User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 9:36 am

link works fine!

User avatar
PeterO
Posts: 5623
Joined: Sun Jul 22, 2012 4:14 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 10:57 am

AndrewS wrote: Even that (still) says "There is a PNP-matched-pair action going on around the polarity FET, but its 3AM and I'm not 100% sure what it's for so I'll wait till I get some rest before doing any analysis." ;)


Looking at this circuit.....

Normal operation:

LHS (left hand side) PNP transistor is diode connected so that means the +5V on the input will hold the base of the RHS PNP up at +4.4V and keep it off. RHS PNP collector is thus at ground (via 1kohm) and the FET turns on.

Pi Powered via GPIO (which is allowed):

RHS PNP emitter is at +5V. LHS PNP transistor is reverse biased as its emitter is at 0V (and assume collector+base are at 4.4V via RHS e-b junction) . Base current for RHS PNP comes via lefthand 1kohm so it rises to 4.4V and is turned on. RHS collector it at +5V so FETs Vgs is 0V so FET is off and 5V from PI does not appear on the input socket via the FET.

The simple "FET only" reverse voltage protection circuit explained in the adafruit video would not stop the PI from supplying current OUT of the input socket.

PeterO
Discoverer of the PI2 XENON DEATH FLASH!
Interests: C,Python,PIC,Electronics,Ham Radio (G0DZB),1960s British Computers.
"The primary requirement (as we've always seen in your examples) is that the code is readable. " Dougie Lawson

jdb
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2238
Joined: Thu Jul 11, 2013 2:37 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 11:54 am

PeterO wrote:
AndrewS wrote: Even that (still) says "There is a PNP-matched-pair action going on around the polarity FET, but its 3AM and I'm not 100% sure what it's for so I'll wait till I get some rest before doing any analysis." ;)


Looking at this circuit.....

Normal operation:

LHS (left hand side) PNP transistor is diode connected so that means the +5V on the input will hold the base of the RHS PNP up at +4.4V and keep it off. RHS PNP collector is thus at ground (via 1kohm) and the FET turns on.

Pi Powered via GPIO (which is allowed):

RHS PNP emitter is at +5V. LHS PNP transistor is reverse biased as its emitter is at 0V (and assume collector+base are at 4.4V via RHS e-b junction) . Base current for RHS PNP comes via lefthand 1kohm so it rises to 4.4V and is turned on. RHS collector it at +5V so FETs Vgs is 0V so FET is off and 5V from PI does not appear on the input socket via the FET.

The simple "FET only" reverse voltage protection circuit explained in the adafruit video would not stop the PI from supplying current OUT of the input socket.

PeterO
It does not, but the maximum current (reverse power) that can be fed out of the micro-USB port is vanishingly small.

The PNP array is Vbe/Hfe matched which means that any slight reverse voltage across the FET will cause an amplified difference to appear in the gate voltage.

Consider the case where Vbe(5,3) is a few tenths of a millivolt above Vbe(6,2) as the current through the FET reverses. The right-hand transistor will pass more of the share of current flowing through R2, resulting in more current flow Ic(5,4), increasing the gate voltage and restricting current flow from drain-source. The gains of the transistors involved make this an extremely sharp cut-off in the reverse direction.
Rockets are loud.
https://astro-pi.org

User avatar
PeterO
Posts: 5623
Joined: Sun Jul 22, 2012 4:14 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 12:09 pm

jdb wrote:
It does not,
Yes it does ! (But if you are going to disagree, it might be helpful if you actually said what you were disagreeing with) :roll:

PeterO
Discoverer of the PI2 XENON DEATH FLASH!
Interests: C,Python,PIC,Electronics,Ham Radio (G0DZB),1960s British Computers.
"The primary requirement (as we've always seen in your examples) is that the code is readable. " Dougie Lawson

jdb
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2238
Joined: Thu Jul 11, 2013 2:37 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 12:26 pm

PeterO wrote:
jdb wrote:
It does not,
Yes it does ! (But if you are going to disagree, it might be helpful if you actually said what you were disagreeing with) :roll:

PeterO
I was confirming that it doesn't prevent reverse current (for some small absolute value of reverse current) - hence agreeing.

Edit: oh I see - the Adafruit video has a completely different circuit.

I'm talking about the circuit on the Pi.
Rockets are loud.
https://astro-pi.org

User avatar
PeterO
Posts: 5623
Joined: Sun Jul 22, 2012 4:14 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 12:40 pm

jdb wrote:
PeterO wrote:
jdb wrote:
It does not,
Yes it does ! (But if you are going to disagree, it might be helpful if you actually said what you were disagreeing with) :roll:

PeterO
I was confirming that it doesn't prevent reverse current (for some small absolute value of reverse current) - hence agreeing.
Ok, :D Starting of with "No" made it seem like the opposite to me :D

PeterO
Discoverer of the PI2 XENON DEATH FLASH!
Interests: C,Python,PIC,Electronics,Ham Radio (G0DZB),1960s British Computers.
"The primary requirement (as we've always seen in your examples) is that the code is readable. " Dougie Lawson

johnclord
Posts: 3
Joined: Fri Jul 25, 2014 1:35 am

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 4:44 pm

The Raspberry Pi B+ A/V output has the ground on ring 3, if you count the tip as pin 1. Some charts are labeled with the tip as pin 4, which would make the ground on ring 2. Either way, the ground on the B+ is the second ring away from the plug grip. Any other cable with the ground on that ring will work, though you may have to swap around the Left, Right, and Video plugs at the other end until you find out which is which.
Cables that will NOT work, are Archos, Gigabeat, Creative Zen Vision series, Cown iAudio, and Apple iBook, Roku, as those have their ground nearest the plug grip, rather than on the next ring out.
Some cables that should work, would be iPod Video, Zune, Standard Camcorder cable, Netgear NeoTV Max, as they all have their ground on the second ring out from the grip, or 3rd from the tip. Different charts count from either end, so pay attention and if you re-describe any, make sure you don't use only the ring # since that leaves people wondering "from which end?"
The ground MUST be correct or the cable will be useless, as the ground MUST be connected to the shield conductor on each of the RCA plugs. After that, it doesn't matter - the colors may be wrong, but you can switch the cables around until you find the video, and then use the audio cables L & R as is, unless you want to find an audio test file (easily Googled for) to see determine exactly which is left or right and then label your plugs for future use.
- I went through all this when I lost a cable to my Netgear NeoTV and had to make up a new cable since Netgear doesn't sell them. I was able to use a standard camcorder cable with my NeoTV, and also made up a custom cable of my own. Of course, then I found the missing NeoTV cable two days later. :geek:

johnclord
Posts: 3
Joined: Fri Jul 25, 2014 1:35 am

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 4:53 pm

Yes, the NEW link works fine.

The original link I reported 404 on was:
http://www.raspberrypi.org/documentatio ... ematics.md
and displayed in the forum as:
http://www.raspberrypi.org/documentatio ... ematics.md

This new link that has been provided and displays in full, does work fine:
http://www.raspberrypi.org/documentatio ... /README.md

adding the /README made the difference between 404 and getting there.

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 5:49 pm

johnclord wrote: The ground MUST be correct or the cable will be useless, as the ground MUST be connected to the shield conductor on each of the RCA plugs.
Totally agree, but after that is just a question of the quality of the contacts inside the connector, does the contact that contacts with the sleeve works just as good as the (GND) contact that makes contact to the second ring (nearest to the sleeve), then it really doesn't matter which one you use for GND, and which one for Video. Yes, using the sleeve for GND initially makes sense it seems, but as there are many more cables in the field that use the second ring then I also would go for that configuration, as long as the contacts are reliable.

User avatar
jojopi
Posts: 3146
Joined: Tue Oct 11, 2011 8:38 pm

Re: Raspberry Pi B+ schematic

Fri Jul 25, 2014 6:56 pm

mahjongg wrote:Yes, using the sleeve for GND initially makes sense it seems, but as there are many more cables in the field that use the second ring then I also would go for that configuration, as long as the contacts are reliable.
I would question that, actually. Both of the TRRS to 3×RCA cables that I own have ground on the sleeve. Ground on ring2 is relatively recent, non-universal, and probably already obsolete in that all outputs are rapidly going HD, even on mobile devices.

By contrast, there does seem to be considerable momentum now behind putting ground on ring2 for mobile phone headsets. Compatibility with current and future headsets is probably more important than output to old SD displays, and it would have been less easy to adapt to using an external dongle.

See also: http://www.raspberrypi.org/forums/viewt ... 23#p577923

User avatar
Burngate
Posts: 6225
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 11:13 am

On a different note, the SD card used to be connected to GPIOs 47 - 53, with 47 being SD_CARD_DET.
I'm not sure how it was driven, as the ARM-Peripherals.pdf doesn't have any Alt functions for those GPIOs.

Now, we're told, GPIO 47 has been re-purposed as the ACT LED.
Has anyone worked out where the card detect is now connected? And have any of the other SD card signals moved, and if so where?

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 11:19 am

don't think a card detect signal makes any sense, a PI can't boot without a card, and after it has booted it cannot practically speaking be removed. Also, before booting there is no code running that can sense the detect signal.

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 11:32 am

jojopi wrote: By contrast, there does seem to be considerable momentum now behind putting ground on ring2 for mobile phone headsets. Compatibility with current and future headsets is probably more important than output to old SD displays, and it would have been less easy to adapt to using an external dongle.

See also: http://www.raspberrypi.org/forums/viewt ... 23#p577923
yes, well that could explain the last minute switch from GND on sleeve to GND on second ring.

User avatar
jojopi
Posts: 3146
Joined: Tue Oct 11, 2011 8:38 pm

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 12:38 pm

The nuked leaked B+ Rev 1.0 link from the first page does show the card detect switch left unconnected.

RISC OS at one time was fussy about the switch being closed. Linux has always ignored it and detected card removal using some other method.

You can hot swap cards safely in Linux, but you need a custom distro image with its root filesystem in RAM, or elsewhere.

dwelch67
Posts: 962
Joined: Sat May 26, 2012 5:32 pm

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 2:07 pm

jdb wrote: There is no access to the ARM JTAG so it's pretty much useless to anyone that doesn't have the Broadcom toolchain/software for VC debug.
The ARM JTAG was easily available to everyone (using free software) I dont know what you are talking about. And not only that a pin was moved so that it could be more readily available from the main connector on the previous boards...Please tell me that the B+ didnt change something to make this more difficult, mine has not arrived yet I would hate to have to send it straight to the bone pile...

Is there a time frame for when the schematic will be available?

David

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 2:14 pm

There is no header anymore for any of the two previous JTAG's. I think they are replaced by needle bed testing pads.

bbodin
Posts: 70
Joined: Sat Jun 28, 2014 3:23 pm

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 3:28 pm

dwelch67 wrote:
jdb wrote: There is no access to the ARM JTAG so it's pretty much useless to anyone that doesn't have the Broadcom toolchain/software for VC debug.
The ARM JTAG was easily available to everyone (using free software) I dont know what you are talking about. And not only that a pin was moved so that it could be more readily available from the main connector on the previous boards...Please tell me that the B+ didnt change something to make this more difficult, mine has not arrived yet I would hate to have to send it straight to the bone pile...
The ARM JTAG is accessible as an alternate function of the GPIO pins available on the P1 26-wide header. On the B+ board, GPIO26 is available on the J8 40-wide header. This makes a little bit easier to access the ARM JTAG signals.

References:
RPi BCM2835 GPIOs
Binh Bui

User avatar
Burngate
Posts: 6225
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 3:52 pm

mahjongg wrote:don't think a card detect signal makes any sense, a PI can't boot without a card, and after it has booted it cannot practically speaking be removed. Also, before booting there is no code running that can sense the detect signal.
jojopi wrote:The nuked leaked B+ Rev 1.0 link from the first page does show the card detect switch left unconnected.
RISC OS at one time was fussy about the switch being closed. Linux has always ignored it and detected card removal using some other method.
You can hot swap cards safely in Linux, but you need a custom distro image with its root filesystem in RAM, or elsewhere.
I can't quite see if there's a track connected to that pin.

RISC OS definitely knows if I take the card out, though it doesn't care unless I try to read from or write to it - it treats it like a floppy.
In fact I could copy from one card to another by drag-n-drop (or at least I could if I had more than one), it'll ask for the cards as it needs them.

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 5:24 pm

I think that Riscos can run completely from RAM, I doubt Linux does.
But yes, for Riscos a card insertion detector might be useful, and I think its simply polling the card to see if its there.
I believe that Linux will normally use such a method too if there is no hardware detector.

plugwash
Forum Moderator
Forum Moderator
Posts: 3546
Joined: Wed Dec 28, 2011 11:45 pm

Re: Raspberry Pi B+ schematic

Sat Jul 26, 2014 11:14 pm

mahjongg wrote:There is no header anymore for any of the two previous JTAG's. I think they are replaced by needle bed testing pads.
There are a row of test points on the bottom with signal names that make it pretty clear they are a jtag port, most likely (though not confirmed) they are the JTAG for the videocore. There is also a non-mounted connector on the bottom which connects to the same place as those test points. I presume the test points are used for factory programming and the non-mounted connector is used by insiders for development work.

I expect they decided that JTAG for the lan951x was pointless.

There is also a seperate arm jtag which is accessed through the perhipheral pin-mux system. On the B the only accessible lines on which this was available were used for the SD card making things awkward to say the least. On the B+ I belive it's available on some of the newly added GPIO pins.

User avatar
AndrewS
Posts: 3625
Joined: Sun Apr 22, 2012 4:50 pm
Location: Cambridge, UK
Contact: Website

Re: Raspberry Pi B+ schematic

Sun Jul 27, 2014 12:04 am

dwelch67 wrote:
jdb wrote: There is no access to the ARM JTAG so it's pretty much useless to anyone that doesn't have the Broadcom toolchain/software for VC debug.
The ARM JTAG was easily available to everyone (using free software) I dont know what you are talking about.
Looking at the original post from which you quoted, I suspect jdb meant that the non-soldered JTAG connector only provides access to the VideoCore JTAG, and doesn't provide access to the ARM JTAG :?: And hence it's the non-soldered JTAG connector that's useless to anyone that doesn't have the Broadcom toolchain/software...

AFAIK there shouldn't be any reason why you can't continue to access the ARM JTAG via pins 15, 16, 18, 22, 7, 13 on the J8 header. Some of the signals are also available on pins 29, 31, 32, 33 and 37.
http://elinux.org/RPi_BCM2835_GPIOs
plugwash wrote:There are a row of test points on the bottom with signal names that make it pretty clear they are a jtag port, most likely (though not confirmed) they are the JTAG for the videocore. There is also a non-mounted connector on the bottom which connects to the same place as those test points. I presume the test points are used for factory programming and the non-mounted connector is used by insiders for development work.
Umm, confirmed here ;)
plugwash wrote:On the B the only accessible lines on which this was available were used for the SD card making things awkward to say the least.
Huh? On the Rev2 Model B the ARM JTAG is on pins 15, 16, 18, 22, 7, 13 on the P1 header. It was only the Rev1 Model B where JTAG access was awkward (with ARM_TMS being on the CSI camera connector).

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 12663
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Raspberry Pi B+ schematic

Sun Jul 27, 2014 7:48 pm

Burngate wrote: I can't quite see if there's a track connected to that pin.
I can definitively say that neither switch is used in any way!

seulater
Posts: 76
Joined: Wed Jul 16, 2014 7:03 pm

Re: Raspberry Pi B+ schematic

Tue Aug 05, 2014 2:41 pm

Has RPi at least given a date when they will release the full schematic.
I want to make some modules for the CM kit. Would like to follow what has been proven already on the B+ kit.

JAvZ
Posts: 1
Joined: Tue Oct 21, 2014 8:01 pm

Re: Raspberry Pi B+ schematic

Tue Oct 21, 2014 8:15 pm

Quick question (since BOM/Schematics are not available yet): what micro-SD connector did you use? (I'd like to procure the same one for my own little project). Thanks for your time.

User avatar
Jessie
Posts: 1754
Joined: Fri Nov 04, 2011 7:40 pm
Location: C/S CO USA

Re: Raspberry Pi B+ schematic

Tue Oct 21, 2014 9:49 pm

seulater wrote:Has RPi at least given a date when they will release the full schematic.
I want to make some modules for the CM kit. Would like to follow what has been proven already on the B+ kit.
Yes the schematics are showing up two weeks after never or after no one is selling a R Pi anylonger. I have never been told anything, but I suspect their manufacturing partners have clauses to prevent a full schematic. So call up Farnell and RS and ask them to loosen their grip. BC might even cover the pinout under the NDA.

Return to “General discussion”