TL431 constant current source for LED from 5V

He said he wanted temperature stable with a 5 volt Vcc. I assumed Vcc was regulated. AFAIK he didn't say that Vcc was going to be all over the place!

If that's another requirement, well I can work with that as well.

I was hoping to find a way to avoid a third external tranny, but that'll work.

But I'm not sure I actually completely understand the OP's problem. It seems maybe more not that the current is changing so much under temp fluctuations, but that the actual light output is fluctuating over temp even if the current is nominally constant, due to LED physics? Or something?

Is the PWM duty cycle set by some type of optical feedback?

If that's the OP's problem and the light output has to be absolutely constant over temperature then I think the correct solution is a cheap uP-driven current source with an onboard temp sensor and a lookup table.

The TO-92 package is pretty expensive - it seems you could do the same thing with an ATTiny uP wth 512 bytes of Flash and a lookup table for about 50 cents...but of course then you'd have to program it

Depending on the final number produced, but also other issues like board space, it does indeed look like it would require careful consideration and might be a tough call.

One advantage I do see for the MCU is that it would be straightforward to use different LEDs if required - just digitize the datasheet curves for a particular LED into a lookup table and interpolate in software.

I'm not able to download these files without signing up. They are just text files. Why not post them here?

--

Rick

BTW, the TL431 has a 2.5 volt dropout plus a transistor if you are upping the current. That is too much for blue/green LEDs on a 5 volt supply. The LM234 can be used with a PNP to give a minimum dropout of around 70 mV plus the PNP sat voltage. It can supply up to 10 mA by itself with a dropout of 0.9 volts or less.

--

Rick

When the register pop-up occurs, just X it away. Then re-click on Download. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Why don't you vary Vcc and see how the current changes? ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Something like this...

...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Could be. My billboards drivers are PWM'd by a uP to compensate for temperature , etc. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Yes, the 5 V is regulated. My concern is the current (really the illumination) being controlled by the uncontrolled variables, process and temperature. If people are sure you can't see a 45% deviation in illumination then I am wasting my time. I expect a 45% change is noticeable though.

--

Rick

Yeah, at temperature goes up, voltage goes down which will result in some increased current depending on the driving voltage. The physics of the LED say there is more recombination at higher temps so less makes it

temperature range. I found another one today that shows something similar for red LEDs, but much less for blue, so go figure.

Did you read my post about the LM234? It is not so cheap, but uses a very simple circuit and does the temperature compensation.

--

Rick

Yes. There is a loop around a "pixel" (RGB) not visible to the viewer. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Yeah, by the time all is said and done I'm not sure the MCU would be cheaper.

--

Rick

Hey, that's way too much stuff. If there's already regulated +5, and you just want same-current biasing of IR through UV emitters, use a current mirror. Two transistors, or one dual, and done.

Input of the current mirror (NPN transistor) is a base-emitter diode drop to ground, about 0.60V; to change the current by one percent, that would have to change to 0.65V, so about 20 degrees C (36 F).

Program the current with one resistor (from +5 to C1+B1+B2), and connect the LED from +5 to C2.