Bike UI on *Flex drivers

[Tidra]

Until recently I was using only the old version of MaxFlex without all that voltage and thermal protections, but now I am almost happy with all that new options.

For me steps between levels are O.K., but I would really like to have a low voltage dimming option, so that Vlow would not blink but just step down and then of course on Vcut-off :poof: :candle::candle::candle:

:grin2::grin2::grin2::grin2::grin2:

Because basically that is what you do when the Vlow occurs and you don’t have any spare battery, you dim the light and that step you repeat until you are on L1 and then you pray that the Vcut-off will not be reached.

I realize that this is complex but would be very useful too. And of course it should be recognizable so you could know if there is a voltage or heat problem…. And yes we are back on the beginning… so maybe the way it is, is not so bad…

:grin2::grin2::grin2::grin2:

of course IMHO,...

LP
Iztok

 

[bbgobie]

Re: Bright as possible without killing the battery

A trail rider on single track doesn't really want noticeable levels while riding except to try to get the most light on the ground that won't kill the battery before you get back. You crank it to max for a minute at a time on the fastest down hill sections(3 cree Q5 at 1000ma. and 7* lenses on the helmet is barely enough for fast sections and bad weather. Still slows lap times at night by 10%), and the rest of the time you try to find a setting to give the most light without killing the battery too fast. You can scrimp by going to 250ma on an easier road section if there are any but most of the time you want the "just right" run time setting which for me would be somewhere between 400-700ma. depending on what I am doing.

If you can't notice a difference in light level, there is no point to be higher than is there?

Having the light levels change so you notice them mean your on high when you need it, med when you don't, and low when you really don't.
Otherwise they become kinda arbitrary. I'm gonna set it on MED even though I don't see the diff between med and high just because it gives me more battery life. Now I'm going fast so I'm going to set it on high even though I can' t see any differently... :shrug:

 

[chris_m]

Just as a data point for you George, I should mention that my custom firmware has exactly (in software terms) doubling of current between each level - though only 4 levels. Given the fact that doubling the current only increases light output by ~60% and the logarithmic nature of vision, this seems a reasonable minimum step size - noticeable but not excessive. I reckon that if I can get away with less light I can might as well double my run time for the difference that makes to output. As Ifor says, the remaining runtime calculation is easier that way.

Ifor - if you're interested I can give you some help with custom firmware (don't currently have temp. sensing implemented, but it would be relatively straightforward to add). I guess if you're integrating with a computer you can do proper runtime display - though my firmware has an "analogue" battery level display via a 3mm LED on the STAT pin.

 

[sendler]

7.2V

I'm a trail rider who rides very technical singletrack and I DO want noticeable levels!!! You need a bigger battery if you are really having runtime issues. Most of our rides are 3 to 4 hours in the dark and my tiny 7.4V 4800mAh Li-Ion battery makes the duration (my helmet light is the same as yours... 3x Q5). I use this setup all the time for riding and racing endurance events with no runtime issues.
7.2V x 2500mah nimh. 218 grams. I will go to 9.6v 2500 for the handlebars when I get my next set built. Waiting for the excellent drivers that make these projects possible. Thanks George! without them we would be stuck paying $350.00 for a commercial light and $150.00 for each spare battery. though with lights on helmet and bars 500ma each should be pretty nice.

 

[sendler]

Wish List

I can see I am totally out voted (you guys must all use giant and expensive batteries) so I will just post my wish list for the record so we can wrap this up. First of all, this is a great chip as it is and shows how some personal inginuity and the internet can improve the quality of our lives. We live in a great time. Thanks George!
On to the wish list. No tables or computations beyond what is already in the defaults, just add user selctable max current in steps of 100ma., user selctable current for the levels in steps of 100ma. We are allowed to pick our own warning voltages so why not the current steps? One other unrelated idea I had was changing the outcome of another click when already on level 5 from a flash to a return to level 4. This would be just as positive an indication that you were already on 5 as the flash, and, would be like running on duo mode and multi mode at the same time. Still, a remarkable product as it is. Thanks Again.

 

[StevelKnievel]

Re: Wish List

Here's a good cheap battery/charger combo:

http://www.batteryspace.com/index.asp?PageAction=VIEWPROD&ProdID=1256

Steve

 

[ifor powell]

Ifor - if you're interested I can give you some help with custom firmware (don't currently have temp. sensing implemented, but it would be relatively straightforward to add). I guess if you're integrating with a computer you can do proper runtime display - though my firmware has an "analogue" battery level display via a 3mm LED on the STAT pin.

Thanks for the offer but I think I am getting there myself. I have resurected a bflex that I broke a year back and have the a reed switch actuated off a wheel magnet wired into what is normaly the status led output. I have it basicaly working but am strugeling a bit fine tuning the mapping from speed to pwm output to my liking. Once I get the bike computer on the road I will be able to display the output level on the lcd and hence have a better feel for things.

Ifor

 

[heatstroke]

George,, the only reason I want more steps related to power is so that I can work out quickly - what life I have left in my batteries as I toggle between high and low.
Right now I hook up a MM to figure out my power draw from the batteries at differnt level and set the low and high based on my desired run time.
I suppose mtb bikers (me) are a bit more concerned about the batt pack weights ?
Anyway- I agree completly with Sendler, if we could program in the mili amps for L & H, I'd be happy too.

 

[Tidra]

Sorry George,

Recently I found out, that it is possible to drive 3xMC-E @ 600mA with MaxFlex, but there is always a but,… If I set it on the 1,2A max, on the 5 level, the first level is O.K. but the second is too high (500lum).

What if the current level settings could be set like:
1.) for every level inside the max output settings, you could set every digit like for mid, low and low cutoff voltage settings, except maximum output,

or

2.) again for every level inside maximum output settings, you could set levels with step 100mA, except minimum and maximum output.

I know there would be quite a lot of clicking to set, but it would be totally self adjustable.
Of course this is more likely to be used in brighter LED setups; because you still need some low settings in case of battery saving and emergency light option.

And no doubt, for me this is the No1 driver in the world, this idea is just suggestion how to make it even better, but of course it is only IMHO.

Than you, George for your effort.

PS: I would really like to see you start making some drivers for P7…

LP
Iztok

 

[georges80]

Sorry George,

Recently I found out, that it is possible to drive 3xMC-E @ 600mA with MaxFlex, but there is always a but,… If I set it on the 1,2A max, on the 5 level, the first level is O.K. but the second is too high (500lum).

What if the current level settings could be set like:
1.) for every level inside the max output settings, you could set every digit like for mid, low and low cutoff voltage settings, except maximum output,

or

2.) again for every level inside maximum output settings, you could set levels with step 100mA, except minimum and maximum output.

I know there would be quite a lot of clicking to set, but it would be totally self adjustable.
Of course this is more likely to be used in brighter LED setups; because you still need some low settings in case of battery saving and emergency light option.

And no doubt, for me this is the No1 driver in the world, this idea is just suggestion how to make it even better, but of course it is only IMHO.

Than you, George for your effort.

PS: I would really like to see you start making some drivers for P7…

LP
Iztok

I'll think about how a menu system for setting specific current levels would work AND be easy to use. I'd add it as a separate menu item for folk that feel competent enough to configure current levels.

P7 eh, only one? Hmm...

I'll let the pictures talk (3 P7's in series running here), been going over 30 minutes and all is happy...

Take a look at the grey stuff under the PCB - neato heat conducting pad.

1st pic - too bright :)

2.8A and what have we here, a momentary action switch going to the PCB, hmm maybe some kind of *Flex driver eh?

Power supply input to driver. On quick calcs, 2.8A to 3 P7's and it's running over 92% efficiency, not 1/2 bad I'd say.

cheers,
george.

 

[Hack On Wheels]

I'll think about how a menu system for setting specific current levels would work AND be easy to use. I'd add it as a separate menu item for folk that feel competent enough to configure current levels.

P7 eh, only one? Hmm...

I'll let the pictures talk (3 P7's in series running here), been going over 30 minutes and all is happy...

Take a look at the grey stuff under the PCB - neato heat conducting pad.

1st pic - too bright :)

*snip*

2.8A and what have we here, a momentary action switch going to the PCB, hmm maybe some kind of *Flex driver eh?

*snip*


Power supply input to driver. On quick calcs, 2.8A to 3 P7's and it's running over 92% efficiency, not 1/2 bad I'd say.
*snip*

cheers,
george.

:drool:Wow! Maybe I will end up going for MCEs setup up with parallel dies... Exciting stuff, that looks great. Any more teaser info? :D

 

[wquiles]

I'll think about how a menu system for setting specific current levels would work AND be easy to use. I'd add it as a separate menu item for folk that feel competent enough to configure current levels.

P7 eh, only one? Hmm...

I'll let the pictures talk (3 P7's in series running here), been going over 30 minutes and all is happy...

Take a look at the grey stuff under the PCB - neato heat conducting pad.

1st pic - too bright :)

2.8A and what have we here, a momentary action switch going to the PCB, hmm maybe some kind of *Flex driver eh?

Power supply input to driver. On quick calcs, 2.8A to 3 P7's and it's running over 92% efficiency, not 1/2 bad I'd say.

cheers,
george.

That is awesome George!

I have been turned off (pun intended!) by the various LED "drivers" for the P7 (and multiple P7s) that can only do about 80-85% efficiency (some drivers even less than 80% efficient!). To me I rather have a DD with a D2Flex than dropping that much power on the regulator/driver. Seeing something at about 92% efficiency is definitely a great sign of things to come ;)

Will

 

[Darell]

It sounds like we have a bit of confusion between using UIB1 and UIB2. But some good input nonetheless.

I can see I am totally out voted (you guys must all use giant and expensive batteries)

I'm still a bit confused as to why half the current draw between steps requires bigger batteries. The fact of the matter is, I use a very small battery, and can control it competently with the known step sizes of roughly 1/2 current (or double depending on which way you go!).

just add user selctable max current in steps of 100ma.

I personally think that 100 mA is just too fine of a filter. There's no output difference for all intents and purposes. Certainly not up at the 1A area. As George will tell you, I've gotten confused more than once in thinking that a 500 mA led should look considerably brighter than one run at 350 mA. But practical demonstrations reveal that the difference is miniscule.

another click when already on level 5 from a flash to a return to level 4. This would be just as positive an indication that you were already on 5 as the flash, and, would be like running on duo mode and multi mode at the same time.

We considered several indicator options for hitting the L5 wall. The reason to NOT bounce back down to L4 is that in this case to be SURE you are on L5, you would click, click, click until it dims once, and then click again to get back to L5. When I want high, I want high, and want it to stay there. I don't even have to look for the indicator - just click a bunch of times, and I KNOW it'll be on L5.

So finally... if anybody wants my input (a bit biased, maybe?) I think the five levels are just what they should be, and have great steps between them. Reading the comments here, some want just TWO or THREE levels, and some say the steps between FIVE levels are too great already. So from those two types of comments, I think we have the best compromise. I love the way this thing works (who's surprised? Raise your hand). And I know I can significantly extend my battery life by just notching down one level. I dread the idea of adding yet another menu setting, and see it as a solution looking for a problem. My 2c.

 

[georges80]

Thanks for stepping in Darell with your take on how the current tables are set up at this time.

I would like to leave things as they are unless there was a compelling reason to change.

I can see the possibe need for specifying drive current in higher power drivers. For bFlex/nFlex and maxFlex3 I think at most it would just require a tweak of L4 to L5, though it is also very nice to know that you're effectively doubling your runtime by dropping down one level.

Anyhow, we'll see how things evolve.

cheers,
george.

 

[Tidra]

George,

nice work, observe my picture,... I use it on all my MaxFlex setup... but my suppliers are from old PC – CD/DVDs,... where can I buy this stuff???

 

[Darell]

Thanks for stepping in Darell with your take on how the current tables are set up at this time.

I hope my message got across, because my writing was terrible! I tuned it up a little bit to make more sense, but having a drink first will still help.

 

[Tidra]

I personally think that 100 mA is just too fine of a filter. There's no output difference for all intents and purposes. Certainly not up at the 1A area. As George will tell you, I've gotten confused more than once in thinking that a 500 mA led should look considerably brighter than one run at 350 mA. But practical demonstrations reveal that the difference is miniscule.

100mA step was mentioned because of L2 could be too high @1,2A settings with 3x MC-E (6s2p), for example I don't need 500lum for uphill riding and could use 250-300lum and save battery,...

IMHO

 

[StevelKnievel]

Darell, you hit the nail on the head!

-and-

George, that prototype driver (uberFlex?)... MWAHAHAHA! :rock: Excellent!!!

 

[znomit]

100mA step was mentioned because of L2 could be too high @1,2A settings with 3x MC-E (6s2p), for example I don't need 500lum for uphill riding and could use 250300lum and save battery,...

IMHO

Turn it down to 750mA max. You'll still be rocking more light than anyone else(~1400lm?) and your batteries will last much longer.