High current digital pot?

[Atom058]

Hello - Is anyone aware of a digital pot that is capable of handling a high current (20 mA)? I need to control the output of an LED but I can't use PWM. The LED is being monitored by a spectrometer and the spectrometer can distinguish the highs and lows of the PWM signal (I've tried it). A digital pot would work nicely, but the ones I find can only handle small currents. I would limit the maximum current with a single resistor and use the pot to vary the current from 0 - 20 mA. This method works with a mechanical pot, but I need to control this via software. I would entertain other options, if anyone has any ideas. Thanks! Jeff

[mark_s]

Hi,
Google "programmable current source". Should do what you want.

[mark_s]

Have a look at this Analog Devices ADN8810.
http://www.analog.com/en/fiberoptic/laser-drivers/adn8810/products/product.html

[Jumper]

What about a RC filter out from the PWM and then any op connected as a buffer (voltage follower).

Then you can use the PWM to set the desired analog voltage and still have as many A as you need. Pick a suitable OP and you can have as much current you need.



R and C depends of the ripple you can allow and the HPWM frequency. I usually use a 2nd degree RC filter and that works great with an OP and led.

Simple, cheap and no need to use a digital pot

[amgen]

1 transister etc would work, find range on pot to work in.

5999

[HankMcSpank]

What about a RC filter out from the PWM and then any op connected as a buffer (voltage follower).

Then you can use the PWM to set the desired analog voltage and still have as many A as you need. Pick a suitable OP and you can have as much current you need.



R and C depends of the ripple you can allow and the HPWM frequency. I usually use a 2nd degree RC filter and that works great with an OP and led.

Simple, cheap and no need to use a digital pot

+1 (& certainly the way I'd approach it)

One other option (if you've sufficient free PIC pins) ....a basic DAC - whilst I'm no expert on DACS, apparently some have buffers integrated, so could likely drive an led.

Also, to get more granularity if going the digipot route, rather than fully supply/ground at either end of the pot wiper, arrange it so the top pot connection has just above the max forward voltage of the LED (so for a blue led with a fwd voltage of 3.2V, arrange it so 3.4V is present at the top of the pot) - the extra bit of voltage is for the Vdrop across your inline series resistor, like wise rather than ground at the bottom of the pot, arrange is to have the voltage where the LED just extinguishes (so for a blue led, probably around 2.8V)....this will give you much, much finer control over the led brightness.

[amgen]

yes, but having the resister on emitter makes the voltage to top of resister about equal to digital pot wiper voltage - .6 v.........
don

[Atom058]

Hello All! Thanks so much for your replies! For some reason, I did not get any notifications of your responses like I used to get - I guess there's a new flag that I need to set on my account to notify me of responses... I think I have solved my problem - It was a combination of the RC filter and upping my frequency. I am still tinkering with it and will post my results soon. Very encouraging so far! Thanks again! It never ceases to amaze me the tremendous knowledge that is available here in this forum!

Jeff

[Charles Linquis]

If it is useful, I have a ckt that will give NEARLY pure DC (100mV p-p ripple) from 0-47V (with 48V in), or 0-11.5V (with 12V in) at up to 20 Amps. It doesn't generate
much heat, either. Of course, it uses PWM.