48v LED circuit

[MonkeyMagic]

Howdy

Thought I'd post this if you wanted to run additional LED's on your ebike.

9 x LED
Parts required:
1 x 2K 1Watt resistor
1 x IN4104 diode
9 x high intensity LED

Tested for half an hour with 9 x 5mm UV LED's no issues

Also tested with taps from LED series (eg. around 12.6v from 4th LED) with resistor to power relay latch etc.

Still use at your own risk lol but it works for me

Just took the design from a ebike 9x LED headlight that was in pieces and followed the PCB tracks so it should be good to go.

Cheers

[Bikemad]

Hey Monkey, you must take into account the forward voltage and current of the particular LEDs that you are using. Most standard 3mm, 5mm and 10mm LEDs require a forward current of around 20mA. The forward voltage for a Red LED is usually between 1.8V and 2.4V max, whereas White and Blue can between 2.8V and 4V depending on the manufacturer (see details attached for more information).

Tapping into a string of LEDs is not recommended as it puts excessive load on the LEDs which are supplying the additional current for the device you are attaching.

I have attached some details below showing some typical LED circuits and information etc. if you're interested. I used the LED series/parallel array wizard for most of the circuit diagrams.

The LED lighting mod was done on my father's bike when he first got it because the original lighting was no good in the dark lanes.
The seven led headlamp made a huge difference and only required about a tenth of the power, and I added the brake lights at the same time.

I have to admit to being a great fan of LED lights, I recently bought a 3 Watt 240Lumen LED torch which has a sliding lens to adjust the focusing of the beam and the light output is absolutely brilliant. I was so impressed that I've bought two of these 300 Lumen headlights to see if I can adapt them for bike use by modifying a couple of cheap cycle light sets.

Alan
 

 

[Cornelius]

Good point and a good writeup.

The best would be to use a constant current regulator, but those are hard to get for voltages above 35-40V sources...

[MonkeyMagic]

Cool thanks for that

I measured the voltage between the LEDs and it was 3.16v which is fine for the ones I'm using, I'll take out that part about the brake light but as long as you restrict the current from the tap and less than the total current of the led string it will be fine (ie, a momentary signal with resistor for a relay)

hey that focus beam light is cool! When you say you ran them on your dads bike what battery was it from?

cheers

[Bikemad]

When you say you ran them on your dad's bike what battery was it from?

It's a 24V 8Ah NiMh which came with his Peerless electric bike (see attachment).

The best would be to use a constant current regulator, but those are hard to get for voltages above 35-40V sources...

It should still be possible to use a simple current regulator using an LM317T (up to 1.5Amps) or LM338T (up to 5 Amps) and a 60 ohm resistor as shown here:


But only if the voltage drop across the regulator does not exceed 40V!

So a 48V battery (55V max) could be used to provide a constant current supply to a series string of between five and twelve 3.6V LEDs.
 
Using it to power just a single LED is likely to kill it, but it might be OK if you had two identical regulator circuits wired in series.
The bigger the voltage drop across the regulator, the hotter it will get (and a suitable heatsink will be needed), so it's far better to use as many LEDs as possible and generate more light and less heat.

The advantage with using a constant current regulator is that the LEDs would never be overloaded, and would always run at a constant voltage, ensuring they would always be the same brightness, regardless of whether the battery was fully charged or almost completely drained.

You can use the Current Calculator to determine what size resistor is required for a specific current requirement.

Alan
 

[Bikemad]

It should still be possible to use a simple current regulator using an LM317T (up to 1.5Amps) or LM338T (up to 5 Amps) and a 60 ohm resistor as shown here:


But only if the voltage drop across the regulator does not exceed 40V!

So a 48V battery (55V max) could be used to provide a constant current supply to a series string of between five and twelve 3.6V LEDs.

I have carried out some basic testing, and can confirm that is possible to use an LM317T regulator chip on a 48V battery pack, and it appears to work perfectly (see attachment below).

For testing purposes I used a 13 LiPo cell pack (48.1V nominal, 54.2V actual) as a voltage supply, and a single LM317T regulator chip, which was wired to supply a constant regulated current to five LEDs connected in series. This resulted in a voltage drop of 36.2V across the regulator, which as far as I'm concerned is within its acceptable input voltage range of between 3 and 40 volts, which the chip is able to handle.

I had the current regulated to 37.9mA in order to give a voltage of 3.6V across each of the LEDs (18V across all five), and after a few minutes of use, the heatsink temperature stabilised at ~35ºC while dissipating the wasted power (54.2V - 18V = 36.2V, 36.2V x 37.9mA = ~1.37Watts).

I then added some more LEDs and carried out the same checks again.
With 11 LEDs, the heatsink was noticeably cooler (~26ºC) as it was only having to dissipate 0.55Watts (14.6V @ 37.9mA) of wasted power in the form of heat.

The current was being regulated correctly and I there was no sign of any problems whatsoever from the LM317T regulator.

Having proved that the current regulator works just as I expected it would, I'm now curious as to whether the same theory can be applied to Voltage regulation, as I'm convinced it should be possible to use the LM317T to regulate a voltage well above its 40V maximum limit, provided the voltage drop across the regulator is less than 40V.
So it should be possible to use a single LM317T on a 60V supply to provide a regulated output for operating 24, 36 or 48V relays etc., so long as the required output remains above 20V.

For lower output voltages, it should be possible to use two LM317Ts connected in series to form a two stage regulator as shown in the attachment below.

Yet another item to add onto my "interesting experiments" list, I just need to organise some extra "play" time.

Alan