2 x P7 DIY bike light **complete**pics

[Goldigger]

I've nearly finished my first DIY bike light..
Its using 1 P7 led, battery pack is AA NiMh 4S 2P 8AA in total 4.8 - 5.6volts..
What driver would you guys use out of a HIPCC or one of these konlux 2800 ?
I think the konlux will give me more runtime as the HIPCC will cut out sooner because of the voltage drop..

 

[georges80]

Re: P7 DIY bike light, nearly complete..pics

I've nearly finished my first DIY bike light..
Its using 1 P7 led, battery pack is AA NiMh 4S 2P 8AA in total 4.8 - 5.6volts..
What driver would you guys use out of a HIPCC or one of these konlux 2800 ?
I think the konlux will give me more runtime as the HIPCC will cut out sooner because of the voltage drop..

With either driver you'll not have full regulation for more than a very short while. Rewire you battery pack to 8S - and then you can use either driver and your light will work much better.

With only 4S/2P you have 4.8V and dropping fast under load. Plus any wiring drops, contact drops and you'll be looking at 4.5V or less. Loaded nimh cells that are only partially discharged will be 1.1V per cell, so you have 4.4V without any other losses. Not a feasible battery/load combo with a buck converter, especially when its also trying to push 2.8A to the LED load.

With 8S you have less current draw from the batteries, so contact resistance and wiring resistance will be less of an issue and you have more 'headroom' to deal with the inevitable voltage losses .

cheers,
george.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

With either driver you'll not have full regulation for more than a very short while. Rewire you battery pack to 8S - and then you can use either driver and your light will work much better.

With only 4S/2P you have 4.8V and dropping fast under load. Plus any wiring drops, contact drops and you'll be looking at 4.5V or less. Loaded nimh cells that are only partially discharged will be 1.1V per cell, so you have 4.4V without any other losses. Not a feasible battery/load combo with a buck converter, especially when its also trying to push 2.8A to the LED load.

With 8S you have less current draw from the batteries, so contact resistance and wiring resistance will be less of an issue and you have more 'headroom' to deal with the inevitable voltage losses .

cheers,
george.

Hi George,
If i have 8in series that will give me 9.6volts correct? wont that fry the P7?
I have both drivers so not a problem which one i use, i have the hipcc already wired up..

Cheers

 

[Linger]

Re: P7 DIY bike light, nearly complete..pics

Well, you can try this one while you decide:
http://www.shiningbeam.com/servlet/the-132/3-dsh-Mode-Regulated-Circuit-Board/Detail
from Shiningbeam. Its the driver from the MG PLI, I've got another in a p60 drop-in, and I bought two more for diy P7. It's very robust, meticulously assembled. (as opposed to some cheap boards, you won't find any questionable solder joints or swarf on the board), less then $7

 

[georges80]

Re: P7 DIY bike light, nearly complete..pics

Hi George,
If i have 8in series that will give me 9.6volts correct? wont that fry the P7?
I have both drivers so not a problem which one i use, i have the hipcc already wired up..

Cheers

That's the whole point of a buck converter... It will take the higher input voltage (9.6V in this case) and do a DC<->DC conversion to step the voltage down. Being current regulators, either board will then adjust the output voltage to maintain a constant 2.8A going to the LED load.

The other 'regulator' posted above this one is a linear regulator - it will 'burn' off extra voltage as heat. That style is cheaper since all it has are linear regulators versus the complexity of a buck converter that requires more components (inductor, switcher regulator and possibly an external FET, schottky diode and related capacitors etc).

cheers,
george.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

That's the whole point of a buck converter... It will take the higher input voltage (9.6V in this case) and do a DC<->DC conversion to step the voltage down. Being current regulators, either board will then adjust the output voltage to maintain a constant 2.8A going to the LED load.

The other 'regulator' posted above this one is a linear regulator - it will 'burn' off extra voltage as heat. That style is cheaper since all it has are linear regulators versus the complexity of a buck converter that requires more components (inductor, switcher regulator and possibly an external FET, schottky diode and related capacitors etc).

cheers,
george.

Why didnt you tell me this earlier lol :laughing:
I thought that i needed to match my battery voltage to the forward voltage of the LED that best i could..

cheers

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

That's the whole point of a buck converter... It will take the higher input voltage (9.6V in this case) and do a DC<->DC conversion to step the voltage down. Being current regulators, either board will then adjust the output voltage to maintain a constant 2.8A going to the LED load.

The other 'regulator' posted above this one is a linear regulator - it will 'burn' off extra voltage as heat. That style is cheaper since all it has are linear regulators versus the complexity of a buck converter that requires more components (inductor, switcher regulator and possibly an external FET, schottky diode and related capacitors etc).

cheers,
george.

George,
I did my maths and with both battery configurations 8S or 4S 4P i came up with exactly the same runtime..
In my previous post i had typed 4S2P should have have been 4S4P 8 cells in total..
Im sure you will correct my maths but either way the runtime is sh?t how can i get 2 hours how about using lions? 2S4P to give me 7.4v @7500ah

P led = 1 x 3.7V x 2.8A = 10.36W
Battery = 9.6V x 2.5Ah = 24Wh
Let's say efficiency is 80% for the driver.
P in = P led / Eff = 10.36W / 0.80 = 12.95W
Therefore Runtime = Battery / P in = 12.95Wh / 24W = 0.53 hours

P led = 1 x 3.7V x 2.8A = 10.36W
Battery = 4.8V x 5Ah = 24Wh
Let's say efficiency is 80% for the driver.
P in = P led / Eff = 10.36W / 0.80 = 12.95W
Therefore Runtime = Battery / P in = 12.95Wh / 24W = 0.53 hours

Why didnt i buy the hipflex...

Cheers

 

[znomit]

Re: P7 DIY bike light, nearly complete..pics

Therefore Runtime = Battery / P in = 12.95Wh / 24W = 0.53 hours

You got that around the wrong way. Should be good for almost 2 hrs (24Wh/13W).
Though I usually drop another 10-20% to account for overly optimistic battery ratings.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

You got that around the wrong way. Should be good for almost 2 hrs (24Wh/13W).
Though I usually drop another 10-20% to account for overly optimistic battery ratings.

Yeah i thought so.. i was looking at it and thought that dont add up..
cheers...

 

[georges80]

Re: P7 DIY bike light, nearly complete..pics

Also, to get the terminology right in this thread for what is generally understood for 4s2p, 8s, 2p etc etc:

XsyP is generally understood to mean "x" devices in series to make up a string. Then we have "y" of those strings in parallel.

So, 4s2p means a total of 8 devices made up of 2 parallel strings with 4 devices in series per string.

So, your original pack of 8 batteries wired as 2 parallel strings of 4 series connected batteries was correctly written as 4s2p. Re-wiring those nimh cells in 8s gives you 8 series connected batteries.

cheers,
george.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

I plan on swapping the hipcc driver to a hipflex so i can have the UIB2 function.
I know i need a momentary switch (waterproof), does anybody know what the contact rating should be? eg: 2amps at 50volts..
I assume this depends on the current draw from the battery?

Cheers

 

[znomit]

Re: P7 DIY bike light, nearly complete..pics

I know i need a momentary switch (waterproof), does anybody know what the contact rating should be? eg: 2amps at 50volts..
I assume this depends on the current draw from the battery?

Cheers

The mom switch is just a signal switch, it doesn't carry any power so you can use pretty much anything.
These are popular as they are waterproof.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

The mom switch is just a signal switch, it doesn't carry any power so you can use pretty much anything.
These are popular as they are waterproof.

Thanks....:twothumbs

This should do the trick
http://uk.rs-online.com/web/search/...ethod=searchProducts&searchTerm=59-511#header

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

Is this the correct way to wire up the hipflex? thought i might as well check before i fry something..
I cant see it being wired any other way but i a may be wrong..

Thanks

 

[Mick]

Re: P7 DIY bike light, nearly complete..pics

You have the LED+ going to the mounting hole. You want the hole below the "LED+" text.
Otherwise your good.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

Amended...cheers for spotting that.. im sure i would have soldered the correct terminal anyway..

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

Ive ordered a couple of hipflexs from George.
I ordered two as i am going to make another so the mrs can have a light to, plus i can also have twin lamps when shes not using it... 1800 lumens max :thumbsup:.
Im just looking at batterys, i must admit im slightly confused regarding configuring batterys to suit 1 P7 and a Hipflex..

 

[slappomatt]

Re: P7 DIY bike light, nearly complete..pics

if weight is a concern I would get Li-ion 18650's and make a pack. if your less worried about weight NimH are good. I bought 8X 5000Mah "C" cells from batteryspace a while back to use with my dual XR-E headlamp and its way overkill. 8.4Hrs runtime. lol

being as the hipflex + P7 are a high current low Volt arangement you might want to just get some high capacity NimH cells and make a pack. The advantage is you can run them down and not have to worry about cutoff voltage.

on the other hand the Hipflex has a low voltage cutoff I beleive so that shouldnt be a big deal.

My pack of 8 5000 MaH NimH C cells would run your light for around 4.5Hrs.
slightly lower cap 4500MaH sub-C's might save you some weight.

or you could get 4 2400 MaH li-ion cells wired series-parallel and should run that light for about 3.2 hours.

 

[Goldigger]

Re: P7 DIY bike light, nearly complete..pics

if weight is a concern I would get Li-ion 18650's and make a pack. if your less worried about weight NimH are good. I bought 8X 5000Mah "C" cells from batteryspace a while back to use with my dual XR-E headlamp and its way overkill. 8.4Hrs runtime. lol

being as the hipflex + P7 are a high current low Volt arangement you might want to just get some high capacity NimH cells and make a pack. The advantage is you can run them down and not have to worry about cutoff voltage.

on the other hand the Hipflex has a low voltage cutoff I beleive so that shouldnt be a big deal.

My pack of 8 5000 MaH NimH C cells would run your light for around 4.5Hrs.
slightly lower cap 4500MaH sub-C's might save you some weight.

or you could get 4 2400 MaH li-ion cells wired series-parallel and should run that light for about 3.2 hours.

How are you charging your pack? weights not really an issue to be honest..

 

[liquidmtber]

Re: P7 DIY bike light, nearly complete..pics

Goldigger, I really like you design on the light. I am new to all this and I am looking at building a light similar to yours. What did you use to take care of the thermal heat? Do you have a heat sink? Also what optics did you use?