What is the accepted limit of power density?

treez said:

What is the accepted limit of power density (in W/cubic inch) for a SELV (all voltages inside <42V) SMPS without a fan, but has potting around all the electronics, and has a metal heatsink.?
Is it about 100W/cubic inch?
Ambient temperature 85 degrees C max

(My 230VAC laptop power supply is 65W in 7 cubic inches. (10W/cubic inch))

Click to expand...

I understood that the question was referring to "power density" as the internally dissipated heat which would need flow across the surface of a certain volume. It was presumed due to statement that no cooling fan is allowable.

However, taking a better look on the example you gave above, now I´m not sure if oppositely, the question may concerns to define what is the maximum energy per volume that can be managed for current SMPS equipments.


+++

Basically , for a certain Wattage, and SELV voltages throughout, how big does an SMPS have to be?
1...With heatsink and fan cooling made available.
2....With heatsink but no fan , but potting compound available.

I think this "power density" is going nowhere. Its the power dissipation that the problem. Taking the 100 W smps, if it only dissipated 2 W, then 2 cu. ins would do, providing you can squeeze all the components in. If it dissipates 10 W, then 2 Cu ins would give a high temperature with only free convection cooling, bolting it to a heat sink would reduce the temperature to a reasonable level providing the internal thermal paths are properly designed, perhaps with a thermally conductive encapsulation.
Frank

Flapjacks graph shows that this is a real concept.
If a customer asks you for a PSU, then at this initial time, you have to be able to give them a rough idea of how big it will be, before you have had the chance to design it.

The diagram is comparing apples to oranges. You can't compare fan and convection cooled devices without specifying the enviromental conditions.

The comparision makes (limited) sense if you look at specific devices, e.g. notebook adapters. As already pointed out, the surface rather than the volume counts for the notebook adapter. But as all notebook adapters have similar size, people know what you are talking about. Extrapolating the number to a different order of magnitude, e.g. 650 W just fails.

Just look in a catalogue, and use some ones elses sizes, unless you can originate a new low loss technology.
Frank

As far as I know there's no standards for speccing power density, since it depends hugely on your test conditions and how you define the volume of the supply itself. Looking over recent literature, I see many works that require fans but don't actually include that in the volume calculation. And even if there are fans, you have to take into account how much clearance the intake and exhaust ports need. And of course most of these tests are done with the thing sitting on a lab bench, rather than tucked inside a rack or case. And I almost never see numbers on the ambient temperature range for which the power density holds (I doubt any of the >200W/in^3 examples I've seen in literature are tested at industrial temperatures).

So it doesn't really make sense to compare apples and oranges, unless you find an example which is extremely close to your own application.

https://uk.farnell.com/tdk-lambda/ls200-15/psu-enclosed-15v-13-4a-200w/dp/1995940

..the above are off the shelf power supplies, if every application was massively different then nobody would ever make such general_for_sale power supplies.
I think we can safely say that nowhere in the world will anyone find one of these type of units with a power density above 40W/inch cubed.
The one above is 4W/cubic inch.

Whilst engineers may demand absolute specs for themselves, managers and marketers talking to actual customers speculatively need to have general rules of thumb....and Watts/cubic inch, as a rough guide, is a useful thing, even of absolute conditions aren't stated.
Generally speaking one needs a W/inch cubed figure to pull the customer in with,then when theyre in, adjust it as per the actual condicitons......if nothing was ever stated in the first place for power density, then the customer will never come to you.

treez said:

https://uk.farnell.com/tdk-lambda/ls200-15/psu-enclosed-15v-13-4a-200w/dp/1995940

..the above are off the shelf power supplies, if every application was massively different then nobody would ever make such general_for_sale power supplies.
I think we can safely say that nowhere in the world will anyone find one of these type of units with a power density above 40W/inch cubed.
The one above is 4W/cubic inch.

Click to expand...

Sure for a modular, general purpose power supply your density is going to be very low. Especially a TDK Lambda, which are very robust. Are you building a general purpose power module? If you're building a specialized DC-DC converter on a PCB, then you should get way more than 4W/ci, and probably more than 80% efficiency as well.

Hell, I've built small SMPS for specialized applications which get around 100W/ci with no forced air. But that doesn't mean I could get the same number for an offline converter. Completely different rules at play.

Whilst engineers may demand absolute specs for themselves, managers and marketers talking to actual customers speculatively need to have general rules of thumb....and Watts/cubic inch, as a rough guide, is a useful thing, even of absolute conditions aren't stated.
Generally speaking one needs a W/inch cubed figure to pull the customer in with,then when theyre in, adjust it as per the actual condicitons......if nothing was ever stated in the first place for power density, then the customer will never come to you.

Click to expand...

I always insist that a customer give some bare minimum of information before I suggest any specifications. For power electronics, the operating environment is always one of the first things I insist on knowing. If they can't say, then I tell them I'll wait patiently while they figure it out. How would you like it if you learned far into the prototype stage that your supply needed to be in an ip67 plastic enclosure?