Supercapacitor backup ?

Tell me why please using a capacitor like this (.047f) 5.5v is either perfe ctly fine or not a good idea on the output of a +5v power supply that power s a Raspberry PI ? The PI draws around 500ma and doesn't like power outages (just had a corrupted sd card because of one -- not fun).

In my mind, I see no problem and at least it would give me a few seconds of backup which may help.

But you guys are the design wizards -- not me. My supply is a standard swit ching supply common these days. I want to put the cap right on the output.

a cap like this --

formatting link
HyDPEj1Wng

thank you.

Reply to
mkr5000
Loading thread data ...

Quote "Typical ESR ... 7-14 Ohm".

That isn't going to work. It won't likely live long and at 500mA the ESR alone would drop the whole 5V. You need a cap that has a very low ESR and is rated at 500mA or more. Check Digikey but those are going to be pricey.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

I've never understood ESR can't get my head around it. what figure of ESR should I look for ? And it's the ESR that plays into the current capability?

Reply to
mkr5000

Think of it like this: your capacitor is in fact a capacitor with a resistor in series, in one package. Then you can calculate how it affects the current capability.

But before you dig too far into this, it has already been discussed on the Raspberry Pi group and this method is not going to work.

Reply to
Rob

Exactamente. Also, a capacitor has a certain maximum current and the datasheet should be consulted.

I don't know what was discussed there and I don't work with any Rasberry Pi. A supercap does work if designed in right. Two main jobs here:

  1. The cap must be low ESR and rated for a large enough surge current.

  1. The cap must either directly provide a rail where the voltage remains above the required minumum long enough or you have to hang a regulator inbetween.

#2 is a balancing act between size, cost and complexity. If using an inbetween switch-mode converter one can get away with lower capacitance and cap size for the job at hand.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

When you put the capacitor directly on the rail, it can provide only very little energy because the voltage can only drop from the rated supply voltage to the minimum supply voltage for the load. And also because the load voltage is so low.

Indeed as you say, it is much better to charge the capacitor to a much higher voltage still within range for commonly available parts, e.g. 40V, then connect it to the load via a switch-mode converter. Maybe another switch-mode converter is required to provide the charge voltage, when the whole thing has to be powered from an existing supply.

It increases complexity but it decreases size and probably also cost. (because energy in a capacitor increases with the square of the voltage)

Reply to
Rob

On Sat, 09 Apr 2016 09:12:25 -0700, Joerg Gave us:

Better to find a place on the rpi itself to incorporate such a "battery", which is what this usage of the "supercap" is being used as.

Fluke uses one to keep the date and time up on a meter that has dead batteries or has the battery pack removed for a battery change out.

So it could be used to keep some critical element of the rpi computer up between a power supply "outage", but that would have to be addressed fairly quickly and so the device would likely lose power anyway.

Mil computers still incorporate a 5 volt standby supply rail and spec that has to stay up long enough to allow for a core dump. Some 150 ms. And that is why there is a separate 5 volt rail in a COTS AT supply spec.

Laptops sense a low battery and the OS can be set to sleep mode the thing before shutting down. An Apple iPad shuts down (safely)at 5% battery threshold no matter what.

Reply to
DecadentLinuxUserNumeroUno

CV = IT is the linear approximation.

So V = I*T/C. Plug in 47mF and 500 mA and you get a droop of 10 volts per second.

Plus the ESR issue that Joerg mentions.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

I'd be very surprised if there was any such terminal. The OP mentioned an SD card write crash. Preventing that requires a bit more energy than a few microamps for an RTC. Essentially you have to keep supplying the whole engine until a write precedure can be safely terminated.

I've done that without separate rails.

Yes, sensing is key. One method is to let the "replenisher switcher" operate at a slightly lower voltage but well within spec. The system then sees this drop and prepares for landing.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

fectly fine or not a good idea on the output of a +5v power supply that pow ers a Raspberry PI ? The PI draws around 500ma and doesn't like power outag es (just had a corrupted sd card because of one -- not fun).

of backup which may help.

itching supply common these days. I want to put the cap right on the output .

_mHyDPEj1Wng

this group is so great -- one of the few that endure since the days of deja news and beyond. you guys are my go to. learn more here than hours of goog le chaos.

thanks.

Reply to
mkr5000

The energy stored in a capacitor is proportional to the square of the voltage.

With 0.47 F across the +5 V line, the energy stored is 0.59 J.

Assuming your board will work down to +4 V, the capacitor stored energy is 0.38 J Assuming lossless capacitor, the energy used when going from 5 to 4 V is 0.59-0.38=0.21 J. Your power consumption is 2.5 W so the capacitor will help for 85 ms or 4-5 mains cycles in a lossless case.

If you intend to use a capacitor as a backup store the load must tolerate a large voltage swing (such a small switcher), such as 2:1. The energy stored at 1/2 initial voltage is 1/4, so the maximum energy that can be extracted is 3/4 of initial energy.

Thus with +16 V nominal voltage across the capacitor and the second switcher letting the voltage drop to +8 V would be fine. Extending the low limit voltage further down is not productive, since the energy below 8 V drops very rapidly.

The practical situation would be even worse, since the second switcher would consume even more current at low voltages in order to deliver the same power to the final load.

Reply to
upsidedown

Not a good idea. Those old capacitors have an impedance at least 10 Ohms. You need the expensive carbon aerogel caps.

Ultra Capacitors have very little power available to devices with a narrow operating voltage range. When I've used them for solar devices, I have a them on the output of a synchronous buck inverter. For example, the inverter maps 4-6 volts of power rail to 0-5.2 volts in two ultracaps using a modified feedback circuit. Being a synchronous inverter, it can move power in both directions with some current derating. This mapping prevents power rail overloads and makes much better use of the capacitors.

--
I will not see posts from astraweb, theremailer, dizum, or google 
because they host Usenet flooders.
Reply to
Kevin McMurtrie

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.