Board versions/revisions, or different assembly lines?


68 posts   Page 1 of 3   1, 2, 3
by marcoalexcampos » Wed Jun 27, 2012 10:19 am
Hi,

I've recently received two RaspberryPi model B boards (from Farnell and RS), and I see quite some differences between them. Picture here: http://195.22.23.82/RaspberryPi.jpg

What tricked me to look close to the boards was the fact that one of the boards has significant stability problems with the USB (LAN/Keyboard/Mouse). Let's call it board "B" for reference.

Note that the tests where ran with the same power adapter (with 5V and 1,25A), bought from RS and it looks very good/stable.

These are some differences I've noted so far:
a) The TP1 and TP2 pins show 4.99 on load in board "A" and 4.89 on board "B" (remember same power adaptor)
b) The RAM is from SAMSUNG in "A" and HYNIX in "B"
c) There are some minor differences in prints on the PCB (look at the picture attached)

I'd like to have some feedback on how I can make the "B" run stable (please don't say power adapter), since I didn't wanted to return it.

Also if you have different boards please fell free to post a picture, to see if we start to make some sense out of these problems :)

Thank you all
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by Troublesome » Wed Jun 27, 2012 12:39 pm
I've been lucky enough to get one from each supplier. Just had a check and the Pi from RS has Hynix RAM and the Pi from Farnell has Samsung. I assume that each company is getting its supply from different lines / factories. However I haven't experienced the instability problems that you have on the RS Pi.
Posts: 6
Joined: Mon Jun 18, 2012 10:07 am
by marcoalexcampos » Wed Jun 27, 2012 2:56 pm
Hi,

Thank you for the reply.

Can you run the same test I did for the voltage under load and post the results?
Also if you board has any difference (even if small) could you post a picture? Or just send by email and I'll make it available.

Later tonight I'll post a photo of the back so you and others can compare with your board(s).

Cheers,
Marco.
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by hippy » Wed Jun 27, 2012 4:26 pm
marcoalexcampos wrote:What tricked me to look close to the boards was the fact that one of the boards has significant stability problems with the USB (LAN/Keyboard/Mouse).

The two USB fuses are marked differently so perhaps from different suppliers. Might be worth taking a multimeter to them with the board un-poweredand everything unplugged and measure what their respective resistances are.
Posts: 774
Joined: Fri Sep 09, 2011 10:34 pm
by marcoalexcampos » Wed Jun 27, 2012 5:54 pm
Will do that. What is the description of the fuses? Something like "F1"? Where is it located exactly?
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by therealeasterbunny » Wed Jun 27, 2012 9:32 pm
Here are more comparisons. My board on the left, from RS received in May, and the board on the right is from my friend which is ialo from RS received mid June.

http://www.therealeasterbunny.pwp.blueyonder.co.uk/pi2.jpg
Posts: 77
Joined: Thu Feb 02, 2012 5:53 pm
by marcoalexcampos » Thu Jun 28, 2012 1:29 am
I've measured the green components on the right of the USB ports with the F1 and F2 references, and these are the values:

Board A: F1: 3.8ohm, F2: 4.5ohm
Board B: F1: 8.2ohm, F2: 5.9ohm

I just got home so maybe this is not the best time to be looking at this. I'll take both boards to work tomorrow and I will be able to make any further tests you might think apropriate.

I will also try to take a more detailed picture of that area (USB/Fuses), and the back of the boards has well.

You all can chat with me about this issue at MSN:mcampos@marcocampos.com or Skype:mcampos2, for more interactive testing.

Cheers,
Marco.
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by marcoalexcampos » Thu Jun 28, 2012 1:38 am
@therealeasterbunny: Do you think you can get a more clear photo, or at least identify the chips (mainly the RAM) that differ from my boards?

At least one of the PCBs is different from the two I have, so we're looking at 3 or 4 (slightly) different boards. :)

Let's call your board on the left "C", just for reference.

@Troublesome: Do you think it's possible to get photos from your boards?

Cheers.
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by therealeasterbunny » Thu Jun 28, 2012 6:07 am
marcoalexcampos wrote:@therealeasterbunny: Do you think you can get a more clear photo, or at least identify the chips (mainly the RAM) that differ from my boards?


JPG Updated with RAM info.

marcoalexcampos wrote:At least one of the PCBs is different from the two I have, so we're looking at 3 or 4 (slightly) different boards. :)

Let's call your board on the left "C", just for reference.
Cheers.


Okey dokey, JPG updated with 'Board C :)

I will try and put up a reverse board photo tonight.
Posts: 77
Joined: Thu Feb 02, 2012 5:53 pm
by therealeasterbunny » Thu Jun 28, 2012 6:41 am
Posts: 77
Joined: Thu Feb 02, 2012 5:53 pm
by marcoalexcampos » Thu Jun 28, 2012 9:38 am
Good morning,

Board A back: http://195.22.23.82/RaspberryPi-Board_A-Back.jpg
Board B back: http://195.22.23.82/RaspberryPi-Board_B-Back.jpg

I will start to compile these photos and info on a Wiki page, to that we can have a clear access.
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by therealeasterbunny » Thu Jun 28, 2012 11:39 am
Cool, sounds like a plan. Let us know the link to the Wiki by return.

I wonder if there are more than three board types out there currently? ...
Posts: 77
Joined: Thu Feb 02, 2012 5:53 pm
by marcoalexcampos » Thu Jun 28, 2012 11:40 am
I've created a section on elinux's wiki with these infos and photos:

http://elinux.org/R-Pi_Troubleshooting#Hardware_versions.2Frevisions

Please feel free to make suggestions or add info. :D
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by RaTTuS » Thu Jun 28, 2012 11:51 am
http://imgur.com/a/3wMKB
my 2Pi
both fairly much stable
1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX - Prosliver FTW
"That's not right, the badgers have moved the goalposts."
User avatar
Posts: 5018
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK
by Troublesome » Thu Jun 28, 2012 1:25 pm
Hey Marco, Sorry for the slow reply! It seems I failed to click watch on this thread.

Ive taken photos of the front and back of my two and put links below. (Warning: The images are about 3mb each)

Source: RS
RAM: Hynix
Front: http://troublesome.dyndns.org/rs.jpg
Back: http://troublesome.dyndns.org/rs_back.jpg


Source: Farnell
RAM: Samsung
Front: http://troublesome.dyndns.org/farnell.jpg
Back: http://troublesome.dyndns.org/farnell_back.jpg

If you need better images let me know and ill take some more with a DSLR when I get home.
Posts: 6
Joined: Mon Jun 18, 2012 10:07 am
by Troublesome » Thu Jun 28, 2012 1:27 pm
Looking at the wiki page you setup and it seems my Hynix board is the same as yours, however my Samsung board is different in that it has the same C6 as the Hynix.
Posts: 6
Joined: Mon Jun 18, 2012 10:07 am
by marcoalexcampos » Thu Jun 28, 2012 2:06 pm
@RaTTuS: Thank you for your pictures. Can you send the references for some of the main components like RAM, USB, C6, F1/F2, etc? And others that you see that differ from any of the boards already posted. If you could measure the F1/F2 we could compare.

Thank you.
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by RaTTuS » Thu Jun 28, 2012 2:29 pm
I cannot measure tp1/tp2 [no voltmeter]
and I'm not sure about my camera for better images [may have to get a magnifying glass to see what they say] - however lets see what I can do [now or in 18hrs]
1QC43qbL5FySu2Pi51vGqKqxy3UiJgukSX - Prosliver FTW
"That's not right, the badgers have moved the goalposts."
User avatar
Posts: 5018
Joined: Tue Nov 29, 2011 11:12 am
Location: North West UK
by marcoalexcampos » Thu Jun 28, 2012 3:48 pm
Thank you. No rush, when you have the time. ;)
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by marcoalexcampos » Thu Jun 28, 2012 3:59 pm
@hippy: What do you make out of the values I've measured? Thank you.

Board A: F1: 3.8ohm, F2: 4.5ohm (<- Ok)
Board B: F1: 8.2ohm, F2: 5.9ohm (<- with random problems on USB/LAN)
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by hippy » Thu Jun 28, 2012 4:13 pm
marcoalexcampos wrote:@hippy: What do you make out of the values I've measured? Thank you.

Board A: F1: 3.8ohm, F2: 4.5ohm (<- Ok)
Board B: F1: 8.2ohm, F2: 5.9ohm (<- with random problems on USB/LAN)

I'm not the expert but other threads have noted voltage drops when the fuse resistances are higher, drops enough to cause some ( but not all ) USB devices to become problematic. That the one with the higher resistances has problems suggests this could be more of the same.

AIUI, the fuses only affect power supplied to the USB port devices so shouldn't be affecting the LAN part of the USB/LAN chip but strange interactions can undoubtedly occur.
Posts: 774
Joined: Fri Sep 09, 2011 10:34 pm
by SN » Thu Jun 28, 2012 4:29 pm
marcoalexcampos wrote:I've created a section on elinux's wiki with these infos and photos:

http://elinux.org/R-Pi_Troubleshooting#Hardware_versions.2Frevisions

Please feel free to make suggestions or add info. :D


You need to add build week ranges on these and change your A,B,C nomenclature to match that.

There were plenty of people earlier than me but my Pi collection consists of:-
Two Build Week 1215 Farnell Boards identical to your Type A but with C6 labelling of "220 16V EEZ" and "n2 220c UD" respectively and having Build Week on front of board
Two Build Week 1218 RS Boards identical to your Type B/C but both with C6 labelling of "2B 220 16H"
So I would suggest its the RAM/SOC is the main variation between suppliers - anything else is parts bin stuff
I can't measure my F1/F2 resistances too well as my digimeter only goes down to a 2Kohm range - I'm seeing c.6-7ohms on the RS boards - the Farnell boards are running so can't measure
Whilst we're looking for differences/patterns, my Serial Numbers show no particular correlation either the Farnell ones being 6F71xxxx & ACA3xxxx and the RS ones 9630xxxx & 6AFBxxxx
Steve N – binatone mk4->intellivision->zx81->spectrum->cbm64->cpc6128->520stfm->pc->raspi ?
User avatar
Posts: 1008
Joined: Mon Feb 13, 2012 8:06 pm
Location: Romiley, UK
by marcoalexcampos » Thu Jun 28, 2012 5:16 pm
@SN: Thank you for the hint! Didn't knew that those numbers wore year/week of manufacturing. Just updated the table.

@therealeasterbunny: Your board is now called B1213 :)
User avatar
Posts: 37
Joined: Fri Jun 22, 2012 2:45 pm
by therealeasterbunny » Thu Jun 28, 2012 6:08 pm
Okey doke :) Thanks. Will take a peek.

(Time passes)

Hey thats great!

I am happy to submit any F1 / F2 info but beyond my usual volt testing AC/DC I have done before in my life, what setting to I put my digital multimeter on for ohms and what do I put my red/blue pokey sticks on in terms of Pi bits to give you the readings to update the Wiki for B1213? Happy to help if I know what to set / poke. Can anyone point me in the right setting?

My multimeter is a Beckman Industrial dM25XL exactly like:

http://www.covingtoninnovations.com/michael/blog/0706/070610-dmm.jpg

... and do I measure ohms / resistance with Pi powered on or off? Sorry... I've just been a volts chap up until now... :)
Posts: 77
Joined: Thu Feb 02, 2012 5:53 pm
by antijn » Thu Jun 28, 2012 9:12 pm
Hi,

I got two cards showing the same problems. One from RS, build week 1218 which has stability problems and one from Farnell, also build week 1218, but works in the same setup.

Not sure what the resistances for the fuses are needed for, but I can give another data point at least.
Values for F1, F2:
RS:
6.4 Ohm , 5.0 Ohm
Farnell:
5.8 Ohm, 3.9 Ohm

I can run other comparative tests too if needed.
Posts: 2
Joined: Thu Jun 28, 2012 8:43 pm