Differences between revisions 6 and 8 (spanning 2 versions)
Revision 6 as of 2011-04-08 09:42:52
Size: 1560
Editor: ?AlexAlex
Comment:
Revision 8 as of 2011-06-11 04:47:12
Size: 4606
Editor: ?AlexAlex
Comment:
Deletions are marked like this. Additions are marked like this.
Line 36: Line 36:
== Quality ==

 * Manufacturers should give a minimum of 1 year warranty for a system running 24/7. Devices that are crap are a danger to the reputation of the FreedomBox.

== Price ==

 * The hardware should be as cheap as possible, but not cheaper...

Line 43: Line 52:


=== Power Supply ===

The Freedombox will need a lot of component that require continue power supply at specific low voltages (usually 12V, -12V 5V and -5V). But the main power supply of the box will be with the 50hz oscillation at 220V (in France) from the plug. So we need something to isolate all the hardware from electric network (that would be a shame to electrocute the guys that touch the box). And give continue power.



For PC 286 - 386 - 486 and Pentium. the power supply used was the AT power supply. It give + and - 5V, + and - 12V. It is an old one, but maybe easier to use. The switch button is on the electric network part, so it is easy to stop.

Nowadays PC use the ATX technology that give + and - 5V, + and -12V and +3.3V, and the switch button is on the motherboard part so all the system can be turned on off by software.

All those voltage are maybe not really useful. It is possible to have component that work only with + and-12V.



Those power supply work well, but they are expensive and not really studied for low power. So it is maybe possible to create our own.



here is the schematic of an ATX power supply studied for 200W. Maybe we can try to inspire ourself from that.



[[attachment:schema_alim_ATX2.svg|{{schema_alim_ATX2.svg||width=100}}]]

those power supply use the Switched-mode power supply technique. it is interesting because there are a bigger efficiency (need to be verified at low power) than linear power supply, but they are using an oscillator that will create a lot of noise with a lot of different harmonic that can be very problematic for our use.



Here is a schematic for a very simple linear power supply.


{{attachment:drawing.svg}}


The transformer is to reduce the intensity and to isolate the circuit from the electrical network (especially the ground). then we use a diode bridge to convert AC current in DC current. and then a capacitor to have a stable intensity.



I don't know if this power supply is stable enough for what we expect, that mean supply component without burn them if there are a jump of current in the electric network.

Maybe that need to be improved before using it.



==== What still need to be done: ====

that would be interesting to have a complete modeling of the system with a non ideal transformer, to know what is the efficiency, and what is the component parameter that we need.
Especially for the transformer and the capacitor.

We can add some other component to make some filter to increase the stability of the system.

Maybe it can help to add an other transformer and diode bridge in the linear system. And add signals with a Pi/2 dephase difference (just before the capacitor) to increase the total efficiency.

Hardware Requirements

/!\ INF: Please help extending the information in this page

The hardware must/should/may ...

Non-functional requirements

  • It must have at least one serial interface
  • It should have at least one JTAG interface
  • It must have at least one network interface
  • Noise level must be below 20 dB
  • It should have over 256MB RAM
  • It should have over 512MB Flash for filesystem
  • It must have storage interfaces, at least one of them:
    • SATA interface
    • USB interface
    • SD/MMC interface

Performance requirements

  • It must be able to execute all Freedom Box software at less than half of the CPU load

Power requirements

  • It should not consume over 15W at full CPU load

Physical requirements

  • It should be reduced in size, not bigger than a set-top box.

Environmental requirements

  • It must work perfectly between 0-40ºC

Quality

  • Manufacturers should give a minimum of 1 year warranty for a system running 24/7. Devices that are crap are a danger to the reputation of the FreedomBox.

Price

  • The hardware should be as cheap as possible, but not cheaper...

INF: At the first version we do not need to be too ambitious, rugged and harsh environment as well as renewal energy powered devices are desired but not at this point

Design of the system

attachment:schematic-of-the-box.png

Power Supply

The Freedombox will need a lot of component that require continue power supply at specific low voltages (usually 12V, -12V 5V and -5V). But the main power supply of the box will be with the 50hz oscillation at 220V (in France) from the plug. So we need something to isolate all the hardware from electric network (that would be a shame to electrocute the guys that touch the box). And give continue power.

For PC 286 - 386 - 486 and Pentium. the power supply used was the AT power supply. It give + and - 5V, + and - 12V. It is an old one, but maybe easier to use. The switch button is on the electric network part, so it is easy to stop.

Nowadays PC use the ATX technology that give + and - 5V, + and -12V and +3.3V, and the switch button is on the motherboard part so all the system can be turned on off by software.

All those voltage are maybe not really useful. It is possible to have component that work only with + and-12V.

Those power supply work well, but they are expensive and not really studied for low power. So it is maybe possible to create our own.

here is the schematic of an ATX power supply studied for 200W. Maybe we can try to inspire ourself from that.

attachment:schema_alim_ATX2.svg

those power supply use the Switched-mode power supply technique. it is interesting because there are a bigger efficiency (need to be verified at low power) than linear power supply, but they are using an oscillator that will create a lot of noise with a lot of different harmonic that can be very problematic for our use.

Here is a schematic for a very simple linear power supply.

drawing.svg

The transformer is to reduce the intensity and to isolate the circuit from the electrical network (especially the ground). then we use a diode bridge to convert AC current in DC current. and then a capacitor to have a stable intensity.

I don't know if this power supply is stable enough for what we expect, that mean supply component without burn them if there are a jump of current in the electric network.

Maybe that need to be improved before using it.

What still need to be done:

that would be interesting to have a complete modeling of the system with a non ideal transformer, to know what is the efficiency, and what is the component parameter that we need. Especially for the transformer and the capacitor.

We can add some other component to make some filter to increase the stability of the system.

Maybe it can help to add an other transformer and diode bridge in the linear system. And add signals with a Pi/2 dephase difference (just before the capacitor) to increase the total efficiency.

Scheduler

This part takes the signal from the plug and divides it to the different part. Maybe it is useful to use a transformer to isolate the electric system.

Signal reception

It is a filter which need to have a good bandwidth and a frequency controlled by the processor

Broadcast signal