0
\$\begingroup\$

Suppose that A and B in the image are connected to the two output pins of LED driver, which can be independently switched on and off a constant current of 30 mA.
D1, D2, and D3 are all the same type of LED and I want D3 to act like the output of an OR gate, i.e., to turn ON when either or both D1 or D2 is ON.
However, in my circuit, D3 will have 60mA and will not have the same brightness when both D1 and D2 are ON.

Is it possible to make D3 have the same current as D1 and D2, i.e. 30mA, and still behave like an OR gate??
Of note, other pins of the LED driver are not available because they are occupied for different purposes.

Edits: These are LEDs and a LED driver I am going to use.

enter image description here

\$\endgroup\$
6
  • \$\begingroup\$ FYI, here is the LEDs I am going to use: downloads.cree-led.com/files/ds/h/HB-CLM1B-BKW-GKW.pdf and a LED driver: cdn-shop.adafruit.com/datasheets/tlc5947.pdf \$\endgroup\$
    – matcha
    Commented Aug 11 at 20:26
  • 1
    \$\begingroup\$ You can edit your question to include the LED datasheets. \$\endgroup\$ Commented Aug 11 at 21:27
  • \$\begingroup\$ Thank you @FabioBarone, I edited my question. Do you have any suggestion about the circuit? \$\endgroup\$
    – matcha
    Commented Aug 11 at 22:36
  • \$\begingroup\$ Yes, I have a possible solution, but I need to draw it, and I am not at my computer yet. Maybe later in the next few hours. \$\endgroup\$ Commented Aug 11 at 23:26
  • \$\begingroup\$ PS: when you edit a question, you should indicate that change, so that previous readers can quickly see the change. Put yourself in the position of a reader, what would you like to see if the author has changed a post since you last read it? \$\endgroup\$ Commented Aug 11 at 23:38

4 Answers 4

1
\$\begingroup\$

This is a way to limit the current through D3 to approximately 30mA, which is not quite what you asked for:

schematic

simulate this circuit – Schematic created using CircuitLab

The TL431 circuit is fairly accurate and stable with temperature but responds slowly compared to PWM. The transistor is less accurate and stable with temperature but responds fast. The effect of "accuracy" is that the brightness may change noticeably when D1+D2 are on vs. just D1 or D2. To reduce that effect, you could limit the current to something like 25mA (increase the resistor values by about 30/25) so that it is always limiting.

\$\endgroup\$
1
\$\begingroup\$

Here I used a single PNP bipolar transistor to act as a 30mA current source:

schematic

simulate this circuit – Schematic created using CircuitLab

Ignore Da and Db, which are only present to constrain voltage across sources I1 and I2 when their switches are open.

When either R1 or R2 passes 30mA, they develop 3V. Diodes D4 and D5 wire-or that voltage to the base of Q1, which places about 1.8V across R3, setting current through it to:

$$ I_{R3} \approx \frac{1.8V}{60\Omega} = 30mA $$

All of that current passes via D3, so if either D1 or D2 is lit (or both), so is D3.

There will be 19V across Q1, so it will dissipate \$19V\times 30mA=0.6W\$. That's quite a lot, so consider reducing the supply voltage to 12V. Your driver IC will be dissipating the same power per channel, so you'll be helping that too.

\$\endgroup\$
0
\$\begingroup\$

No.

You have defined a hypothetical question with some current-mode logic using diodes combined into one so-called "diode-OR" voltage logic. That would violate Kirchoff Laws to do what you asked.

Maybe. You could limit diode current with a precision current sensor, reference voltage and switch on an active shunt. But no simple RTL logic solution as you might be thinking.

A simple CMOS OR/NOR gate could replace this attempted question but use only 3.3V to 5V logic rather than waste 24V.

\$\endgroup\$
2
  • \$\begingroup\$ Do you mean it is possible with 3.3V or 5V? Could you show me an example? Eventually I have to use 24V because D1, D2, and D3 is actually not a single LED, but a series of 3 or 4 LEDs, but an example with 3.3V or 5V may give me some idea. \$\endgroup\$
    – matcha
    Commented Aug 12 at 0:44
  • \$\begingroup\$ Until you define all the I/O interfaces, channels and voltages, I’d rather not speculate but multi input NOR logic can sink 1 LED current from low Vdd. Is your string 12V or 24V or TBD? Your PWM Mux uses a serial register. You may need to include in your query for intended usage. \$\endgroup\$
    – D.A.S.
    Commented Aug 12 at 3:53
0
\$\begingroup\$

Here is a way to limit the current in D3 to 30mA, while using the same current that flows in the other LEDs - ie: D3 does not consume more current from the power supply than what is already being used by D1 and D2.

schematic

simulate this circuit – Schematic created using CircuitLab

Circuit Description
When D1 or D2 has 30mA of forward current, the same current is passed through D3, as set by the current source formed by Q1, Q2, R1, R2. However, when both D1 and D2 are on, the extra 30mA bypasses D3 via D4 (zener) and D5.

The components R1, R2, Q1 & Q2 form a constant current source of about 30mA, which is only energised when either of the current sinks for D1 or D2 (or both) are turned on. If both D1 and D2 are off, then it remains in the off state.

You could introduce another LED in series with D4 to indicate the AND function, ie: when both D1 and D2 are on (refer D5).

The voltage required for zener D4 should be selected based on testing of your circuit. The value shown (5V) here is merely a starting point. Note that D4 will dissipate more power as its break-over voltage increases. Select the breakover voltage of D4 to be as low as possible while still providing voltage margin for the current source and LED D3 to work correctly. The anode of D3 only needs to be about 1V below Vcc_24V for the current source to work properly.

R1 is selected to give 30mA of current into the emitter of Q1. R2 is selected to provide sufficient base current for Q1, and needs to be adjusted if supply voltage (24V) is varied significantly.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.