UCC28064A: UCC28064A Recommended Operating Ranges

Hello Yoon, 

You pose a difficult question to answer with a single definitive reply.

Although the two statements that you find in other datasheets may not be explicitly found in all TI datasheets, in general these statements can be considered to implicitly apply also to the TI Recommended Operating Conditions table values.  

 In general, parameters higher or lower than recommended may result in operation outside of the normal expectations for the control circuitry and may or may not overstress internal circuitry.  Circuits may be forced out of their linear ranges, for example. In some cases, operating outside of the recommended values results in relatively benign misbehavior; in other cases, the consequences may be more severe.  I do not wish to address each line item in the table shown above since it will take too much time to go into all of the details. 

Basically, the recommended limits are established for good reasons and we urge designers to stay within them. We cannot support designs that exceed these limits other than to identify any such situations to be corrected, even if the design appears to be "working" okay.  

Regards,
Ulrich

Hi.

Thank you for your kind reply.

I have an additional question.

I understood that you meant that you should not use values ​​exceeding the Recommended Operating Conditions.

Please carefully look at the HVSEN input voltage value in Table 7.3 of the Datasheet.

The maximum value is 4.5V.

That means, if a value exceeding 4.5V is input, the IC may malfunction and TI recommends not to design it this way.

Please look at the Function Block Diagram in 8.2.

The value input to Pin8 is to be compared with the value 4.87V/4.67V.

4.87V is designed for Over Voltage Protection (OVP).

So the value input to Pin8 can be up to 4.87V.

This value contradicts the HVSEN input voltage (max 4.5V) in Table 7.3 of the Datasheet.

If 4.87V is input for a very short time and OVP operation is performed, the IC can withstand the momentary stress.

However, if 4.86V is input continuously instead of 4.87V, what will the IC do? (Assume there is no tolerance in the input value.)

I think the HVSEN input voltage (max 4.5V) value in Table 7.3 should be modified to a larger value than 4.87V.

I will wait for your reply.

Thank you.

Hello Yoon, 

I agree that you have uncovered contradictory recommendations regarding the HVSEN input voltage. 
Obviously, the HVSEN function cannot operate unless the voltage is allowed to exceed 4.5V.

I believe that this maximum recommendation value is intended to prevent nuisance triggering of HVSEN shutdown during normal PFC operation, given that the minimum threshold for HVSEN can be as low as 4.64V.  Although well-intentioned, this recommendation can be confusing if the phrases 1 and 2 are strictly applied.  I think this confusion can also apply to several of the other recommended limits, where exceeding them may not result in damaging conditions, but simply could be considered poor design practice.   
And some limits may be applicable to the "usual" PFC design ranges, but not applicable to special-case design ranges.

It appears that I will have to address some of these recommendations for this part after all. 

First: HVSEN can definitely accept >4.87V without ill effects, up to the 7-V absolute-maximum rating.  At that point, continuous application of 7V may trigger continuous leakage of the ESD protection for this pin and damage it from excess power dissipation. 

Other limits where the recommended limits may seem to be contradictory, confusing, or arbitrary (in my opinion) are:
1.  ZCDA, ZCDB series resistor:  These resistance limits (20kR to 80kR) are suitable for expected ZCD-winding turns ratios for the usual "universal" line voltage range, but may not be appropriate for special-case input voltages.  The true limit would be ZCDx pin current (ABS MAX = -5mA) and I'd recommend to choose a resistance to limit the current to -3mA maximum.  
2.  PHB threshold voltage: (0V to 2V) Since maximum output power is delivered when COMP = 5V, the 2V recommended limit is intended to avoid having a single phase deliver more than 50% of the maximum power, with some margin for tolerances.  2V seems to be a reasonable limitation, but this arbitrary limit leaves no room for higher values when the designer may have a special case.  
3.  BRST threshold voltage:  (0V to Vphb-0.6V) This is intended to shed a phase first before entering burst mode, to avoid bursting with 2 phases running. 
Burst-mode with 2 phases may not be ideal operation, but I don't foresee any functional stresses or reliability issues in this condition. 

The remaining parameters of Table 7.3 that I did not address have appropriate recommended limits, which I believe should not be exceeded or unreliable operation may result. 

Regards,
Ulrich

Hello.

Thank you for the detailed explanation of the various pins.

However, I am not interested in the good intentions of the manufacturer.

I only trust what is written in the datasheet and ask the manufacturer for help when needed.

I cannot agree with you.

Customers trust your datasheet and design based on it.

If there is no information about Normal-Case and Special-Case in the datasheet, the customer does not know about it.

How can the customer know that there is no problem if they design beyond the limits?

You specified the ranges in the datasheet and you confused the customer to design beyond those ranges.

Samsung Electronics is already using this IC in many products and it is designed to input more than 4.5V.

If there is a problem in the product, you may also be responsible for proving that this IC is not a problem.

I suggest the following.

1. Your apology for using contradictory expressions regardless of your intention.

2. Add a statement to the datasheet that it is okay to use up to 4.87V on that pin.

I will wait for your reply.

Thansk.

Hello.

Thank you for your reply.

I have no intention to raise this issue within Samsung.

Because it has already been used and produced in many products.

For the continued trust and partnership between Samsung and TI,

I expect TI to take prompt action.

And I hope Samsung engineers will design without confusion in the future.

Please let me know TI's decision as soon as possible.

Thank you.