and I’m putting it all together to try and come up with a good way to configure the config UI without being complicated.
Still open to ideas, but i definitely want to keep it flexible. I’m set on the ramping config, variable mode count, adjustable cutoff voltage, two+ current sens resistors (ideally one populated, one+ empty for easy modding).
I have some questions for everyone. I thought of using a poll or survey(better) to do that. It would be much easier to track than asking in here. Once I set it up, I’ll post a link for everyone to fill out.
Do y’all know the “Flupic” board and firmware?
It’s no longer available — guy who made them retired a while back — but a wonderful driver of some years ago.
Google “flupic firmware” and look for the PDF for V2.2, the last one I think — showing the modes and settings.
The ideas here are similar — and the presentation of how to navigate could be done like those, which were/are very helpful.
quick link to flupic
Introducing FluPIC V2.2 firmware - User Guide UP! | Candle Power Flashlight Forum!
its cpf, I shivered…
Thanks Relic. Maybe you can post your UI, or I can make suggestions for entering initial program mode… its up to you. Or feel free to PM me if you want.
Im looking forward to your poll. But just so we dont miss any criteria, can you post your subjects?
All in all, the UI ideas sound intuitive and easily described in a youtube vid of the finished product; for those unclear about the operation. But above all else, just having any good 4A boost driver with high PWM frequency solves a huge amount of shortcomings.
These questions may be premature but I’ll go ahead and post anyway.
Will the driver continue to still boost effectively while running at anything less than 100%?
If so, why would we need resistor mods unless it is to exceed 4A?
Hi FlashPilot, I didn’t get alot of time on this tonight. However, I can answer some questions. I hope to get more time on the weekend for this and get the UI nailed down. I want to have the survey up and running for tomorrow too.
The main subjects of the survey will be regarding the UI. I want to gather as much input as I can using questions that tailor to multiple choice answers. There will also be an open entry section to add stuff that users feel was missed. Examples are, how many quick-presses to enter programming. Max. number of modes, timeouts. There will also be hardware related ones like max current you would need, soft switch option, gang option, etc. Some are y/n, some are multiple choice.
Regarding the dimming, I’m considering a different option to handle the dimming. If my theory is correct, we may be able to dim the output by adjusting the feedback voltage (or a reference voltage if available) into the boost controller. This would allow ‘linear’ dimming with no PWM on the output. It’s a common way to control typical Switch-mode voltage regulator ICs. I think it will also work for this type of IC (this is why I need to talk to Bill, badly). there may still be a need for PWM down at the moonlight levels, but for anything down to 20% or so, no PWM means safe use around motors and fans. Fingers crossed everyone 
The resistor mod would be there to raise or lower the max output current. Mostly to raise it. If we can do what I suggest above, there may not be a need for resistor mods. Just adjust the max output by adjusting the reference voltage. Go as high as you want until the boost inductor saturates (we’ll probably need to add a safe limit here).
Don’t get too excited about this just yet; I need to discuss this with Bill. He may know of a reason to not do this (I’m a software guy, so hardware is not my strongest suit).
Hiya Relic,
The creativity in problem solving and innovation just keeps pouring in. While its a great idea to get as much as we can out of this driver, I think we should pause for a moment and consider who the mediator, suppliers, engineers and manufacturers might be. We certainly dont want to overwhelm them and have them tell us to get lost. When you get right down to it, a simple 3 mode H/M/L driver that can effectively and reliably boost at 4A on a single cell would be very much appreciated by most and still sell like hotcakes if priced reasonably. With that thought in mind, would we do better to have them create the hardware and supply the package with a recommendation for an easy to use USB programmer to modify the firmware later? At this point, focus would shift specifically towards hardware and its ability to do what we want of it - well into the future, with easy access to a clip-on a programmer in such a way that anyone can do it. The How To Build a Flashlight With Perfect Modes thread has spawned many of its own interests.
Here’s why Ive mentioned this sudden change in direction: I truly believe that with all the options that we desire and with the inherent level of technical difficulty, there will be no way of supplying a stable firmware release within the first few revisions. It will most likely be a long evolutionary process of firmware tweaking to come up with an excellent product. At that point, who owns the source code and who has the knowledge, time and desire (passion) to keep tweaking out the bugs and making the code available to the masses? Dr. Jones has been very generous in offering us his time but I havent seen him comment for a while in this thread. Id really appreciate his expert opinions in all of this.
We are a group of perfectionists and everyone has an opinion. Its just the nature of the hobby and of those that revel in it. From previous experiences of observing the group, I feel strongly that many will be very dissatisfied with whatever solution is provided when the inevitable firmware glitch shows up. Most of us are used to some bashing and negative comments but this might quickly offend the provider and make him run. If we kept firmware development “in-house”, there would be more than enough collaboration and knowledge to get the most out of this driver. In all likelihood, it seems logical that there would be several firmware versions available for specific purposes to satisfy the group. A great how-to guide in doing some of our own rudimentary programming to effect our own changes would also be helpful.
You’re the liaison and will have the best knowledge of what Bill might be able to have accomplished. A possible scenario might be that you find out how far he can take hardware/firmware development and then provide you with a copy of the firmware source code so that we can continue the development. We and they would have access to the latest firmware releases, which guarantees continued interest and sales of this driver. After all, we are the enthusiasts, they are the sales people. We both win… or we both lose.
What are your thought? Anyone else also please comment…
Dr Jones, PM sent.
I agree with this, in principle anyway. I didn’t really expect Bill’s team to implement the entire firmware functionality. Hopefule might have been a better way of describing it. I would like that to be the case, and if he can that’s great.
With that out in the open, it is likely that this will be an open source, collaborative project involving some BLF users that have code writing experience. This can be managed online quite easily (Google Code, Github, etc.). The survey will include design abilities as part of the questions. I know of a couple of members that have experience with coding.
Because of this, I need to ensure that we select an MCU that works for us. A lot of members probably have a preference (more survey questions), and I’m sure my preference does not align with any of them (I’ll explain this more some day; in short, I’m biased because of my work experience). When Bill signs back on we’ll iron this out.
Regarding a right-out-of-the-shute driver, getting the hardware right is the primary objective. Basic firmware that gets the board up and running (i.e. four fixed modes, low voltage cutoff) is all I will aim for to start.
In the meantime, the survey will proceed to gather ideas for the advancement of the driver.
Bill, I hope we (I) haven’t scared you away. There have been no big changes that will make this more difficult for your team. When you are back online, we can finalize the hardware and move forward with the basic version. I look forward to hearing from you.
Sounds perfectly logical and reasonable to me.
Up to this point, it sounds like all the bases have been thoroughly covered. Well done! It will be interesting to see what the MCU preferences are after the survey and the reasoning behind them.
+1 :)
Hello Mr. Zang,
Again, thank you for joining the BLF forum. We are all still very excited to buy the 4A boost driver from LightMalls when it becomes available. Will you please respond to Relic38’s private messages? He could also use your email address for better communication. Please feel free to ask any questions of us. As relic mentioned, the simple UI will do for now.
Mr. Zang, thank you for reading. Please come back to visit us soon.
Edit: PM sent