Dual BCM Boost PFC chip is a poor choice?

[cupoftea]

Hi,
This is a Interleaved dual BCM PFC chip. (FAN9611)

https://www.onsemi.com/download/data-sheet/pdf/fan9611-d.pdf

Surely this is a waste of time?

On the first page of the datasheet it should say.............
1...Cannot be (cheaply & realistically) used with powder core inductors
2...Needs two Gapped_core/bobbin/clamp sets to be assembled for the 2 boost inductors (with all the associated assembly costs)
3....Needs two gapped cores for the ferrite cores which are a pain to source and are more expensive with long lead times.

So, Both stages need an auxiliary coil on the inductor to indicate the zero crossing.
This means powder core torroids cannot realistically be used (due to coupling concerns) , and a full gapped-core/bobbin/clamp set
needs to be used.
This makes the product assembly more awkward, and it is simpler to use two Powder-core ACM_CCM Boost PFC stages
in parallel with each having 2 current sense transformers to reconstruct the ground return current.
They can be sync'd to be interleave switching.
A simple share circuit on the output will assure equal current sharing.

Not to mention that any design with gapped cores is more awkward and expensive to kit for than one
where there are no gapped magnetics.

One could use the UCC28070 dual ACM_CCM controller but every one i ever ordered arrived with snapped off
pins or ESD damage....so two sync'd ACM_CCM PFC controllers with total of 4 current sense transformers
is the way to go. A current sense transformer is extremely simple to wind...can be done on a small split bobbin
with pri in one section and sec in the other.

The sharing circuitry for the paralleled ACM_CCM Boost PFCs can comprise one stage with
divider regulating it to 410V....the other to 390V...then an error amp pulls current through an extra upper divider resistor
(in the "390v" stage)to equalise the output currents
of each booster.

 

[ahsrabrifat]

If you are continuously facing ESD damage while using UCC28070, You can use NCP1632 as an alternative.

 

[cupoftea]

Thankyou but NCP1632 again needs the auxiliary winding on the torroid....and you cant realistically have an auxiliary coil on a torroid inductor...its coupling will be too poor.
UCC28070 is literally the only ACM_CCM Boost PFC chip in the world today.

For reasons above we cannot use it.

The alternative use of two separate anti phase sync'd ACM+CCM PFC chips is not great , but its the best there is.......even though it needs all the following auxiliary circuits with it.....

Anti-phase synch circuit
Output current share circuit
4 current sense transformers (2 for each booster)
A duty cycle clamp circuit so max duty is not too high for the CST's.
___ ____ _____________________________________
The UCC28056 is a BCM PFC chip, though unfortunately not dual_interleaving. As can be seen on page 45 of its datasheet (below)....they dont use an auxiliary coil to sense inductor discharge, and so they have used a torroid inductor for the boost inductor....(as you would, since a torroid is far cheaper than a gapped_core_set_and_bobbin.)

https://www.ti.com/lit/ds/symlink/ucc28056.pdf?ts=1738941571290&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FUCC28056

 

[Easy peasy]

FAN9611 & 9612

we did a PFC front end with this, variable Vout to match mains Vin, 2 x RM10i chokes, 2.2kW, seemed to work very well, except it is prone to asynchronous noise

and the no load operating freq is fairly high - otherwise works OK, very low switching losses in the mosfets and diodes - such that we could use slow but low Von diodes ( nice eff improvement there )

 

[cupoftea]

Thanks, another way, similar to that 9611 way , is to just use two paralleled ACM Booster PFC stages...and just design it so its in shallow continuous mode at mains peak, max power.....this way, you benefit from a very much reduced reverse recovery than deep CCM (not as good as your BCM but perfectly satisfactory) . Plus you may use powder core torroids...and no need to bother with the bias coil and gapped ferrites and core sets and bobbins and all that paraphinalia.

 

[Easy peasy]

you can use iron powder as the chokes without issue - as the uH simply go down a bit with increased current - this does not unduly affect the controller - adding a small TIW wdg to a toroid iron power boost choke is pretty easy too.

 

[cupoftea]

adding a small TIW wdg to a toroid iron power boost choke is pretty easy too.

Thanks, are you referring to adding an auxiliary coil to an iron powder torroid?...for the purposes of doing zero crossing detection as part of BCM operation?
The coupling will be poor....compared to with a core set and bobbin.

 

[Easy peasy]

The coupling will be just fine - no matter how you do it - that is the nature of toroidals - almost all of the flux circulates in the core and therefore inside any windings on that core - also as this aux wdg is lightly loaded - you still get a good look at the signal - even if the coupling was marginal.

 

[cupoftea]

almost all of the flux circulates in the core and therefore inside any windings on that core

Thanks, though i was just thinking of common mode chokes, with separated coils, and how the "leakage" inductance can be quite significant...but i see what you mean, even if not well coupled it should still work.

 

[mtwieg]

From what I recall, these BCM PFC converters are based on the assumption that average input current is linear with on time and peak current (looks like this chip modulates on time, not peak current). Therefore so long as the inductors are very linear, you can get very low THD without needing any ACM feedback and its associated limitations (especially at very light load).

So using ungapped core materials with gradually decreasing permeability (as is often the case for iron powder toroids) may be a problem for this type of controller. IMO this could be seen as a downside of iron powder, rather than a disadvantage of this controller. Iron powder relies on ACM feedback to get very low THD, while normal ferrite does not (so long as you stay away from Bsat).

But I'm betting you could use iron powder toroids with this controller, would just have to choose Bmax carefully (probably lower than you would with a CCM/ACM controller).

I don't see coupling as being an issue with the aux winding. The datasheet doesn't describe the impact on coupling, but I doubt it would matter much since you aren't transferring any significant energy.