Ni-Mh 'AA' cells- a question

[underground]

Following my ongoing trouble with my Speleotechnics lamp, I'm gonna build myself a new lamp out of an Oldham headshell and a Petzl battery box.
What I'm thinking of doing is using one of those Luxeon focussed LEDs that are mentioned on another thread, in the 'main' bulb position, and something else in the pilot (if I need to), or if the main bulb has enough throw, maybe a 7 LED reflector instead.

Just wondering about burntime and suitability of rechargeable Ni-Mh cells, since I found some in the house yesterday, in light of comments saying that 4.5v flatpacks can break the LEDs...

I think I should check the no. of mAH of the cells shouldn't I? For arguments sake, let's say they're 2000mAH ones.

Can anyone suggest whether these would do the job- maybe how long they'd last off a single charge?

 

[Johnny]

Well me old mucker...

I dont know huge ammounts about it but I think it goes like this;
2000mah = 2ah (2000mah/1000=2ah) or two amp hours.
Which means that you can run a 1 amp bulb for 2 hours (2ah/1a=2h).
If you add more cells, in series, then you add the amp hours, two cells at 300mah = 600mah = 0.6ah which means that you can run a 1amp bulb for 0.6 hours (0.6ah/1a=0.6h) or a 0.5 amp bulb for 1.2 hours (0.6ah/0.5a=1.2h).
Volts are the same in that you add them together, in series, you add the voltages. One NiCad cell = 1.2v three NiCad cells = 3.6v. Your bulbs/LED's need to be rated at or around the voltage given by the batteries, any less and the bulb will not run efficiently and will be dim, any more and the bulb will burn brighter but for less time before it burns out.

What I dont know about, having trained as a mechanical and not electronic engineer, is charging the buggers.
Can anyone out there shed any light on this one?
Does anyone know what voltage the Speleoshitnics 7 LED arrays should run at?

 

[Huw Groucutt]

3.6 V i think

 

[Johnny]

Your right Huw, it will deffinitely be 3.6v as that is the voltage of the headlight and FX3 cells (3 NiCad's = 1.2vx3=3.6v). But are they able to go higher?
The difference between 3.6v and a flatpack at 4.5v is only 0.9v. If the arrays are rated at 4v then the difference is only 0.5v.

 

[paul]

Your right Huw, it will deffinitely be 3.6v as that is the voltage of the headlight and FX3 cells (3 NiCad's = 1.2vx3=3.6v). But are they able to go higher?
The difference between 3.6v and a flatpack at 4.5v is only 0.9v. If the arrays are rated at 4v then the difference is only 0.5v.

According to Phil Brown, at 3.6v the LEDs are already being run a bit over their recommended voltage - I don't know by how much.

 

[Johnny]

According to Phil Brown, at 3.6v the LEDs are already being run a bit over their recommended voltage - I don't know by how much.

Is he on commission? :wink:

 

[paul]

According to Phil Brown, at 3.6v the LEDs are already being run a bit over their recommended voltage - I don't know by how much.

Is he on commission? :wink:

Surely not!

 

[Roy Oxlade]

Hi all

LED's are not voltage dependant, you can run one from 1 volt to 100 volts so long as you use a current limiting resister.

An LED will be specified as needing say 20 mA and 1.5 volt forward voltage drop. So the resister required for a given application will be

( supply voltage minus forward volt drop ) divided by required current.

So running this from a 12 volt battery would require

( 12 - 1.5) / 20 mA gives 525 ohms, so use neatest prefered value below this.

hope this helps

 

[Dave H]

Most of the LED lighting applications I have seen have used a switching regulator to accept a wide range of input voltage, and yet still provide 'optimal' output voltage to the LED array.
Roy's suggestion of just using a resistor will work, but you will be wasting a lot of energy in the resistor (heat), the LED will not like getting too hot in a confined headset, but worst of all the current will alter dependent on the charge of the batteries, which they will not appreciate.
So far as charging NiMh batteries (assuming that you are going to use AA size): They are ten-a-penny, so do you really have to charge them? Otherwise just grab a charger from Maplins (I use a simple 4 way plug-integrated Uniross charger)
There have been lots of discussions about LED lighting in the CREG Journals.

 

[potholer]

I went for a different approach to the swicthing one - running off 3xNiMH/NiCd/Alkaline cells, I use a microprocessor-controlled hybrid linear /PWM regulator to run 2 Luxeons at up to 1W each

It seems to have worked pretty well in the units I've sold to mates so far (I'm wary of elaborating for fear of posting anything too commercial here), but did require a significant amount of electronics, which as far as I could see was only really practical in an Oldham headset by taking a pretty different approach to the whole unit. I'm not sure how feasible similar electronics would be in with a fully standard system using a side-emitting Luxeon in place of a bulb to generate the spot beam I doubt *I'd* be able to make anything compact enough for that kind of setup.

Presumably, most of the home-made switching regulator units are either in custom headsets, or behind a large LED array, without a spot beam?

 

[Dave H]

I went for a different approach to the swicthing one - running off 3xNiMH/NiCd/Alkaline cells, I use a microprocessor-controlled hybrid linear /PWM regulator to run 2 Luxeons at up to 1W each

PWM = Switching !

I'm not sure how feasible similar electronics would be in with a fully standard system using a side-emitting Luxeon in place of a bulb to generate the spot beam

Yes, I'm with you on this one. I don't know which micro' you are using but the 8-pin PIC's might be up to the job and are small, you might even be able to include a dimming function to avoid the need for a side-light (place a variable resistor on the on/off knob - but again it's changing the function of the lamp!)
Still, my main worry would be heat sinking of the Luxeon "they don't like it up them Mr Mannering!"

Is the light from the two Luxeon's sufficent to replace a 1W halogen bulb?

 

[potholer]

Ah - it depends what you *mean* by switching.

It's not a switch-mode regulator in the sense of voltage conversion, inductors, heavy RFI, etc.

Also, I use pure constant current (linear) regulation for my two higher brightness settings, and only use PWM regulation for the lowest setting, since using constant low current would tend to end up with the LED flickering, whereas PWM regulation gives a steady light level.

I use the LEDs independently at up to 1W each, one for a very wide light, good for nearby stuff, crawls, etc, and one in the reflector to give a strong spot beam as well as good downlighting in front of the caver. Using half the reflector for the main beam allows the electroncis to be mounted in the other half on an aluminium plate which also acts as a heatsink for the LEDs, and seems to be effective.

At 1W, the spot beam seems essentially to be in the same ballpark as even a good FX3. Nominally, it should be kicking out about 30 lumens compared to a 2.4W bulb's 40, but it'd be difficult to say which was brighter if looking at them in turn.

I'm using the PIC 12F675s now, which are ideal for the job, and I even managed to keep a pin free, until my recently-started and continuing work to make it more proof against bad Headlite connections.

 

[Dave H]

Sounds spot on!

Keep up the good work!

 

[potholer]

Would anyone have any experience of how a Nova performs with a flaky battery connection?