TL4050A50-Q1: TL4050A50-Q1

Hi Dave,

I've just ran a quick test on a TL4050 with the same cathode current (10mA) and did not see a 250mV or 100us overshoot.

What kind of load is the TL4050 driving? Do you have any load capacitors?

What about the input? How fast is you power on transient?

The main factor in shunt references is the cathode current, which is determined by the source voltage and the resistor. That being said, the higher the current the higher your overshoot can be.

The part would not be able to handle 20V through a 500 Ohm resistor without a load, since the current would be 30mA. The maximum cathode current is 15mA.

Regards, Diego

Hi Diego,

There is a 0.47uF capacitor connected between the cathode and ground. 

The load is a few OpAmps that draw a total of 1.5mA.

The 10V input rise-time is about 500uSec.

Thanks,

Dave

Hi Dave,

Could you share a scope shot of the supply rail, cathode, and anode.

I tested the TL4050A25 under those conditions and I did get around 5% overshoot, but it only took about 50us to settle.

If you want to decrease the overshoot I'd recommend slowing down the input transient, which can be done with a simple RC circuit up front. 

If the issue is the length of time I'd recommend decreasing the capacitor between the cathode and ground, or adding a small resistor in parallel with it to empty the capacitor after overshoot. 

One last option is to add a small series resistance at the output, while this will help with the overshoot and decrease the noise, it will cause a small decrease in you reference value. 

Regards, Diego

The red trace is the Cathode input, through 500 Ohms, and the yellow trace is the Cathode voltage.

The overshoot itself is not a problem, provided i know what causes it and that it is limited to 0.5V.

Can you explain why there is overshoot and will it vary from part-to-part and temperature?

I thought it might be related to the regulation response speed of the part, but you folks are the experts.

Thanks,

Dave

For TI.pdf

Hi Dave, 

You are correct, the overshoot is due to the loop response time.

This is not a value we can guarantee since it is heavily dependent on the application. We have a couple of graphs on the LM4050-N-Q1 datasheet (section 6.11.1) that go over the usual response for very fast input transients. 

If you would like, I can run a test on three LM4050A50-Q1's to see how they behave under various start-up times.

Regards, Diego Lewis

Hello Diego,

In many applications knowing how large the overshoot could be is important to safeguarding connected devices.

Will a transient simulation show this behavior? 

If so, is there a SPICE model for this part?

Thanks,

David

Hi David,

We don't have a SPICE model for the TL4050, but we do have some transient SPICE and TI-TINA models for some of the LM4050-N's.

Regards,

Hi Diego,

TI-TINA models for the LM4050 would be excellent. 

Can you please post them.

Thanks,

David

Hi David,

 

Here is a link to the models.

 

Regards, Diego

Thank you. The models were helpful to studying the overshoot.
Thanks,
David