John Fields said:
I have a requirement for a constant current supply to drive 7 high power
LEDs wired in series.
The voltage applied to the LED series string must be adjustable between 25
to 33 volts and the current regulated at 300mA.
Any circuit ideas, chips etc?
---
Try this: (view in Courier)
LM317
+-------+
VIN>--+--|IN OUT|---+
| +---+---+ |
| | [3R9]
| | |
[0.1µF] +-------+
| |A
| [LED]
| |
GND>--+--------------+
If your LEDs have a Vf of 3.5V, that'll be a total of 24.5V, so with
a Vin of 25V you'll only have 0.5V across the regulator, which will
cause it to drop out of regulation. It needs about 2V across it to
work at room temp with a 300mA load, so your lowest supply would
have to be 26.5V. If you absolutely _have_ to have compliance down
to 25V it can be done with an opamp and a MOSFET and I'll give you a
schematic if you need one.
With a 33V input, the drop across the regulator would be 8.5V and
the current through it would be 300mA, so it would be dissipating
about 2.6 watts and you'll need a heat sink.
Thanks for that info John.
I can't quite see how you derive the constant 300mA with that LM317 cct
above.
The LEDs are 4.1 volt 1W devices. I made up a PSU using 6 x LM317T regs,
one for each bank of 7 LEDs. This drove the LEDs "sort of" OK, not quite
getting to their required 300mA, a bit short on voltage.
The customer now insists that the LEDs be "current driven" and produced some
chips suggesting I build a PSU using them. They are DD311 - I don't know if
you know them. Quite interesting devices, except I can't seem to get the
buggers to work as I would expect. (Data sheet available at
http://www.tenrod.com.au/htdocs/products/download.asp?pid3302 ).
The way I see it is (refer to page 11 of the .pdf using the diagram at the
top of the page) if the cct was set to provide a constant current of 300mA,
then if load ILED was increased, no additional current should flow, it would
be limited to 300mA.
However when I hook it up using a stiff voltage souce for VLED/REXT and a
REXT resistor as specified using the chart on page 6 it doesn't work out. If
I increase the load then the current in the path VLED (pin 2) also
increases? I would expect it to stay at 300mA?
I use load resistors in place of LEDs for the test - could this be the
incorrect way of testing and be the cause of this result?
Your comments would be appreciated.
