Raspberry Pi Single Board Computer

Documentation!

cat /proc/cpuinfo returns BCM2708 Revision 0010 (along with other things like processor type).

Can I get from here to any documentation on the rasp pi that I have? and if so how?

Thanks,

very much a newbee. Ian D

what happens when one of the tiny cheap and designed to be not flush with the pcb connectors breaks

how easy are they to repair and replace?

why couldnt the first version have been sold with the ability to add ram to it?

when can we see the internet bandwidth we are paying for onscreen along with the pathetic 1-10% of what we are paying for that we actually get

what USB peripherals are compatible with this device is there a list somewhere

what are the specs of the GPU

thanks

http://www.youtube.com/watch?v=zsjTpFR0oYQ

PS the circuit board, for a wide area around both the USB and RJ45 connectors were painted with flux.

This is a bad...Ive personally seen excess flux crash a 3,85 MHz clock in a VCR...and the signals on the Pi are a much higher frequency, on the order of 12 MHz on up, possibly to high UHF.

http://alphacpmd.com/~/media/Files/CooksonElectronics/TP%20-RF%20Characterization%20of%20No%20Clean%20Solder%20Flux%20ResiduesWWE%20SM000%20CNP%2020100917.PDF

There are hardware failures in the Pi due to faulty design. The below is written by a hardware engineer with massive RFI and circuit test./design experience, plus FCC cert and manufacturing.

Those of you noticing USB problems, its not you, the Pis designers ignored the design requirements of SMSC for the LAN9512 IO chip, heres the long list, since there may be EEs floating through here.

Management at Newark need to take notice, your Company and others are selling faulty product that is mis designed and is NOT USB COMPLIANT.

Main problems:

* false rumours of the need to use a powered USB hub - not relevant to failure of a wireless USB receiver drawing 28 mA

* false rumours of power supply problems in the Pi (there arent, the supplies are clean and stable on an oscilloscope)

* Numerous and serious design defects in both the circuits and CBA layout of the Pi - the Pis designers refused to adhere to SMSCs design requirements,

  especially inexcusable since SMSCs website offers design evaluation.

Design defects in the Pi B 512: (this is an engineering level R-E and document, non engineers may not understand it)

Refer to SMSCs documents on the LAN9512 and App notes (engineers will know how to find those)

For non engineers or programmers:

The USB connector on the Pi is too far from the power connector, many of the electronic components

associated with the USB chip are incorrectly installed, circuit board layout is wrong (these issues are critical on high speed circuit boards)

and the USB clock circuit is too close to the USB data lines. These design faults can lead to both hardware and software crash or instability.

=============

List of design failures in the Raspberry Pi, 512 Mb B version:

Source documents:

Raspberry-Pi-Schematics-R1.0.pdf  (schematic of the Pi, incorrect in some areas)

an1920.pdf (from SMSC)            (Ethernet/USB chip)

9512_sch.pdf (SMSC)               ( " ")

NCP1117-D.pdf                     (from ON Semiconductor) power supply regulator 

  Comparing the Pi schematic, the Pi board itself and the documentation by the component manufacturers

we see many design mistakes in the Pi that are causing it to violate the USB design standards

and not work with common wireless Keyboard and mouse sets.

  It is absolutely pointless to make and sell a piece of computer equipment that wont even work

with a mouse and keyboard.

  There are so many serious electrical design mistakes on the Pi B board that it is pointless

to even consider using it to develop code.

  The popular rumour on the Pi Internet Forums is that the Pi requires use of

a powered hub as the Pi cannot supply enough current to USB devices. This is an admission

that the Pis design violates the USB standard, and is also false in that the Pi fails with 

USB devices that are very low power that do not require a powered hub.

Design analysis:

1.) There is nothing wrong with the Pis power supply (RG2, the bulk regulator that drives

the USB Hub, 3.3V regulated supply- (NCP1117)). Oscilloscope and DVM (lab quality equipment,

not hobbyist equipment) prove the power supplies are flat and clean. Bulk supply is 4.95V from

a switching regulator supply (MCM 28-13060), about 100mV noise floor, the 3.3V supply is about 10 mV.

There is no power supply droop when a 28 mA USB receiver is hot plugged. This indicates RFI causing signal corruption

in the controller, Ive personally seen that and fixed it in a HVAC controller design.

    a.) The power supply connector is too small and fragile.

    b.) the power supply as on the Pi is correctly designed according to ON Semiconductors data sheet for the

    NCP1117.

  That the Pi cannot supply 600 mA to a 3G modem is a separate topic and is not excusable as this violates the USB standard.

2.) The Pis USB ports do not work with an MCM 83-13110 wireless KBD and mouse or a Microsoft set.

Both sets work fine with two laptops. The MCM wireless receiver plugs directly into the USB port, and has a receiver 

and a 12MHz clock, which happens to be the same frequency (?) as a low speed USB device bus.

  The USB receivers clock at 12 MHz, and the USB signal frequency being 12, with the SMSC device clock at 25 MHz,

may be allowing interference at 2f especially since the Pis circuit board is improperly designed as in 

Sect. 3 (B)(5) of this Document.

  When the MCM receiver is used on the Pi, the mouse pointer jumps to the bottom of the screen and stays there.

  When used with a Microsoft wireless set, the keyboard exhibits a 'stuck key' symptom when the User attempts to input Username

and PW. This has been reported on a Pi Forum.

  The MCM wireless KBD/mouse receiver draws 28 mA DC quiescent current. This is nearly zero and can not be 

the source of board crash when the receiver is hot plugged. 

    a.) This 28 mA figure is well under the limits in SMSCs document AN1920.pdf, Sect. 4.5.1. and far less

    than the test amount in Sect. 4.5.2.2 of 100 mA.

  The problem tended to be mitigated (but not totally eliminated) by using a USB extension cable. The cable changes

the tuning of the USB lines by adding capacitance and inductance, but more importantly, moves the receiver away

from the USB socket on the board.

  The Pi works with an older Logitech wireless set with a 3 foot USB cable in between WITHOUT a powered hub.

  The story is that the Pi must be used with a USB hub, the fact that simply using a USB extension tended to mitigate

the problem indicates the USB hub story is exactly that, a cover story that hides the failure in the Pi

3.) specific design problems in the Pi:

    Being a hardware engineer particularly experienced in RFI and FCC type acceptance, and having the manufacturers

data sheets and Pi schematic and two Pi boards, I note the following design flaws in the Pi:

(NOTE: the Pi boards I received do not have the Polyfuse port switching circuit, the USB power is direct to the 5V supply

and the supply is correct and clean)

  a.) the USB connector is located in the wrong place, it, if used with a 3/4G modem for example, will draw the highest current

  and/or highest total energy of any device on the board, so it must be located next to the power connector.

    Other such boards are designed correctly:

    https://www.olimex.com/Products/OLinuXino/iMX233/iMX233-OLinuXino-MAXI/ (power and USB and other connectors on the same side of the board,

    power does not traverse the signal paths. Not the best layout because the USB connector is across the board from the power, but its better)

  Notice the folk at Olimex know of the failures in the Pi - https://www.olimex.com/Products/OLinuXino/

  People with hardware engineering experience can catch such failures, persons who only write code cannot.

  Another example of 5V supply near the USB connector- http://www.embeddedarm.com/products/board-detail.php?product=TS-7800

  An example of doing it badly wrong as in the Pi - Beagle Bone - http://beagleboard.org/bone 

Read the Wiki on their previous release that was absolutely plagued with design failures- http://elinux.org/BeagleBoard (Revisions section)

  (get a copy of their Reference manual and read the disclaimers http://beagleboard.org/static/beaglebone/latest/Docs/Hardware/BONE_SRM.pdf

  They admit to serious RF immunity problems.

  Why is the connector location wrong?

  Code hacks and hardware designers know the USB power source is DC. They do not realize that the USB supply CURRENT is AC and VERY noisy.

  It is not acceptable to route the USB current past or through sensitive, high speed circuits as the noise from the USB current can corrupt signals.

  b.) See the SMSC document AN1920 (an1920.pdf)

    This documentation sets out REQUIREMENTS (not suggestions) of how the circuits must be designed. The Pis designers

  ignored them and now have a faulty product.

    1.) ibid, Sec. 4.3 requires the USB connector shield to ONLY be AC coupled, the shield on the Pi is soldered to DC ground.

    This allows USB device current to mix with USB signal currents and can cause data corruption.

    2.) Sect. 2.9.1 forbids using 90 degree turns on circuit traces, one USB trace on the Pi board has a 90 degree turn.

    Sharp corners radiate noise energy from very high frequency signals.   

    3.) Sect 2.8.4 requires bias resistors be on the component side, R23,24 and others are not on the component side.

    4.) Sect. 2.5.1 requires crystal circuit components be on the top layer, they are on the bottm of the board.

    A corrupted clock signal could crash the entire IO system and maybe even the processor.

    5.) Sect. 4.2, requires DP/DM trace isolation, these traces are routed near R39 and C41, IN THE CRYSTAL OSCILLATOR CIRCUIT.

    This fault, by itself, can crash the whole board by the USB port signals corrupting the clock signal in the hub.

    6.) Sect 4.5.3, requires at least 150 uF capacitance for the USB connector, the Pi has a 47 uF, only half the required value.

    7.) Sect 2.2, requires one decoupling capacitor for each power pin, there is one missing on page 3 of the Pi schematic

    with the group of 7 power pins including pin 5- 7 pins, 6 capacitors is not "one for each pin."

    8.) Sect. 2.2.2 requires CALCULATION of individual decoupling capacitor values, this was not done on the Pi,

    the designers apparently assuming that all such caps are 0.001 uF; this is a common mistake made by designers and experimenters.

    9.) Sect 2.1 & 4.5.1 require sufficient power supply. If the Pi cannot supply power to operate USB devices, then its design

    violates both SMSCs requirements and the USB standard and is misleading to advertise to sell as USB -operable when it is not.

4.) See Pi schematic P.3 and Document 9512.pdf. GPIO pins were left floating, this violates the first rule of digital design

that "every pin must go somewhere". If these lines are left tri state or OD in software, they could couple noise into the package.

5.) See Pi schematic P.3 and 9512.pdf Sect. 2.3. PU/PD pins must have external pull up.down resistors, the Pi schematic

does not show an external resistor on pin 14. Apparently this entire circuit has been omitted as it didnt function properly.

6.) The USB standard requires hot plug-able operation, hot plugging the KBD/mouse wireless receiver from MCM

into the Pi crashes the entire board back to a hard reset. This is NOT due to power supply bounce, I have tested

that with an oscillloscope. It is due to signal corruption.

7.) False and misleading information on the Pi Forums on the web:

    a.) USB problems are blamed on lack of a powered hub. They are not, as proven by oscilliscope measurements

    and the fact the wireless KBD/mouse receiver draws 28 mA. Tests prove that even a USB extension cable improves

    operation without being powered,

    b.) claims that Poly fuses were use don the USB ports for over current protection are false, the Polyfuses

    according to Document 9512.pdf, Sect. 2.2.2 are for port power switching, not overcurrent protection.

  This Pi device is not suited for the intended purpose, especially as it is advertised /sold as USB compliant and

is not, and there is nothing that can be done to remedy it without a complete redesign of the circuits and circuit

board to comply with the USB specification and requirements of SMSC.

  As a result of the Pis designers negligence, thousands of these units have been sold to Consumers who are being

deceived into thinking that the Pi works correctly and they must use a USB hub. This is a false premise.

You can log in with:

Username: pi

Password: raspberry

Then you can use "sudo" to execute commands as root.  "sudo su -" will give you a root shell.  "passwd" in the root shell should let you change the root password if you really need to.

I just need to find the root password from the latest wheezy image file.  Please???

try type Raspberry pi 3D dimension under google.

The RPI-B is bit too slow, wish the RPI-C CPU can up to 1.2GHz and SDRAM up to 1GHz, otherwise, better to overclock it, anyway, it's good for command prompt type, don't expect too much for graph mode when tablet all more then 1GHz CPU. Also, don't expect chinese language on it.

I'd like to see a few goodies posted:

1. Schematics. Most of the stuff I'd like to do involve hardware (adding FPGA peripherals, drivers, etc.). I can back-engineer all this stuff if I have to, but having an official set of documents would be great. I'd also like to add some memory, but that doesn't look like it's feasible.

2. Docs on drivers for peripherals such as USB disks

Raspberry PI is a great platform for teaching, but it could make a good platform for more serious development because it has a relatively stable OS.