New Pi 4B v1.2?

[trejan]

Somebody on Reddit said they've got a Pi 4B 4GB v1.2 with revision code c03112?

[Cob]

Brand-new Reddit account and no photos after 5 hours?


Hmmmmmmmmmm...

[trejan]

Brand-new Reddit account and no photos after 5 hours?


Hmmmmmmmmmm...

Go to their userpage to see the photo they posted of another board but no confirmation this is a v1.2. It must have gotten caught in the spam trap.

https://i.imgur.com/wHA6AZJ.jpg for anybody who doesn't want to go via Reddit. https://www.cnx-software.com/2019/06/24 ... o-4gb-ram/ has a good photo of the v1.1 board if anybody wants to compare. The PCB looks identical to me so if this is really is a v1.2 then it doesn't appear to have the USB-C CC resistor fix.

Will need to see if they ever post a photo of their own board.

[Cob]

They’re both the same board.

Obvious spam.

[trejan]

They’re both the same board.

Obvious spam.

Yes. It is probably a troll post but there is a chance they're telling the truth. A photo of the revised Pi 3B with the heatspreader topped SoC from the 3B+ appeared early as well.

Will need to see what if anything they post. Easily debunked by Pi engineers anyway.

[trejan]

This is looking legitimate. As mentioned in the Reddit comments, there is an issue ticket at https://github.com/rpi-ws281x/rpi-ws281 ... /issues/30 which talks about a Pi 4B v1.2 as well. https://bugs.launchpad.net/ubuntu/+sour ... omments/20 is another one.

A September pastebin of a boot log + crash from a LibreELEC dev using a 4B v1.2 at https://pastebin.com/jR6q2Au9

Any Pi engineers want to let us in on what this new 1.2 revision has changed?

[W. H. Heydt]

Any Pi engineers want to let us in on what this new 1.2 revision has changed?

Another thing the revision might be would be newer firmware.

[Imperf3kt]

Any Pi engineers want to let us in on what this new 1.2 revision has changed?

Another thing the revision might be would be newer firmware.

Unlikely... You can update the firmware on a Pi4b without needing a hardware change.

[icecold]

Any news on whats been updated in this revision since it obviously exists now?

[Gavinmc42]

Working USB booting

[davidcoton]

Any news on whats been updated in this revision since it obviously exists now?

Only "fake" news so far, I fear.

[HawaiianPi]

Working USB booting

It's getting closer every day...

Bootloader guy is now working fully on USB boot. No timescales as yet.

The USB-C fix might not even warrant a new revision number, because it has no effect on software, firmware or certifications. Small discrete components on PCBs often change over time. Ever wonder why there are empty component spaces on most PCBs? This revision 1.2 thing (if true) is probably a red herring that has nothing to do with the USB-C fix.

[hippy]

Another thing the revision might be would be newer firmware.

Unlikely... You can update the firmware on a Pi4b without needing a hardware change.

The term "firmware" is used quite loosely for Pi. There is the ROM bootloader firmware in the SoC which cannot be changed without a silicon respin, the firmware in the Boot Eprom, firmware for the GPU which loads from SD Card at boot, firmware for the WiFi/BT module, for the USB host chip, possibly the LAN chip, TV Hat, PoE HAT and other smart peripheral components.

Revision code haven't historically changed when the functionality has remained the same, the change is transparent to the software, so my guess would be a step change in SoC, a change in the ROM bootloader code, or some change the software which is loaded needs to know about so it can configure itself correctly.

[ragin_koala]

I do have a 1.2 c03112 that I got at the 18/10/2019 maker faire, thought it was just like any other one

[W. H. Heydt]

Jamesh is the new board revision out in the wild?

More to the point.... *is* there a new board revision? (And I note that picture posted doesn't show a version number at all, which is also true of the original/existing Pi4B boards.)

[trejan]

More to the point.... *is* there a new board revision?

I'm guessing yes. The two issues were posted by established users that were registered years ago and the LibreELEC kernel log is from September. If it was a troll post then I'd expect one of the Pi engineers to have said it was a fake by now.

(And I note that picture posted doesn't show a version number at all, which is also true of the original/existing Pi4B boards.)

Yeah. The original Pi 4 was v1.1 at release but it wasn't marked on the silkscreen. Something has changed to bump it to v1.2 but it wasn't the PCB artwork. If the photo is accurate then it doesn't have the USB-C CC resistor fix yet.

Somebody else has posted to the Reddit thread saying they've received a v1.2 back in October.

Odd that nobody from Pi Towers is confirming or denying this.

[emma1997]

Not surprising to me at all. I think RPF/RPT may want to 'flush the pipes' to avoid Osbourne Effect. I know I wouldn't mind treading water for a month or two, maybe even longer, in order to get latest and greatest.

[GaryH1]

Don't forget that there are some components on the underside of the PCB, so the additional resistor fix needed for the USB-C issue may be there. We haven't seen any pictures yet of the underside of the supposed rev 1.2 version. As it's an engineering change I would expect the revision number to be bumped.

[trejan]

Don't forget that there are some components on the underside of the PCB, so the additional resistor fix needed for the USB-C issue may be there. We haven't seen any pictures yet of the underside of the supposed rev 1.2 version. As it's an engineering change I would expect the revision number to be bumped.

I'd expect the silkscreen to be updated like previous Pi boards if the PCB was altered. There isn't any clean way of doing the USB-C CC resistor fix without altering the board as the connection between CC1 and CC2 is under the socket itself. Besides, the complaints would get even louder if you couldn't identify a fixed board without staring at SMD components.

[laurent]

This revision may been planned before the USB-C "issue" discovery, and address some/many other minor points that doesn't need any staff explanations. If this revision really exists, and doesn't correct the USB-C "issue", a further one will do.

[HawaiianPi]

I'd expect the silkscreen to be updated like previous Pi boards if the PCB was altered.

Except that, unlike previous Pi boards, the Pi 4B has no revision number on the silkscreen that would need updating.

They do have a year, however, which makes the supposed photo of the rumored 4B r1.2 board highly suspect, as it also has 2018 on the silkscreen (since the 4B launched early in 2019, I would expect any later board revisions to have a 2019 date).

There isn't any clean way of doing the USB-C CC resistor fix without altering the board as the connection between CC1 and CC2 is under the socket itself.

Got any verifiable links to that information (not criticizing, just curious)?

[trejan]

Except that, unlike previous Pi boards, the Pi 4B has no revision number on the silkscreen that would need updating.

It is odd it doesn't say v1.1 on the silkscreen but I expect them to add it or it will be a nightmare trying to determine if you've got a fixed board or not.

They do have a year, however, which makes the supposed photo of the rumored 4B r1.2 board highly suspect, as it also has 2018 on the silkscreen (since the 4B launched early in 2019, I would expect any later board revisions to have a 2019 date).

The new 3B with heatspreader topped BCM2837B0 still says v1.2 with a 2015 date as the PCB wasn't changed. If the v1.2 board didn't have any PCB artwork changes then no silkscreen changes isn't that strange.

Got any verifiable links to that information (not criticizing, just curious)?

I looked at the board under an inspection microscope. There is no visible trace connecting the two pins together on the top. R1 and R79 are on the bottom with a via in the middle but there is no sign of the CC pin connection. There are some x-rays of the Pi 4 and one of them is a close up of the USB-C connector area. You can just about see a trace going between CC1 and CC2 underneath the socket.

Lifting one of the CC pins and soldering a resistor is fine for a one-off but you're not going to want to do this on a production line. Putting a thin interposer PCB under the socket isn't viable either for production reasons and it'll likely throw off alignment of the USB-C socket for cases. If you're going to fix this issue then it'll be new PCB artwork hence why if v1.2 is real then I don't think it has the USB-C fix.

[W. H. Heydt]

I do have a 1.2 c03112 that I got at the 18/10/2019 maker faire, thought it was just like any other one

Is there any marking on the board to indicated that it's a v1.2? Can you post a picture?

[HawaiianPi]

It is odd it doesn't say v1.1 on the silkscreen...

Just noticed the 3B+ doesn't have a revision # either.

3BPlus_silkscreen.jpg (65.55 KiB) Viewed 27237 times

On the other hand, the Zero W has PCB silkscreens with 2 different dates, but the same revision.

pi_zero_w_v1.1_2016-2017.jpg (195.33 KiB) Viewed 27237 times

¯\_(ツ)_/¯

[rpdom]

I think the date change and the silkscreen changes are because of additional approvals. The 2017 has D170026007 and 007-AF0055 markings, which are the UL approval marks for Japan and would be required for selling into that market.