Ucc2818 Voltage error amplifier

[txbob]

Hi,
In one of the existing design I was working (pfc boost with UCC2818D controller) , I noticed that soft start is implemented by pulling the voltage error amplifier down through a diode (which will make multiplier output to zero) as shown in blue lines. This is done during startup and also some transient events when the boost duty cycle needs to be drastically reduced
The uc2818 does have a dedicated soft start pin which can be activated by OVP/EN pin (Pinno:10) but for some reason initial designer didn’t use that feature for enabling soft start, may be the duty cycle will not go down all the way to zero….
My question is there anything wrong pulling down on error amp output ? The datasheet is not saying that it is a transconductance amplifier.


https://www.ti.com.cn/cn/lit/ds/symlink/ucc3817.pdf

 

[KlausST]

Hi,

If you switch the bjt ON then the In- is pulled low.
Therfore the opamp output wants to drive a high.
Now the diode gets conductive and the opamp output is fighting agains the bjt.

It is like shorting the opamp output to gnd.
Avoid this.

Klaus

 

[txbob]

Hi,

If you switch the bjt ON then the In- is pulled low.
Therfore the opamp output wants to drive a high.
Now the diode gets conductive and the opamp output is fighting agains the bjt.

It is like shorting the opamp output to gnd.
Avoid this.

Klaus

Klaus,
Thank you for the reply
In pwm controllers like uc3843 we can pull down the output of error amp.
http://www.prisall.narod.ru/bp/3843.pdf Page 12

I understand that op-amp o/p shouldn't be shorted ,with context from datasheet will you please explain in detail why it can be done in uc3843 and cannot be done in ucc2818.
Thanks

 

[KlausST]

Hi,

In one picture the opamp output has a weak drive to high (1mA current source)
This can be overridden by external circuit.

The other picture seems to work like a regular opamp. Pulling the outpu lo is like shorting the output.

Klaus

 

[txbob]

Hi,

In one picture the opamp output has a weak drive to high (1mA current source)
This can be overridden by external circuit.

The other picture seems to work like a regular opamp. Pulling the outpu lo is like shorting the output.

Klaus

Klaus,
Thanks for your time and effort
In order to make achange in the design I hve to prove that the current way of puuling it down hurts the chip(ucc2818) so..
My last question is ,is there any where in UCC2818 datasheet you can see that the voltage error amplifier o/p is not current limited.. or anythong to show that there is no inherent protection within the IC against the short.....or is this something I should get from the manufacrurer (texas instruments)
Thanks again