TL431 vs Zener

4 posts / 0 new

Dir Sir/Madam, We used TNY278GN for a three phase power supply. During the tests using Zener diode as reference in the feedback, the drain voltage is normal; you can find them in fig.1, attached.

Using TL431 as reference in the feedback, the drain voltage is an on/off as shown in fig.2. Enable pin voltage is in the same state. In the on state the drain voltage seems to be working normal fig.3.

  1. Please let me know whether it is ok and due to the TL431 on/off characters.
  2. In case it was ok, please let me know whether using TL431 may affect the SMPS lifetime.

Looking forward to hearing from you. Regards

I'm not sure what you mean by the TL431 having on/off characteristics. 

 

It looks like the state machine of the TinySwitch is just operating in slightly different states between the two references.  This is reasonable since the two references are going to have slightly different set point not to mention their own unique operating characteristics.

 

Looking at your scope plots and the time bases involved, the TL431 pulse groupings appear to be at approximately 120Hz.  What you're seeing is probably just the TL431 responding to the line voltage ripple on the output of your power supply.  The TL431 is a much more accurate voltage reference than the zener, so the control loop of the TL431 is probably just responding to the line voltage ripple and compensating for it.  You can verify this pretty easily by looking at your your power supply output ripple (the low frequency ~120Hz ripple) and seeing if it's better on the zener or the TL431.

 

If you have any other questions, please let me know.

 

Regards,

The Traveler

HiThank you for your support.You told:”It looks like the state machine of the TinySwitch is just operating in slightly different states between the two references.  This is reasonable since the two references are going to have slightly different set point not to mention their own unique operating characteristics.” Please let me know what those two references are.About the circuit, if you check fig.2 you can see that using TL431, we have 500us nearly full-working periods every 8ms and it is because of the TL431 responding to the line voltage ripple on the output of the power supply.First of all, I want to be sure whether it is normal. Next, I want to know whether it, 500us full-working periods every 8ms, affects the switching power supply components life and operating time.Regards

 

salamisz - 

 

The two voltage references you're using are a TL431 and a zener diode. 

TinySwitch is designed to continuously move through different controller states in response to different operating conditions.  This is how an on/off controller with an internal state machine is going to work (generally).  The slightly different operating states shouldn't affect the life of the controller or of the power supply. 

However, as with any power supply design, you're going to want to come up with a thorough set of design verification tests. If you have the time in your design scedule, you might want to compare the two voltage references against each other and see if there are any advantages of one over the other in your design.

 

Regards,

The Traveler