supercapactiors

[Joeledmund]

i think if i make a bank of 5.5V 1F caps on a power cord could have a small battery backup unless if someone has a better idea i just dont want to solder on my pi

[jasonclark]

I've got a lithium battery pack for charging my phone. This has a built-in microusb and can source 2A.
It can also take a charge from a MicroUSB charger...
From Amazon my 9000mAH unit was ~£25

I dont think capacitors would work very well without a boost regulator, once the supply drops to 4.6V, the PI will become unstable very quickly.
If you can get a DC/DC converter fed from these, terminated to a MicroUSB, it would work.

[techpaul]

You would probably want some diodes and resistors to control the charging that large a bank of capacitors, otherwise you might have to wait some time for them to charge up, and will take longer for Pi to shut down and be careful of stored charge, you may need a bleed resistor across the capacitors to slowly discharge them to avoid sparks.

Remembering dealing with largish switched PSUs larger than PC PSUs and worrying about removing the caps on them as screwdriver ends used to regularly vapourise.

[Burngate]

You'll need to work out how many 1F capacitors you're going to need, and that means how much current for how long, and with what voltage drop.

Q=CV is the starting point for capacitors, and Q=It for your current

So It=CV

I for the Pi is about 750mA - use 1A as that's easier to calculate with, and gives you a safety margin
The Pi should possibly work down to about 4½v, which as luck would have it is 1v below your starting point of 5.5v - again easier to calculate with

So to find C all you need is t - how long (in seconds) do you want it to last?

[Joeledmund]

well i only need it to last at the longest is like 10seconds no power so when i use in car i can restart car with out restarting computer (like running car on battery to turning on engine) the two capacitors worked for a millisecond then video went out out but power LED stayed bright for like 20seconds so 2F is not enough i dont need a diode because they are dc and only flow one way someone on line did solder a 1.5F straight to the p1 and p2 on the pi and theirs is fine i just dont want to solder on the pi so i am using a female USB and a male USB kinda like and extension cable

[Joeledmund]

I've got a lithium battery pack for charging my phone. This has a built-in microusb and can source 2A.
It can also take a charge from a MicroUSB charger...
From Amazon my 9000mAH unit was ~£25

I dont think capacitors would work very well without a boost regulator, once the supply drops to 4.6V, the PI will become unstable very quickly.
If you can get a DC/DC converter fed from these, terminated to a MicroUSB, it would work.

i have a battery pack too but this is a fun side project to work on for a big power hick up in my car

[techpaul]

well i only need it to last at the longest is like 10seconds no power so when i use in car i can restart car with out restarting computer (like running car on battery to turning on engine) the two capacitors worked for a millisecond then video went out out but power LED stayed bright for like 20seconds so 2F is not enough i dont need a diode because they are dc and only flow one way someone on line did solder a 1.5F straight to the p1 and p2 on the pi and theirs is fine i just dont want to solder on the pi so i am using a female USB and a male USB kinda like and extension cable

Using Diodes means you can use it like a battery backup, so only flows out of cap when PSU voltage drops. Also means Pi reacts quicker to power resumption as the capacitors slowly charge up again, also reduces startup surge in current to charge them at startup, reducing stress on PSU.

With experiments like this you really do need current probes to watch the current transients at the transitions. Either on a logger or a digital scope

[Joeledmund]

well i only need it to last at the longest is like 10seconds no power so when i use in car i can restart car with out restarting computer (like running car on battery to turning on engine) the two capacitors worked for a millisecond then video went out out but power LED stayed bright for like 20seconds so 2F is not enough i dont need a diode because they are dc and only flow one way someone on line did solder a 1.5F straight to the p1 and p2 on the pi and theirs is fine i just dont want to solder on the pi so i am using a female USB and a male USB kinda like and extension cable

Using Diodes means you can use it like a battery backup, so only flows out of cap when PSU voltage drops. Also means Pi reacts quicker to power resumption as the capacitors slowly charge up again, also reduces startup surge in current to charge them at startup, reducing stress on PSU.

With experiments like this you really do need current probes to watch the current transients at the transitions. Either on a logger or a digital scope

ok then what type of diode would use like a 1amp zener , or a rectifier or blocking i dont want a bleed resitor tho if it will reduce efficiency but the diode would is on the input or output side

[techpaul]

Using Diodes means you can use it like a battery backup, so only flows out of cap when PSU voltage drops. Also means Pi reacts quicker to power resumption as the capacitors slowly charge up again, also reduces startup surge in current to charge them at startup, reducing stress on PSU.

With experiments like this you really do need current probes to watch the current transients at the transitions. Either on a logger or a digital scope

ok then what type of diode would use like a 1amp zener , or a rectifier or blocking i dont want a bleed resitor tho if it will reduce efficiency but the diode would is on the input or output side

You need Schottky diodes for minimal Vf drop, at least 1A rating, with Vr at least 10V (easily achievable).

If you have two diodes, one on output and one on input you control current flow. With small value reasonable wattage resistor you control the charge rate of the caps

This also keeps charge voltage level below 5.5V