LP5521: Modulating LED output with external FET.

Hi Mark,

Actually, I didn't figure out your problem.

For your connection, I also didn't figure out why you want to use an external FET to control it.

Since based on my understanding, I think if you only connect a MOSFET in serial, the function of this FET is the same as PWM dimming. Could you please tell me more detail information why you need to connect the MOSFET in serial?

BR, Jared

Hi Jared, I reliase I didnt actually put a question Sorry about that :-)

Why am I using an external FET? => We need to be able to precisly control the operation of the LED in the time domain. Including ON/OFF modulation at 38KHz, which is too fast for the LP5521. 

In effect we are simply using the LP55211 as a programmable current source on channel R, which we turn on and off using our external FET. 

My question is would the LP5521 support operation in this mode? 

I appreciate we could  solve this problem by replacing the LP5521 with a bespoke/discreet circuit, but the footprint to accomodate this and the drive for the other 2 LED's (G & B) is much larger. We need a very small footprint. 

Hi Mark,

Sorry to late reply. I think it should be possible.

BR, Jared

Thanks Jared. Can you see any reason why it would not work?

As mentioned the only issue I can see is there might be a small voltage spike on the LED going from FET off to FET on.

Is there any LED open fault detection circuitry inside the IC which might catch. us out?

Hi Mark,

Please give me some time since I am OoO now, I will get back to you ASAP.

BR, Jared

Hi Mark,

Sorry for late reply.

Actually, LP5521 doesn't have the LED open protection. Could you please capture some waveforms to help me locate the problems?

Or I'd like to recommend you to use LP5811, the next generation of LP5521, since LP5811 has the LOD protection and it has better performance than LP5521.

BR, Jared

To be clear I do not want the IC to show a fault on Led Open Circuit. This would cause a problem when we modulate at 38kHz using the external FET. 

I will re-write the situation: 

1. We want to be able to precisly control the LED on channel R in the time domain, by turning it on and off at 38kHz in a custom pattern. 

2. We will configure LP5521 to drive a fixed current on channel R (say 5mA). Then control the LED ON/OFF using the external FET.

3. I can see that when we turn off the FET then the voltage on channel R will rise up close to VBATT. When we turn the FET on it will take a brief time for it to drop to the correct voltage for 5mA. so we may see a breif higher current blip when turning the FET from OFF to ON. This is ok.  

4. I am looking for some confirmation that there is no LED open fault circuitry inside the LP5521 which would disable channel R when we use our FET. We want the LP5521 to keep driving channel R whether the external FET is on or off. 

No I will try and answer some question you may have: 

A. Why dont we design or use another chip that is simply a constant current source? => We need a very small footprint and want to use channels G and B in the normal mode. 

B. What is the problem we are seeing? There is no problem at the moment, our circuit is being prototyped as we speak. We are looking for confirmation on the inner workings before we get too invested in this solution. I think you have confirmed there is no open LED fault detection which is good. 

C. LP5811? This is no good We need precise timing control at 38kHz.

My question: 

* Is there any fault mode in the LP5521 which may cause a problem when we turn off our FET, then back on again? 

Hi Mark,

Please refer the below comments for my answer:

1. First, I can see that the package size of LP5811 is smaller than LP5521.

2. Actually, as you said, there is no problem at the moment. I think it should be no problem in my side. I have some questions because I thought you met some problem at first.

3. For your precise timing control, I am not sure what's your pattern. Could you please tell me what does it look like? Since LP5811 has internal program animation function, maybe it can help you to figure out the timing control.

In a word, I think LP5811 will be a better solution than LP5521.

For your last question, does LP5521 has any problem. Please give me some time to test it on EVM board since I am OoO. It maybe take some days. Thanks for your patience.

BR, Jared

Thank you for your time on this Jared.
I am glad you do not see a problem with the LP5521 from your side.
As soon as I have built our prototype boards I will let you know the result of the test.

I appreciate your suggestion on the newer LP5811 part. But I think we are better to stick with the LP5521.
Incase I have misunderstood something these are the reasons

1. I agree the IC package on the new LP5811 is smaller:
LP5521TM : 2.09mm x 1.73mm = 3.6mm2
LP5811xYBHR : 1.84mm x 1.43mm = 2.6mm2
But the need for external inductor and larger capacitors increases the footprint quite significantly.
As a solution I believe the LP5521 is smaller.

2. One of our timing requirements is to turn the LED on and off at 38kHz.
i.e. on for 13.1us, off for 13.1us.
It looks like the fastest the LP5811 can switch the LED on and off using its animation block is 24kHz.

3. If the fault detection features of the LP5811 shut down the channel this would prevent us using an external FET.
We would have to use the animation features which are too slow for us as described above.

What do you think?

Mark.

Hi Mark,

Thanks for such detail explanation!

1. For the first question, could I know what's your input voltage?

2. Yes, if the timing requirement is very strict, I think it can only use external FET to realize it.

3. Actually, LP5811 has an option to disable the LOD behavior, so I don't think LOD protection will affect it.

Looking forward your test result, hope you are doing well!

BR, Jared

1. I think I can guess where you are going on this, unfortunatly we need to keep the boost. The input voltage is a Lithium cell Vin is 4.5V down to 3V. We do not need a boost for our IR LED (LP5521 Red channel lets us fix VBatt on R channel), unfortunatly the other colours need a boost when we get down to 3V.

2. Agree.

3. I didn't spot this on the datasheet first time around. Good to know the LP5811 may still be an option. Thank you. 

Hi Jared, If possible I'll add the test result to this thread tomorrow or Thursday, then close this ticket. Just so other followers on this thread will know the outcome. 

Sure!

Jared Zhou I have built up and tested 8 prototype circuits, all work fine. Modulation is working with the external FET as expected. There does not appear to be any race condition - External FET and LP5521 can be enabled in either order. Thank you for your help on this.