I stumbled across this... A 330uF 200V flatpack capacitor.
It's 1/2" thick. And also costs over $100 ea (mouser) Yikes!
I need flatter...
I'd like to know if I can get 150uF @200V crammed below 10mm without resorting to paralleling many capacitors. Can be axial, radial or SMD. Any mounting orientation is ok.. Currently Googling and browsing Digikey...
But surely three 47 uF caps in series gives an effective capacitance of one third of 47 uF, that is 15.7 uF. Also, connecting electrolytics in series to achieve greater working voltage than the rating of a single cap may require voltage-splitting resistors in parallel with the caps to avoid any one having to bear more than its rated voltage.
UCC/NCC used to offer a high voltage FTK series that was radial lead in an oval package - still 12.5mm minor dimension. Don't see these advertized anywhere, now. 10mm can often be wangled to 12.5 by removing printed wiring in the viscinity of the part body.
Flatness is often a bullshit requirement - research it carefully.
What's the actual requirement? Filtering? Energy storage? Hold-up time? These can often be adressed more intelligently than by just adding uF.
I've fitted caps (and other parts) within routed board cutouts to save extra height. If you did this with a flat RB radial cap in a board cutout you could use up to say a 19mm diameter cap at any length.
UPZZD820MPD ?? I tried that in the search box on the Nichicon site.. Nadda.. Mouser...nadda Digikey...nadda Back to Nichicon site using the Parameter Search...82uF sorting by diameter. Ugh... it's UPZ2D820MPD... Z2 not ZZ
I guess I could grind off some of the plastic coating to get under
10mm. Huhhh.. I have to do a lengthy analysis to see if 82uF will be ok.. I might have to buy a faster computer just to simulate the effects of using less than 150uF.
I suppose it would be hold up time. If the ripple current is too high my smps drops out of regulation.
150uF was the min. capacitance determined. Also, and this is something I have to do homework on... I believe the more ripple current, the shorter the capacitor lifetime. My app has to run for about 7 years continuously. IIRC the capacitor ripple current is about 0.5A peak-peak.
Nothing at that power level 'needs' to be 10mm thick. It's not hold-up time if it happens on a cycle by cycle basis - hold-up is a one-shot deal at power down.
That said, increasing the stored input voltage may reduce the part body size for the same CV^2 /2 energy storage. That's one of the features of pfc boost - it can also be a feature without regard for pfc. Simple voltage doublers get the same effect. The idea is to never let the storage energy get that low in the first place - it should only occur for brown-out, ride-through or hold-up time conditions.
Extending regulation range (lower drop-out) gets more energy out of the capacitor. At lower powers it's no big deal, so long as your 10mm thick package has sufficient surface area to dissipate the losses.
Pyrotechnics to alleviate LF dropout effects can also be performed on the secondary side, where real-estate is higher density, voltages are lower and technology is your friend.
This 10mm thick package is likely custom - there are thieving methods applicable to body wall thickness that can be used to scavenge extra millimeters for internal parts. Open up a flat battery pack.
I think there's an unstated rule of thumb with commercial electronics that the thinner something is, the quicker it gets busted, which sort of works against the 7-year aim.
I've been thinking about the form factor for electrolytics.. They tend to be more tall than wide. Is this to reduce the ESR? If an electrolytic were made like a disk say ...10mm tall x 30mm wide then the foil length is longer. I think that would make capacitor more inductive and resistive than other forms??
There are a number of factors driving the demand for electrolytic capacitor shapes:
In order to produce the first two, the sealing method must be long-lasting and practical. The rubber seals with minimal diameter perform that function in a can shape.
Temperatures are reduced if the surface area is maximized. A longer cylinder has that feature without complicating the sealing.
Parts with a single minimized dimension facilitate the third requirement. Mounted vertically in skyscraper fashion, they occupy minimal real estate. Mounted horizontally, they have a minimal profile.
Round containers also do not distort under increased internal pressure (vs square flat features). They are already optimized for material volumetric efficiency and material cost. Note that the pressure release features of cylindrical electrolytics tend to make use of this weakness of the cylinder's remaining flat surfaces (unless built into the sealing hardware), although slits are also inscribed in the sides of some types.
The flat electrolytics you were looking at are used in applications where the parts are actually clamped into high-vibration, temperature-controlled environments where economics are not the first consideration. The sealing method and low manufacturing volume dictate their price. Environmental limitations of their chemistry have always limited their acceptance in the intended hi-rel market.