Flyback SMPS control chip datasheet does not give value for internal pullup resistor?

[treez]

Hello,

I was just admiring the Sanken integrated FET flyback controllers when I realised that the datasheet has insufficient information to make them actually useable. This is because the datasheet does not describe what is the value of “R(pullup)” in the controller.

As you know, “R(pullup)” is the value of pullup resistor used in the controller through which the feedback opto pulls current through in order to regulate the flyback smps. “R(pullup)” can be seen in the linked document called “Designing with the TL431”, and is called “R4” throughout this document..

Designing with the TL431:
**broken link removed**

STR-W6000S series integrated FET flyback controllers:-
http://www.semicon.sanken-ele.co.jp/sk_content/str-w605xs_ds_en.pdf


Do you know what is the value of R(pullup) for any of the STR-W6000S range of flyback smps controllers?

 

[Easy peasy]

from the data sheet the max feedback current is typically 230uA (150 - 340uA range) with 7V drive, so the effective resistance is obviously quite variable, from (7v - 0.5V / 340uA) to 7V-0.5V / 150uA) = 19k117 to 43k333

Which is perhaps why they don't state it explicitly...

you need to design for worst case.

 

[treez]

Thanks, I see your point, indeed that does seem to be what the value of the R(pullup) should possibly be…however, there is not a pullup “resistor” there, instead, as you allure, it is a “current source”….and as you know, current sources have infinite resistance, so I actually don’t believe that an “equivalent” resistor such as that that you have calculated is applicable here..surely you agree?

Another point is that they do not give the attenuation ratio of the potential divider which must exist inside the STR-W6000S range of controllers. This divider is the one that divides down the voltage at the FB pin so that it can go into one input of the PWM comparator. After all, the peak current detection threshold is 0.78V, so when the feedback pin is at a few volts it needs dividing down before sending to the PWM comparator.
The STR-W6000S datasheet doesn’t even show what attenuation ratio this divider has, and this information is needed for the feedback loop calculation. Therefore, the STR-W6000S control chip simply cannot be used until this information is forthcoming from Sanken.

-For an example of the divider inside a control chip needed to divide down the voltage at the FB pin, the diagram on the first page of the UCC28C42 datasheet shows what I am referring to….
UCC28C42 datasheet
https://www.ti.com/lit/ds/symlink/ucc28c42.pdf

 

[Easy peasy]

if you look carefully they show a resistor and a current source...

 

[treez]

thanks, yes, i see that resistor, but nowhere in the datasheet does it explain what value that resistor is, or even what it is for. The current source , next to that resistor is the only thing that the datasheet relates to the FB pin feedback loop operation, and it says it has a nominal value of 230uA.

And as you know, they havent given the feedback pin divider voltage ratio, so from that point alone a feedback loop calculation cannot be done.
These chips get used in Panasonic televisions, and i wouldnt mind betting that they dont actually care that outsiders to panasonic cant work out the feedback loop for these chips, because they dont want their competitors to know about these things. Sanken is closely allied to Panasonic.

The STR-W6000S range of controllers is unusable without these bits of information, which Sanken obviously knows, and could easily divulge, but is not doing so.

 

[Easy peasy]

the resistor and the current source each form part of the 150 - 340 uA available at the pin

while the specs are loose - it does not render the chip un-useable - plainly, as it is in common use, the divider ratio is pretty obvious on close inspection, as per usual and assoc feedback loop needs to be slow and damped to give acceptable performance over the tolerances of the IC...

 

[treez]

Thanks, but with the STR-W6000S controllers, the internal divider ratio is not at all obvious. It is obvious in other chips such as the UCC28C4X range or the NCP12XX range, but not in the STR-W6000S range.

From the STR-W6000S datasheet, all one can possibly know is that the “regulation range” of the FB pin is somewhere between 0.95V and 8.1V. Goodness knows how far below 8.1V you have to go before you get into that range of FB pin voltages which result in adjustment of the peak current threshold.
The datasheet simply does not say.
They may be in use, but if so, then they either have “insider knowledge”, or are just “jacking it” and being lucky with their feedback loop.

- - - Updated - - -

I(FB) min is actually 7uA, and I(FB) max is 340uA , so its an extremely huge range of currents.
This seems a bit strange, a current source is supposed to be the same value all the time, (unless stated to be time variant, which this one isn’t) so why is it varying so much?, and which value should we take?, even if we could assume that we can use it to calculate an “equivalent” R(pullup) resistor…which , in fact, I don’t believe we can because a current source is just that, and has an infinite resistance (or very very high resistance).

- - - Updated - - -

And its worth noting the absolute lack of any kind of supporting App Note for the STR-W6000S range of chips. This to me, says one thing, "Closed Shop"....enter at your own risk.

 

[treez]

OK, Easy Peasy has called it right again, Easy Peasy says that the effective R(pullup) resistor value is gotten by doing an “Ohm’s law” style V/R calculation, and page 12 of AN-4150 confirms that this is indeed correct (AN4150 concerns the FSQ0265 flyback controller but the situation is the same as for the STR-W6000S controllers)

AN-4150
https://www.fairchildsemi.com/application-notes/AN/AN-4150.pdf

FSQ0265 datasheet:
https://www.fairchildsemi.com/datasheets/fs/fsq0365rn.pdf


Now this “V/R” thing works for the FSQ0265 controller because it has a well defined (well toleranced) “IFB” value of 900uA.

However, the STR-W6000S flyback controllers have an IFB value which ranges anywhere from 7uA to 340uA. -That is just too wide a range from which to calculate any kind of sensible value for R(pullup).
(And again, the STR-W6000S datasheet doesn’t give the attenuation factor of the feedback pin voltage inside the chip)

So again, the STR-W6000S range of controllers are not reliably useable given the lack of information in their datasheet.

 

[Easy peasy]

If one had to use the chip and was worried about the low current (or variable current) into the attached optocoupler collector, then one could put an external resistor (or current source) limited to 7V drive, and feed this to the pin also, say 1.5mA, this would largely swamp the 340uA max of the chip...

- - - Updated - - -

because of the 7V line used on this pin, one can guess that the max pwm is likely obtained in the 5-6 volt region - and would be easy enough to measure in a test jig.

- - - Updated - - -

Minimum feedback current, presumably for maximum o/p, is in the range 7 - 30uA, drawing less than this will cause the protection circuits to kick in...

- - - Updated - - -

C3 on the opto, limits the dynamic response pretty severely, so any volt loop controller will have to be pretty slow and damped, this is not intended as a fast dynamically responding power supply...

 

[treez]

Problem now solved also by Sanken, who have replied (twice) reference the STR-W6000S internal pullup resistor , and the divide down ratio.
…………………………………………………………………….
Dear Sir,

Thank you very much for your inquiry.

The pull-up resistor value for the STR-W6000S is 27kΩ.
Also, the feedback pin voltage gets divided down internally.

Best Regards,
Sanken Electric Co., Ltd
…………………………………………………………………

Dear Sir

VFB is divided by the internal resistor at 1:3 ratio.
6.2/4=1.55V and OCP voltage (0.78V) is compared at the comparator.


Regards
Sanken Electric
……………………………………………………………..