UIB2 and maxFlex

[georges80]

Now that I have the maxFlex bug fixed and maxFlex will soon be shipping (this week) with UIF/UIP/UIB firmware (NOT uib2) I have to decide whether the bike folk are interested in maxFlex.

UIF/UIP/UIB is the current nFlex/bFlex firmware version that has the Fixed lighting/Portable lighting and original Bike firmware (UIB).

For folk that don't know, maxFlex is a boost converter (steps up) and you can find technical details on my website (link below).

The question is whether there's enough interest by folk to have a boost converter (maxFlex) with UIB2 in it.

Porting the code should be relatively easy since the uController core interface of nFlex/bFlex and maxFlex are VERY similar. So the real issue is whether there's enough demand for me to a) do the port and b) have to stock a 2nd 'flavour' of board.

Now, before you all get excited thinking you can run that 5 Cree monster from 2 AA's - there are physical limitations (at least in this universe) about what configurations are realistic in a boost environment. You have to keep the input current to maxFlex at 2.2A or lower to retain reasonable efficiency. You will also need to heatsink maxFlex if you start pushing 8W+ into the load. Also, boost converters are harder on the batteries since they always draw more current than what is being provided to the load and as the battery voltage drops the current rises higher. Anyhow, I'm not a great believer in a boost converter for bike lighting where # of batteries isn't as big a deal as a handheld flashlight.

So - opinions/thoughts??

TIA!

cheers,
george.

 

[Darell]

So - opinions/thoughts??

I think you are a madman. (Do my thoughts help in any way?)

 

[georges80]

I think you are a madman. (Do my thoughts help in any way?)

Yeah - that helps a lot. Guess I'll wait for the weekday CPF folk, now that the peanut gallery has supplied their view

The silence so far is evidence that I shouldn't bother with a UIB2 port - which is fine by me, but I'll give it a few days...

george.

 

[dom]

Hi George
I'll have the UIB2 if you're making some with it
Be good for an helmet light with lighter batteries you can strap on the helmet.

Top job on fixing that

Cheers
Dom

 

[Darell]

You're right. Dom's comments were better than mine.

 

[Zero_Enigma]

I think it's a good idea as well for those DIY'ers that are thinking of making a backup light for that 'oh crap' situation that you can run off fewer AA cells to get you out of a situation.

I'm all for more choices and damn it George to killed my dream of 5 Cree Q5's @ 1000mA with a maxFlex running off 2xAA. :mecry::kiss:


Zero_Enigma

 

[SurfHenk]

Congratulations on debugging!
I am absolutely interested in one, in fact you can write down an order for one when available. I have other lighting systems (and therefore batteries) running on 7.2 volts so the maxflex with UIB2 would be the best driver for a DIY 3 or 4 LED bikelight I would like to build.

 

[znomit]

I think *maybe* I can see a use for this...
http://www.bikeforums.net/showthread.php?t=349977

 

[ktronik]

Hi George
I'll have the UIB2 if you're making some with it
Be good for an helmet light with lighter batteries you can strap on the helmet.

Top job on fixing that

Cheers
Dom

watt hrs are still watt hrs...weather you boost or stepdown. for the same runtime you need the same watt hrs... but by using boost converters you are always worse off...IMHO

if you wanted to use a smaller battery with a stepdown, you could use 4 cr123A or AA lithium... still the same weight as 2 18650's using a boost controller...

I would rather be taking less current from more batterys than more current from LESS batterys...voltage sag under load will play a bigger role when using boost controllers...

I can't see any GOOD reason to use boost for bike apps... happy to be proven wrong...& this is just IMHO...

but as dom says for less runtime you have a lighter system...


K

 

[Movemint]

I'm in to it, if only to not carry a 1lb battery pack in my jersey pocket any more.

Excuse my ignorance, but what kind of runtime what I get if I was to build my new light with 3 Cree XR-E Q5s and a Maxflex? I assume I would have a nice small battery pack, but what sort of batteries could I use to get a good 2 hour ride in? Is that even possible?

Again I don't know enough about the electronics to know what I would need to do to make my light build work with the Maxflex. Should I just another Bflex and run a big battery?

Jordan

 

[Doh!Nut]

I can see a use for it both in Bike lights and "sports lights" like diving where you are busy doing something else, a feature like dual mode where you just toggle between high and low (for example) would be usefull.

There is a argument that you can never have enough light in both of these sports and 5 or more emitters is a reasonable specificiation but this gives you a Vf of 18+volts.
I use 4Cell Li-Ion as you can get a cheap charger for it, a 4 cell pack is small and light and an 8 cell (2*4) pack lasts a really long time.

So you need a boost convertor:thinking:
I am using a shark, shortly to be converted to Hi/Low with a second switch.


I am currently designing a 6 or 7 7LED dive light based on my bike light, but in a waterproof housing. The shark would be borderline with the Vf for 6 LED and the MaxFlex wiould be on the limit with 7LED


http://www.candlepowerforums.com/vb/showthread.php?t=167735

having said all that my next bike light is a triple Seoul helmet light which will use a buck converter. But then again I have just calculated that two 18650s would power those three LED for around 2 hours with a boost :mecry:- I am going to lie down.

 

[georges80]

I can see a use for it both in Bike lights and "sports lights" like diving where you are busy doing something else, a feature like dual mode where you just toggle between high and low (for example) would be usefull.

There is a argument that you can never have enough light in both of these sports and 5 or more emitters is a reasonable specificiation but this gives you a Vf of 18+volts.
I use 4Cell Li-Ion as you can get a cheap charger for it, a 4 cell pack is small and light and an 8 cell (2*4) pack lasts a really long time.

So you need a boost convertor:thinking:

Ha - but there is a scheme with bFlex that I will be implementing in about 1 month. Already tested on the bench and it works great.

Master + slaves(s). You take say 2 bFlex. One you leave alone, the other you configure as a slave. You then run 2 control wires from between the master and the slave(s). Now you essentially have have 6 LEDs running from you 14.4V li-ion pack since you have 3 per bFlex. You could also have say 3 LEDs on one bFlex and 2 on the other - they are independently current regulating - all the master/slave does is fully sync them for parallel use.

So, you don't always need a boost Of course you would need to buy 2 bFlex instead of one maxFlex (or other boost)...

cheers,
george.

 

[georges80]

I'm in to it, if only to not carry a 1lb battery pack in my jersey pocket any more.

Excuse my ignorance, but what kind of runtime what I get if I was to build my new light with 3 Cree XR-E Q5s and a Maxflex? I assume I would have a nice small battery pack, but what sort of batteries could I use to get a good 2 hour ride in? Is that even possible?

Again I don't know enough about the electronics to know what I would need to do to make my light build work with the Maxflex. Should I just another Bflex and run a big battery?

Jordan

It's just power. Work out the total Vf of your 3 LEDs & multiply by the current you want to typically drive through them. Divide that by 0.85 (conservative efficiency rating of bFlex) and that gives you the input power. Take the input power & divide by the battery pack typical voltage and that gives you the input current and then multiply by the number of hours you want your light to run and that's the amp.hour rating that you need from your battery pack.

cheers,
george.

 

[Zero_Enigma]

It's just power. Work out the total Vf of your 3 LEDs & multiply by the current you want to typically drive through them. Divide that by 0.85 (conservative efficiency rating of bFlex) and that gives you the input power. Take the input power & divide by the battery pack typical voltage and that gives you the input current and then multiply by the number of hours you want your light to run and that's the amp.hour rating that you need from your battery pack.

cheers,
george.

George,

Sorry for my poor understanding of the electronics. I'm learning bit by bit slowly as I play with the LED's.

So if I understand you right with what you said above that would mean if I was using 3 x SSC P4's that would be:

3 x 3.7v SSC P4 = 11.1v / .85 n/bflex = 13.0588xxxx


This much I'm clear on. Now the next step I'm confused on.

Take the input power & divide by the battery pack typical voltage and that gives you the input current and then multiply by the number of hours you want your light to run and that's the amp.hour rating that you need from your battery pack.

So does this mean I take 13.0588 divide 1.2v typical voltage off the AA cells which gives me 10.8823 then I take that and multiply say 6 hrs which gives me 65.2941 :thinking: I plan on using a bunch of AA's at first then using a 14.8v li-ion later.


Thanks in advance for the help.

 

[georges80]

George,

Sorry for my poor understanding of the electronics. I'm learning bit by bit slowly as I play with the LED's.

So if I understand you right with what you said above that would mean if I was using 3 x SSC P4's that would be:

3 x 3.7v SSC P4 = 11.1v / .85 n/bflex = 13.0588xxxx


This much I'm clear on. Now the next step I'm confused on.



So does this mean I take 13.0588 divide 1.2v typical voltage off the AA cells which gives me 10.8823 then I take that and multiply say 6 hrs which gives me 65.2941 :thinking: I plan on using a bunch of AA's at first then using a 14.8v li-ion later.


Thanks in advance for the help.

Ok the 13.0588 is your watts output (assuming 1A - that's a LOT of light by the way).

How many AA's. Say 12 nimhs for bFlex with the 3 LEDs. Then you have 12 x 1.2 = 14.4V.

Now 13.0588/14.4 ~ 0.9A. So, if you want 3 hours of runtime (at 1A output) you need 0.9 x 3 = 2.7Amp.hour nimh cells.

Now, say we do the same calc for maxFlex - it's a boost so you need lower input voltage than output voltage. So, lets use 9.6V (8 nimh cells).

Now we have 13.0588/9.6 ~ 1.4A. So, if you want 3 hours of runtime (at 1A output) you need 1.4 x 3 = 4.2amp.hour nimh cells.

cheers,
george.

 

[znomit]

Yeah but LEDs are cheap and more efficient at lower currents. For a 14V battery instead of running 3 at 1A off bFlex why not 6 at 500mA off maxflex. More light, same runtime.

Oh, I know, because when you get out on the trails its too tempting to bump it up to 1A :thumbsup:. Now you have a stupidly bright light and curse your wimpy battery pack.

 

[georges80]

Yeah but LEDs are cheap and more efficient at lower currents. For a 14V battery instead of running 3 at 1A off bFlex why not 6 at 500mA off maxflex. More light, same runtime.

Oh, I know, because when you get out on the trails its too tempting to bump it up to 1A :thumbsup:. Now you have a stupidly bright light and curse your wimpy battery pack.

In that case you can run 6 off one bFlex.

Run bFlex at the 1A current range and wire the 6 LEDs are 2 parallel strings of 3 per string.

Each LED would then see approx 500mA (if the Vf's are reasonably close).

That still gives you a buck converter (more friendly to the batteries) and a shipping version of UIB2.

cheers,
george.

 

[znomit]

Yeah I do that on my MTB light. 2 strings of 2 at 500mA, though I have a 3ohm resistor I can switch in to one string to make it 700mA front/300mA side. I think its the way to go.

 

[SurfHenk]

George,

Could you tell us whether you have decided yet regarding the UIB2 on the maxflex? As said, I am interested. In fact I would happily use the UIB(1), except that I really would like my LEDs to be protected from overheating. An option only offered in UIB2 if I am correct..

 

[Zero_Enigma]

I wished that the driver would come out with TWO SWITCHES option. Clicking the same switch over and over is annoying and sometimes getting the timing right is hard (trying on my mates sensitive clicker).

Makes for programming easier and scrolling up/down in brightness as well.

Hope someone has this in the works.