Is there any reliable way of dynamically controlling the LED current for the STCS2A?

[SThor]

Dear all

Because the STCS2A have separate pins for the FB and Source it is attempting to Think by manipulating the FB pin's voltage feedback to change the actual current through the leds.

I tried to connect an operational amplifier between the source pin and the FB pin and changed the gain to "fool" the FB pin to think that therefrom normal current to too much (which would mean the current through drain/source should go down to match the 0.1V requirement) but I do not get a reliable result.

Is there an easy solution to this that I have not seen?

I choosed the STCS2A because it have a fault pin (DISC) and it is a "LED current driver".

Regards

[SChak.1]

We are doing a similar thing with this chip. We have just put a resistor divider in parallel with the current setting resistor. Then feed the resistor divided voltage to the FB pin. The ( 1 + ratio of division ) will be the current gain .

We have infact put a I2C programmable potentiometer for the top resistor of the divider. This gives us programmability through code

No Amps necessary. Also not a good idea since we dont know the loop dynamics.

[Vikipat]

Can you please share circuit?

[Peter BENSCH]

First of all, this thread is already three years old.

Then: a schematic for a voltage divider?

Well, connect two resistors in series, then these two in parallel to RFB (see data sheet) and finally connect FB to the junction of the two voltage divider resistors.

If you build the voltage divider with e.g. 100ohm resistors, you get the divider factor (1+100/100) = 2. Therefore, only 1/2 of the voltage of RFB arrives at pin FB.

This procedure maps the LED current to the setpoint 100mV, but quadruples (with voltage divider=2) the power dissipation at RFB, because the voltage is squared in the power calculation.

Theoretically, the idea of @SThor​ would be better, but you also need opamps that can work well around an output voltage of (100mV/factor), usually at a few millivolts. Apart from special rail-to-rail opamps, only opamps with negative VCC- are able to fulfil this requirement.

However, it is also questionable whether current adjustment through the LED makes sense, since

• the brightness of the LED does not scale linearly with the current
• massive chromaticity shifts can result, especially but not only with white LEDs

Regards

/Peter