Should i try using an ultracap?

Hi Mark,

Difficult problem. This is what I would try/do.

Energy storage on the low voltage output of the power supply is difficult. You need to have a very large tank (a battery would do that) but if you want to run the application from a capacitor you have serious problems. You talk about 12V / 600mA / 30seconds. That is 216J. A really large amount of energy. Probably your application is sensitive for voltage drops, so a discharging 12V capacitor (12V will decrease slowly) will cut out the application before the capacitor is drained.

You can try the following. Buy a power supply with an PFC circuit in. This is an input voltage booster which charges the primary elcap to 400V, independent from the input voltage. This circuit is standard incorporated in 75W+ power supplies. Storing electrical energy on the primary side is in your case 1000 times more efficient and... the power supply does not mind that the 400V capacitor is discharging slowly under mains voltage removal. The primary controller adjusts its regulation setpoint to keep the output voltage steady on 12Vdc. You can add extra capacitance on the 400Vdc bus if desired. Additional you should examine if your application should run with 100% functionality or if it is allowed to switch off certain circuits (to lower the 600mA consumption). You can easily make a mains dip detection circuit on the secondary side of the power supply. This can act as a signal to switch off certain circuits.

For example: in LCD-TV's the backlighting is cut-out in case of mains dips, this is 100W less load... It can be substantial.

Maybe this can help you.

Best regards, Enrico Migchels

A PFC smps! of course! that's brilliant!

Thank you for your help!

any advice on which one to buy? ...

Hi Mark,

Buying a SMPS with PFC circuit alone won't be enough. You have to tweak both the application and the smps.

If you need 12V @ 0.6A for 30s this means 216J (240J transferred to the primary input of the smps, efficiency 90%)

If the smps is able to be in regulation at 100VDC this means that the capacitor voltage is allowed to drop from 400V to 100V in 30s with a constant load of 8W. This would mean that you need a PFC elcap of about 5mF. This is a very big and expensive device :-)

If you detect mains failure with an opto-isolator (very simple circuit) and you can cut-out a large portion of the load the value of the PFC elcap comes down.

For example: load consumption in mains failure mode 100mA. This would mean 1.3W input power. The capacitor value comes down to 800uF. And so on.

I would say: focus on the application first and then.. focus on the SMPS.

Best regards, Enrico Migchels

I've talked to my employer, and we've bargained back and forth. when he realized how expensive this setup would be he allowed me to go down to 10 seconds instead of 30.. the 600mA is a bare minimum i'm afraid. But this still reduces the Joules needed by 2/3, so i'm quite pleased with this. what do you think? .. can it be done?

For the PFC solution, you need to add substantially to the primary storage capacitor value. This will of course, mean that it no longer operates in PFC mode, as it is charging the caps to full voltage all of the time i.e. the charge current will not be sinusoidal. A battery really IS the way to go. 'Supercaps' are going to present a voltage that falls immediately once power is removed.

I agree with Enrico here - you really have to tailor the application to be as 'backup friendly' as possible before trying to power it. You also realise that a $65 UPS in front of the PSU will do all of this, if you send it a shutdown signal after 10-30 seconds?

I hate 'cost is no object' projects, because if you cannot satisfy the client's bizarre requirements, it reflects upon one's abilities as an engineer. Many times it is simpler to say "no, it cannot be done". In your case, it's more a matter of "you should have thought about this before you tried to get me to bypass any semblance of common-sense and try to break the laws of physics"!

Hello Mark, usually the best way is place an small battery in the output for get the perfomance of 30 seconds hold up time for output load current requeriment, I think the same of Jonathan.

The ULTRAcaps or SUPERcaps are complex for charguing and need best way form constant current method. It depend on your controller for PWM control, can make overload protection and can see hiccup mode due the current demand the ultracap is high if not are controlled by any IC.

Regards,

José Antonio

Hi Mark,

Lowering your holding time to 10s would mean an capacitor of about 1500uF/400V. Still a large container. You could bring it down to 1000uF if the output voltage is somewhat higher (450Vdc), but that would mean a lot of tweaking....

A pitty that the 600mA is not open for discussion.

Jonathan, A PFC-controller does not mind having a large capacitor at its output. The input current of the circuit will always be nicely matched to the input volatge.

Best regards, Enrico Migchels

Yes, you can try the solutions with the ultracaps.

For your spec, I=1A, dt=10s, for example dV=15V-12V=3V; following dV=I*dt/C, you could use six 2.7V/25F ultracaps in series for your project.

And there is dedicated charger IC for ultracaps in the market.

These respones have all been very informative, and i see that i'm not the only one who's new to this exciting field.

I made som PFC + large cap calculations , and also considered the six 2.7V/25F cap solution, and when i handed the considerations to my client, he started groaning. And all of a sudden economics are an issue

So now it's back to a li-po cell with a simple charging curcuit.. heh.. so if i hadn't learned a lot, whis would have been a waste of everyone's time ...

thank you all...