Post Go back to editing

Hey all

I'm not happy with a current backlight solution I use on a custom board, it tends to blow up the driver IC for yet unknown reasons.

So I looked around for other LED drivers and found the ADD5211.

I've looked into the datasheet and also downloaded the calculation sheet, but now I'm not sure because I get completely different values than the example values in the datasheet and beside that I'm not sure which values I should use for calculation. I suppose I have to use the worst case values even if I might run the backlight at lesser current?

I have to drive 4 LED strings of a TFT backlight.

Each string runs at around 100 mA, max string voltage is 54V, at 100 mA probably less than 50V.

I also want to have an option to alter the string current if I have to use a different panel, so I might need more or less than 110 mA.

(In my current solution I just adjust the resistors that set the string current.)

As an inductor I want to use SRR1280-220M (22µH, Isat = 4.1A), which -if I'm not wrong- should be sufficient.

I selected SS5P9 as diode (If = 5A, Ifsm = 150A, 90V, could be replaced by SS5P10 if 100V would be better).

IRF7488 was chosen as NMOS switch (5A contin. drain current, 50A pulsed, Vds = 80V). Here I'm not that sure if it's okay to use this mosfet, but looking at the calculated values, it should be?

Switching frequency should be around 1000 kHz. (1070 kHz if Rfreq is 16 kOhm). I wouldn't mind changing the switching frequency, if anyone here says it's better to use more or less than that.

Minimum input voltage is around 23V.

Now, I'm not sure about the current sense resistor and the compensation parts.

Using the calculation sheet, Rcs is around 0.165 Ohm. What resistor should I use? 0.16 Ohm (are available as 1206 and 250 mW), or more/less than that?

If I calculate it correctly, Prcs is around 174 mW, so 250 mW should be safe?

About Cout: what values are suggested? I'm not going to use PWM dimming. Would it be better to use 4.7µF or rather 10µ or 22µF?

Altering the Cout capacity also alters the values of Rc and Ccomp. Rc's suggested value is 100 Ohm, is it okay to safely use other values such as 330 Ohm?

A quick calculation:

with Cout = 22µF, Rc would be 330 Ohm and Ccomp around 2.2µF. These values are common and available.

with Cout = 10µF, Rc would be 150 Ohm and Ccomp around 2.2µF, or 330 Ohm and 1µF. These values, too, are common and available.

But again, what Cout would be best and is it safe to use the just calculated values for Rc and Ccomp depending on Cout's value?

Whew, I hope this makes sense to you

I'd really appreciate some hints so I can design a better solution than my current one that causes troubles now and then ...

Thanks in advance and best wishes,

Robert

Parents
• Hi Robert,

Yes, there is very comprehensive protection against shorted conditions.  You can see the summary of them in Table 7 of the datasheet.  Exactly which of them kicks in depends on how (or where in the LED string) the fault occurs. So this covers all the cases of LEDs being shorted out, opened up, etc.

However, there is one thing in particular that you can't do to a boost regulator:

If you short the boost output (top of the LED string) to GND, then the OVP resistor divider sees it and shuts down the boost switching.  However, there is still a path from Vin, through the inductor, through the diode, to the output (which is shorted to GND).  In this case, prolonged operation here will probably damage the diode.  But the ADD5211, and the power NMOS, will not be damaged (because we detected this and stopped switching).

So that issue is common among all asynchronous (which means it uses a diode, not another FET, for the connection from switch to output) boosts.  However, if it is going to be a big problem for your application, then there are a few things you can do:

1.  The ADD5211 has a FAULT pin which gets pulled low during this event.  If you place a PMOS in the path from Vin to the inductor, then you can use the FAULT pin to shutdown Vin when a fault occurs.  But this will also trigger in the event of a shorted LED string or open string.

2.  Probably the most comprehensive protection is to use the ADM1270:

This will give you adjustable current protection, with a programmable delay.

• Hi Robert,

Yes, there is very comprehensive protection against shorted conditions.  You can see the summary of them in Table 7 of the datasheet.  Exactly which of them kicks in depends on how (or where in the LED string) the fault occurs. So this covers all the cases of LEDs being shorted out, opened up, etc.

However, there is one thing in particular that you can't do to a boost regulator:

If you short the boost output (top of the LED string) to GND, then the OVP resistor divider sees it and shuts down the boost switching.  However, there is still a path from Vin, through the inductor, through the diode, to the output (which is shorted to GND).  In this case, prolonged operation here will probably damage the diode.  But the ADD5211, and the power NMOS, will not be damaged (because we detected this and stopped switching).

So that issue is common among all asynchronous (which means it uses a diode, not another FET, for the connection from switch to output) boosts.  However, if it is going to be a big problem for your application, then there are a few things you can do:

1.  The ADD5211 has a FAULT pin which gets pulled low during this event.  If you place a PMOS in the path from Vin to the inductor, then you can use the FAULT pin to shutdown Vin when a fault occurs.  But this will also trigger in the event of a shorted LED string or open string.

2.  Probably the most comprehensive protection is to use the ADM1270: