Ultimate driver board.

PeterScowcroft

Newly Enlightened
Joined
Oct 1, 2006
Messages
73
Hi,

Not sure if this is in the right section so please feel free to budge.

My friend has been nice and gone to the effort of designing me a driver board for my light.

The problem with most current driver boards (sorry bad pun) is that they are aimed at the current tech. My n-flex which was spot on a few months back now can't provide the current I need for my design.

The board that has been prepared is specifically over eginered so that it not only works for todays LED's but anything within reason that the market can throw at it.

So basically it is a buck/boost driver with 90%+ efficency that can regulate current upto 4-5A The voltage in can be anything from a couple of AA to a few lead acids.

Covering all bases basically.

Now as you probably have guessed this board costs quite a bit to assemble.


So I thought you guy could help me with a bit of research.


How many people would be interested in this driver and what is the maximum you would pay?

If there are enough people with enough cash, we might put into production, if not just one for me and one for him.

The board costs less the more that sign up and is fully reprogramable for each new led designed.

at the moment it costs more to make than most of the competition cost to buy so this wouldn't be cheap, but hopefully the best in the market for years!

Now picture 5 sets of five P4s in parralel running from a few Sub-C cells.
Thats a good 25 x 240lm.


So,

Who would want one, and how much could they reasonably afford?
We are not looking to make much of a profit but the board is expensive to make (for obvious reasons)
 
We would like more info.

From what I can make out
It can take a voltages from 1.5v (AAcell) up till 12 volts(lead acid)
Current up to 5amps ..

Sounds good already

Some questions
Can the output current be controlled ?
Any other features like short current ,open circuit , size of the board

Is a picture possible of the board that you already have ?

Price would not be a problem if you can bring out aboard that can do the function of
a couple of boards.

More information needed please ..
 
Hi,

The basic circuit has been designed.
I know the output is designed to be modified by either (or both) a pot )for modding oldham lights) or a push switch digital thingy (for my light).

The board is designed to fit inside a petzl duo so will be quite small, probably rectangular.

I do not know the exact details of the board since the redesign was done last week and I am flued up!

The final design has not been put into production yet so if you can think of any extra features put in; now is the time!

So what I am looking for is:

If it can do this:
list list list

Then I would pay up to £xx

it's to see if this will be commercially viable.
If 100 say yes could be quite cheap, if 10, probably not.
 
I would be interested if you would do this. I prefer larger clusters of LEDs for extreme brightness, so this driver would be perfect for me. I have plans for a second LED headlight project (or LED foglights if my friend changes his mind) using the new SSC P4s (about 10 will give you the same light as higher wattage after-market halogen headlight bulbs), so that would be great. The current drivers for higher power use nowadays are the Fatman by George (boost driver -up to about 20W on the output as long as the input stays under 2A), the Shark driver (more powerful boost driver, but seems to be funky to use -never touched it yet), and Andrewwynn's Hotdriver circuit (12V very high wattage FET controlled linear voltage regulator -originally to power high wattage bulbs in bigger flashlights from high power battery packs). As you can see, there is nothing like what you have. You have something unique! I am using that Hotdriver in a dual 18 K2 headlight cluster setup and works great, although the regulator, as well as the K2's being used are pretty inefficient. The Fatman is great for Quad LED flashlight mods (like 4 SSC P4s powered at 1A each), but other than that, its uses are limited.

So will this driver circuit boost and buck the voltage, and do so automatically, depending on input voltage? I always hear about buck/boost drivers being slightly less efficient than just boost or buck drivers. As long as this thing doesnt go into direct drive mode, unless that is desired for a certain application. Will the current be fully regulated and controllable (via a potentiometer)? What is the max voltage input and output it can handle? So how many watts can this beast handle as a load? Will there be a max input current, besides what the battery can dish out?

This does sound amazing -the "Holy Grail" of LED drivers. Small-ish size, mucho power, and full-o-features. Awesome!!! As for price, $50 (~25 GBP or 37 Euros) seems to be the max that I may pay. Notice I said maybe. I may pay more if this is for real. I understand that development costs and parts and etc will be expensive, but anything too expensive would be insane. I would rather research and build my own.

So, I suppose, full current control (via a fixed resistor or a potentiometer) would be nice. Boost capabilities so at least 4 series mounted higher wattage LEDs can be used, powered from a small-ish battery pack (NiMH). The ability to work as a buck driver so one can power something like a few 5W LEDs in parallel with a car's 12V system (an odd scenario). The ability to efficiently power those 25 P4s from 12V with no problem. You covered most of this, but I was just saying what would be ideal. Oh, make sure the max input voltage is pretty high, in case of slight variations and spikes from the power source (some external suppressing TVS diodes will help protect the circuit). Automotive power can be nasty, but several freshly charged charged li-ions can be too. I dot know.

How about a user settable low voltage cutout option? Maybe it can be set using a calculated set of resistors or something. This would be good to protect you rechargable NiMH and Li-ion batteries, as well as prevent killing your lead acid car battery. A simple formula can be used maybe to find this low-voltage cut-off point, so any type of batteries can be used at different quantities (2 Li-ion or 4 NiMH for example).

This is great. I hope others are interested. If this thing does not sell, I will have to keep in touch with you anyway. Maybe you can build a few even if few are interested or give instructions on how to do this yourself (how about a kit?). Please keep us updated! Have you talked to George from TaskLED or Wayne from YC and/or the Sadwich Shoppe? How about Newbie or Andrewwynn? They are all pretty knowledgeable in the area of different driver circuits for LEDs and incandescent. Maybe they can give you extra tips or maybe you can school them (you never know). Good luck! Sorry about the long post.


-Tony
 
Last edited:
The info so far is a little vague, could u post some info to the following:

1. Input voltage/current to output voltage/current chart.

2. More accurate technical specifications, such as size, adjustment settings and user interface.

3. Show a few practical applications or complete setups for this board at work.
 
Thanks Gryloc, thats a lot of good info.
I am hoping it has open short circuit protection and a low voltage cut off too.
Mine will be for a caving light so the light will be taking repeated hard impacts potentially be run underwater and I need protection againt low voltage.

George's circuit is great, just wish he could fix the max flex problem! I am impressed that his crcuit runs fine underwater ( my torch leaked and filled with water, it still ran fine!

I do not know the exact details on the board as the designer made some changes a few days back and I have been sick since then.

The board on a small order will cost £23 a piece to get the board made, we then have to do alterations and set the spec of the order to client so lets hope more interest turns up!
 
Last edited:
I am looking for a driver that takes 2 x NIMH has input and will drive a Seoul SSC P4 at 1 amp. I had assumed that the Sandwich Shoppe Nexgen or GD could handle this but I am told they are limited to 1.5 amps on the input. Another CPF'r has a board that will do the job - but he wants $40 for it and I found that to be too steep. Also I need a round board less than 26 mm in diameter (fits inside a mag-c body).

My application is a 2-C mag, 2 NIMH, aluminum heatsink (LITEmania and soon to be others), Seoul SSC P4. My guess is that this would be a very popular configuration and value priced. If you could pull together a board for this it should be very popular. I would be interested.

THanks ...
 
Any progress, Peter? This sounds really interesting and you have something that could be a hit. Many people may be interested! Have you thought about starting a poll and putting it in the LED section or some other section of these forums? I bet that this general flashlight section is pretty slow and you may get more desirable results in a different CPF section. Please keep us updated. That would be awesome. Good luck and Thanks...

-Tony
 
I'd be interested, though I personally wouldn't be willing to pay too much - nothing to do with the board in particular, I'm just not the type that likes to invest too much into a flashlight. And I've only got very limited experience in modding lights, so I wouldn't want to go too overboard on making a modded light if I'm not certain that the finished product would work properly :p . Some people have no problem spending $300 on a good light, I'm more in the $30 range :D . So realistically I can't see myself spending more than about $20 or maybe slightly more on a board, this would certainly be interesting to me though.
 
No offense, but we see a lot of these "theoretical" things that have been built up and someone is looking for feedback, but for some reason no pictures can be given, etc.

Having done enough power supply design in my life, I have a good idea that something like this is:

- Not trivial to design
- Will not be that small
- Will not achieve 90%+ efficiency under all conditions (and if it is a true buck-boost or SEPIC), may only work that way in very specific conditions

So, before I provide valueable marketing information, I think some proof that this actually exists since the claim has been made that version 1 exists is not out of order.

Semiman
 
Top