Heater
Posts: 14431
Joined: Tue Jul 17, 2012 3:02 pm

Re: New Pi 4B v1.2?

Wed Dec 11, 2019 12:08 pm

chwe,

What are you ranting for? Trifling details. Nobody cares. Pi works fine. I'm going to buy the ones you don't.
Memory in C++ is a leaky abstraction .

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 24934
Joined: Sat Jul 30, 2011 7:41 pm

Re: New Pi 4B v1.2?

Wed Dec 11, 2019 12:26 pm

chwe wrote:
Wed Dec 11, 2019 11:57 am
jamesh wrote:
Tue Dec 10, 2019 4:44 pm
And what video/USB3/Wifi problems? Almost certainly software, whatever they might be. Links to github issues would be the best way on indicating where your problems are, I can then see how we are dealing with them.
Maybe this one?
https://www.cnx-software.com/2019/11/29 ... esolution/
https://www.enricozini.org/blog/2019/hi ... esolution/
https://twitter.com/assortedhackery/sta ... 3898029061
Please don't expect us to crawl around the internet looking on everyone else's website for bugs. We have a number of github issue trackers, if it's not in any of those, it's not been officially reported. Stuff that appears on our forums may or may not been seen (except some specific bug reporting threads), so github is the only guaranteed way of reporting an issue so it won't get lost.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I own the world’s worst thesaurus. Not only is it awful, it’s awful."

chwe
Posts: 136
Joined: Tue Jul 31, 2018 1:35 pm

Re: New Pi 4B v1.2?

Wed Dec 11, 2019 6:30 pm

at least your boss spotted it (from the second link):

https://www.androidauthority.com/raspbe ... n-1059644/

commenting that:
“Hard to know what the problem is from the blog post (sic), but this is exactly what I’d expect if someone was using poorly shielded HDMI cables,” Upton told Android Authority in an emailed response. “A poorly shielded cable radiates ~20dB higher than a properly shielded one. I’d be very surprised if this is a hardware issue with the Pi itself.”
but the original (https://www.enricozini.org/blog/2019/hi ... esolution/) stated:
We reproduced this:
- on both microHDMI outputs
- with two different cables: one with a microHDMI to HDMI dongle adapter; one direct microHDMI to HDMI. The latter is the one I bought in the Raspberry Pi Store in Cambridge together with one of the Pi 4 units.
- with three different RaspberryPi units
so it seems to be an issue affecting more users (using a HDMI cable from your own store). I can't reproduce the issue by myself (I have a HDMI screen with this resolution but don't own any microHDMI cables - my use-cases are headless and the original cable is rather expensive here (~17$) so it won't make much sense for me to buy one just to reproduce a bug which seems to be already reproduced by different people around the world. Cause you already saw this, I don't think there's a need to open an github issue anymore, especially cause it's not known yet if this is a hardware or software issue.

Any comments on the revision history part?

User avatar
PeterO
Posts: 5440
Joined: Sun Jul 22, 2012 4:14 pm

Re: New Pi 4B v1.2?

Wed Dec 11, 2019 6:52 pm

chwe wrote:
Wed Dec 11, 2019 6:30 pm
so it seems to be an issue affecting more users (using a HDMI cable from your own store). I can't reproduce the issue by myself (I have a HDMI screen with this resolution but don't own any microHDMI cables - my use-cases are headless and the original cable is rather expensive here (~17$) so it won't make much sense for me to buy one just to reproduce a bug which seems to be already reproduced by different people around the world. Cause you already saw this, I don't think there's a need to open an github issue anymore, especially cause it's not known yet if this is a hardware or software issue.

Any comments on the revision history part?
You should try to keep up with current events and announcements rather that relying on out of date stories you find on the internet....
PhilE wrote: Friday's rpi-update firmware release includes a patch to choose an HDMI mode for 2560x1440@60Hz with alternate timing (reduced blanking) to avoid harmonics in the 2.4GHz band.
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

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 24934
Joined: Sat Jul 30, 2011 7:41 pm

Re: New Pi 4B v1.2?

Wed Dec 11, 2019 8:37 pm

chwe wrote:
Wed Dec 11, 2019 6:30 pm
Cause you already saw this, I don't think there's a need to open an github issue anymore, especially cause it's not known yet if this is a hardware or software issue.

Any comments on the revision history part?
We know what the issue is. Workaround mentioned above. More details will follow in due course.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I own the world’s worst thesaurus. Not only is it awful, it’s awful."

hippy
Posts: 6709
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 9:15 am

andrum99 wrote:
Sun Nov 24, 2019 10:44 am
Anything is possible. The fact that nobody has said what has changed in 1.2 would certainly lead one to suspect it is related to a future, as yet unannounced, feature.
Has anyone checked to see what make/size of Boot Eeprom is fitted to the 1.2 boards ?

It might be the same as 1.1 but I can see reasons it might change, why that would lead to a revision code kick, and why no one's currently saying what the change is.

Component relocations might just be incidental to what the actual change relates to.

andrum99
Posts: 1021
Joined: Fri Jul 20, 2012 2:41 pm

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 12:38 pm

hippy wrote:
Thu Dec 12, 2019 9:15 am
andrum99 wrote:
Sun Nov 24, 2019 10:44 am
Anything is possible. The fact that nobody has said what has changed in 1.2 would certainly lead one to suspect it is related to a future, as yet unannounced, feature.
Has anyone checked to see what make/size of Boot Eeprom is fitted to the 1.2 boards ?
They said when the Pi 4 was launched that they weren't committing to staying with the same size of EEPROM, although I'm not sure that would require a new board rev.

hippy
Posts: 6709
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 12:58 pm

andrum99 wrote:
Thu Dec 12, 2019 12:38 pm
They said when the Pi 4 was launched that they weren't committing to staying with the same size of EEPROM, although I'm not sure that would require a new board rev.
It's not a good idea to flash something larger than what it's going into so it would be desirable to know the size of Eeprom to avoid doing that. Using the revision number to indicate what size Eeprom the board has would be one way to do that, may be easier than using other means of determining size.

trejan
Posts: 1112
Joined: Tue Jul 02, 2019 2:28 pm

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 1:12 pm

hippy wrote:
Thu Dec 12, 2019 9:15 am
Has anyone checked to see what make/size of Boot Eeprom is fitted to the 1.2 boards ?
I think it is the same. The parts are still Winbond and the fuzzy lettering in the photos is plausible to match with the expected 4Hxxx 0Gxxx for the W25X40.
hippy wrote:
Thu Dec 12, 2019 12:58 pm
It's not a good idea to flash something larger than what it's going into so it would be desirable to know the size of Eeprom to avoid doing that.
You don't need the revision for that. You query the manufacturer ID and device ID from the chip itself.

hippy
Posts: 6709
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 2:17 pm

trejan wrote:
Thu Dec 12, 2019 1:12 pm
You don't need the revision for that. You query the manufacturer ID and device ID from the chip itself.
True, and I did say "Using the revision number to indicate what size Eeprom the board has would be one way to do that, may be easier than using other means of determining size". If the revision code is readily available and can simply be compared with some value it may mean less ROM code than having to interrogate the chip for example.

trejan
Posts: 1112
Joined: Tue Jul 02, 2019 2:28 pm

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 2:33 pm

hippy wrote:
Thu Dec 12, 2019 2:17 pm
True, and I did say "Using the revision number to indicate what size Eeprom the board has would be one way to do that, may be easier than using other means of determining size". If the revision code is readily available and can simply be compared with some value it may mean less ROM code than having to interrogate the chip for example.
Encoding it into the revision code would imply that the SoC boot ROM has changed to check for revision = 2 then new smaller EEPROM. Any changes to that SoC boot ROM are extremely expensive as you need to make a new mask for it and wait for the fab to build more. It would be easier + cheaper to directly probe the EEPROM capacity to allow for the part to be easily swapped out and just account for this in the boot ROM. It wouldn't be much more code.

Besides, there are OTP registers left over which is where low level configuration options like this would be kept. If the EEPROM size does need to be known then that would be where it is stored.

User avatar
rpdom
Posts: 16082
Joined: Sun May 06, 2012 5:17 am
Location: Chelmsford, Essex, UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 4:49 pm

The boot ROM doesn't need to check the revision. The first bit of bootcode in the EEPROM can do that. The bootrom code doesn't need to know how big the EEPROM is to execute the first bit of code in it.

IIRC the revision code is held in OTP, so no changes to the ROM needed for that either.

trejan
Posts: 1112
Joined: Tue Jul 02, 2019 2:28 pm

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 4:52 pm

rpdom wrote:
Thu Dec 12, 2019 4:49 pm
The boot ROM doesn't need to check the revision. The first bit of bootcode in the EEPROM can do that. The bootrom code doesn't need to know how big the EEPROM is to execute the first bit of code in it.
It is bootloaders all the way down. But yes, that would be a sensible way of doing it.
rpdom wrote:
Thu Dec 12, 2019 4:49 pm
IIRC the revision code is held in OTP, so no changes to the ROM needed for that either.
Yep. It is. The mask change would be if the SoC boot ROM loading the whole thing in one go and needing to special case revision 2 to handle that. Not needed if you do your method of another bootloader stage.

emma1997
Posts: 497
Joined: Sun Nov 08, 2015 7:00 pm
Location: New England (not that old one)

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 5:23 pm

Traditionally 'revision' number related to the PCB itself and not so much components on board or software (we had 'version' for that). Of course there were many exceptions by creative engineers and companies, making life slightly less convenient for everybody else, but what ya gonna do?

I think if any changes here were of any import at all then there would be announcements or at least straight answers.

hippy
Posts: 6709
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 5:42 pm

rpdom wrote:
Thu Dec 12, 2019 4:49 pm
The bootrom code doesn't need to know how big the EEPROM is to execute the first bit of code in it.
The Boot ROM would need to know if it were checking the size of the Boot Eeprom before flashing a Boot Eeprom Image from SD Card to Boot Eeprom. Though it wouldn't necessarily have to, could just write it all and not care if it doesn't work when it begins executing what it's loaded. Though that's not a particularly sensible idea given 'random code' can cause 'random damage' to anything it has access to.

It could be that the current Boot ROM writes 512KB regardless which could potentially cause a problem with smaller Eeprom, necessitates a change in Boot ROM.

I'm not convinced ROM and mask changes are all that expensive in the grand scheme of things and, if the 4B has sold 3 million or there about, that's likely at least £3 million profit of which some can be spent on changes.

Besides, if it needs to be done, it will be done, and when returns outweigh the cost, then why not ?

But the idea that the revision change might reflect a change in Boot Eeprom size was just a thought. I am well aware that it might not be, doesn't have to be, that there are other ways to determine Eeprom sizes.

trejan
Posts: 1112
Joined: Tue Jul 02, 2019 2:28 pm

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 5:47 pm

hippy wrote:
Thu Dec 12, 2019 5:42 pm
The Boot ROM would need to know if it were checking the size of the Boot Eeprom before flashing a Boot Eeprom Image from SD Card to Boot Eeprom.
The boot ROM doesn't do the flashing. It loads recovery.bin from the SD card which then handles the flashing of the bootcode EEPROM and the VL805 firmware EEPROM.

The recovery mechanism is that the boot ROM will look for recovery.bin on the SD card before loading bootcode from the EEPROM.

hippy
Posts: 6709
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: New Pi 4B v1.2?

Thu Dec 12, 2019 6:36 pm

trejan wrote:
Thu Dec 12, 2019 5:47 pm
hippy wrote:
Thu Dec 12, 2019 5:42 pm
The Boot ROM would need to know if it were checking the size of the Boot Eeprom before flashing a Boot Eeprom Image from SD Card to Boot Eeprom.
The boot ROM doesn't do the flashing. It loads recovery.bin from the SD card which then handles the flashing of the bootcode EEPROM and the VL805 firmware EEPROM.
You are right. I appear to have misunderstood the exact process :oops:

glenk
Posts: 6
Joined: Sat Aug 17, 2019 1:33 am

Re: New Pi 4B v1.2?

Wed Dec 18, 2019 11:12 pm

One person has now indicated that they have a RP4 with an apparently fully working USB C power input, not sure how word that correctly but you get my gist.

https://www.raspberrypi.org/forums/view ... 3&t=259629

They haven't confirmed that they have a v1.2 RP4 yet but it would seem likely.

Surely the Raspberry Pi Foundation can now tell us what the changes with v1.2 are.

W. H. Heydt
Posts: 11460
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: New Pi 4B v1.2?

Thu Dec 19, 2019 12:02 am

glenk wrote:
Wed Dec 18, 2019 11:12 pm
One person has now indicated that they have a RP4 with an apparently fully working USB C power input, not sure how word that correctly but you get my gist.

https://www.raspberrypi.org/forums/view ... 3&t=259629

They haven't confirmed that they have a v1.2 RP4 yet but it would seem likely.

Surely the Raspberry Pi Foundation can now tell us what the changes with v1.2 are.
In that thread NONE of it is clear. The OP hasn't responded with data to support the Pi4B version being tested and hasn't done anything to support the idea that an e-marked cable is being used. I don't think that thread even rises to the level of anecdote, let alone actual data.

folays
Posts: 6
Joined: Tue Dec 17, 2019 7:17 pm

Re: New Pi 4B v1.2?

Sun Dec 22, 2019 2:53 am

W. H. Heydt wrote:
Thu Dec 19, 2019 12:02 am
glenk wrote:
Wed Dec 18, 2019 11:12 pm
One person has now indicated that they have a RP4 with an apparently fully working USB C power input, not sure how word that correctly but you get my gist.

https://www.raspberrypi.org/forums/view ... 3&t=259629

They haven't confirmed that they have a v1.2 RP4 yet but it would seem likely.

Surely the Raspberry Pi Foundation can now tell us what the changes with v1.2 are.
In that thread NONE of it is clear. The OP hasn't responded with data to support the Pi4B version being tested and hasn't done anything to support the idea that an e-marked cable is being used. I don't think that thread even rises to the level of anecdote, let alone actual data.
I'm the OP of the linked thread above.
I admit that in my original post, I did not fully entered in fully "I will figure out what is going on" mode.

I stated my final words on https://www.raspberrypi.org/forums/view ... 9#p1583459
As I said in this post, the route on the back on the board is not the same near R1 and R79.

I also stated that I measured the resistance of both R1 and R79 : 5,1 kOhm.
And especially that I measured 10,23 kOhm between "left leg of R1" and "north leg of R79", which is the correct "addictive resistance value" of both of them being in "serial" and no longer incorrectly in the orignal-incorrect disposition of CC1 and CC2 sharing R1 and R79.

Regarding "the OP hasn't done anything to support the idea that an e-marked cable is being used", I expressly stated that :
- works with USB-C PD power supplies : Macbook 12" 2015, Anker PowerPort Speed+ Duo + "USB-C charger cable (apple)", LG 32ul950 screen (with a USB-C non-TB3 cable, plugged on the USB-C-TB3-70W-able port of this screen)
I guess that all "USB-C charger cable" from apple, no matter which "USB-C PD charger (apple)" they were sold with, would be e-marked.
In this case, I immediately adds that it works :
  • with the 27W USB-C apple charger from a MacBook 12" 2015 + USB-C<->USB-C cable officially distributed with it
  • with the 87W USB-C apple charger from a MacBook 15" 2018 + USB-C<->USB-C cable officially distributed with it
    Being 87W, and needing to be able to deliver the 20V power profile, I guess that the cable is e-marked
    (https://commons.wikimedia.org/wiki/File ... erRule.svg),
  • with the most expensive official USB-C<->USB-C TB3 cable from Apple plugged to "Anker PowerPort Speed+ Duo"
  • with the most expensive official USB-C<->USB-C TB3 cable from Apple plugged to a TB3 port of "MacBook Pro 15" 2018"
    (yes, the MacBook Pro "on battery, not plugged to A/C" is now able to power my RPi4. with the same USB-C-TB3 cable that is otherwise able to deliver screen to two ul32950 4K screens daisy chained together, which considering the specs, two 4K screens on the same USB-C port means that the port+cable are both TB3 capable.)
I aim to personally only own USB-C PD charger and "good quality cable that won't fry themselves or my hardwares"

The previous RPi4 4GB that I owned before (an original one from September 2019) only booted with a "USB-A iPad mini 3 charger" + "a stupid no-name USB-A<->USB-C cable". There was NO other combinaison of charger+cable that made the original RPi4 4GB works.

The new RPi4 4GB that I posted about (delivered on the 16th of December 2019) , works with everything USB-C that I own. (all PD capable, with the best rating of the Google spreadsheet guy who rated different brands, and also works with all TB3 capable things

ejolson
Posts: 4249
Joined: Tue Mar 18, 2014 11:47 am

Re: New Pi 4B v1.2?

Sun Dec 22, 2019 7:38 am

folays wrote:
Sun Dec 22, 2019 2:53 am
W. H. Heydt wrote:
Thu Dec 19, 2019 12:02 am
glenk wrote:
Wed Dec 18, 2019 11:12 pm
One person has now indicated that they have a RP4 with an apparently fully working USB C power input, not sure how word that correctly but you get my gist.

https://www.raspberrypi.org/forums/view ... 3&t=259629

They haven't confirmed that they have a v1.2 RP4 yet but it would seem likely.

Surely the Raspberry Pi Foundation can now tell us what the changes with v1.2 are.
In that thread NONE of it is clear. The OP hasn't responded with data to support the Pi4B version being tested and hasn't done anything to support the idea that an e-marked cable is being used. I don't think that thread even rises to the level of anecdote, let alone actual data.
I'm the OP of the linked thread above.
I admit that in my original post, I did not fully entered in fully "I will figure out what is going on" mode.

I stated my final words on https://www.raspberrypi.org/forums/view ... 9#p1583459
As I said in this post, the route on the back on the board is not the same near R1 and R79.

I also stated that I measured the resistance of both R1 and R79 : 5,1 kOhm.
And especially that I measured 10,23 kOhm between "left leg of R1" and "north leg of R79", which is the correct "addictive resistance value" of both of them being in "serial" and no longer incorrectly in the orignal-incorrect disposition of CC1 and CC2 sharing R1 and R79.

Regarding "the OP hasn't done anything to support the idea that an e-marked cable is being used", I expressly stated that :
- works with USB-C PD power supplies : Macbook 12" 2015, Anker PowerPort Speed+ Duo + "USB-C charger cable (apple)", LG 32ul950 screen (with a USB-C non-TB3 cable, plugged on the USB-C-TB3-70W-able port of this screen)
I guess that all "USB-C charger cable" from apple, no matter which "USB-C PD charger (apple)" they were sold with, would be e-marked.
In this case, I immediately adds that it works :
  • with the 27W USB-C apple charger from a MacBook 12" 2015 + USB-C<->USB-C cable officially distributed with it
  • with the 87W USB-C apple charger from a MacBook 15" 2018 + USB-C<->USB-C cable officially distributed with it
    Being 87W, and needing to be able to deliver the 20V power profile, I guess that the cable is e-marked
    (https://commons.wikimedia.org/wiki/File ... erRule.svg),
  • with the most expensive official USB-C<->USB-C TB3 cable from Apple plugged to "Anker PowerPort Speed+ Duo"
  • with the most expensive official USB-C<->USB-C TB3 cable from Apple plugged to a TB3 port of "MacBook Pro 15" 2018"
    (yes, the MacBook Pro "on battery, not plugged to A/C" is now able to power my RPi4. with the same USB-C-TB3 cable that is otherwise able to deliver screen to two ul32950 4K screens daisy chained together, which considering the specs, two 4K screens on the same USB-C port means that the port+cable are both TB3 capable.)
I aim to personally only own USB-C PD charger and "good quality cable that won't fry themselves or my hardwares"

The previous RPi4 4GB that I owned before (an original one from September 2019) only booted with a "USB-A iPad mini 3 charger" + "a stupid no-name USB-A<->USB-C cable". There was NO other combinaison of charger+cable that made the original RPi4 4GB works.

The new RPi4 4GB that I posted about (delivered on the 16th of December 2019) , works with everything USB-C that I own. (all PD capable, with the best rating of the Google spreadsheet guy who rated different brands, and also works with all TB3 capable things
Thanks for the detailed reply and analysis. Because people trying out a Pi for the first time will likely want to reuse existing power supplies, it is good to know that the updated version of the 4B works. I guess those who have been holding off on buying a Pi 4B because of the USB-C power blunder desribed here now have a reason to buy one for Christmas.

User avatar
Imperf3kt
Posts: 3219
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: New Pi 4B v1.2?

Mon Dec 23, 2019 10:15 am

I would say more of a reason to buy one for 2020
Christmas is a bit close to get one delivered (unless you're lucky enough to live somewhere you can actually walk into a store and buy one in person)
55:55:44:44:4C
52:4C:52:42:41

User avatar
PeterO
Posts: 5440
Joined: Sun Jul 22, 2012 4:14 pm

Re: New Pi 4B v1.2?

Mon Dec 23, 2019 10:34 am

ejolson wrote:
Sun Dec 22, 2019 7:38 am
I guess those who have been holding off on buying a Pi 4B because of the USB-C power blunder desribed here now have a reason to buy one for Christmas.
Considering that they have been selling as fast as they can be made, I would estimate that the number of people who " have been holding off on buying a Pi 4B because of the USB-C power blunder" can be counted on the fingers of one hand.

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

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 24934
Joined: Sat Jul 30, 2011 7:41 pm

Re: New Pi 4B v1.2?

Mon Dec 23, 2019 10:39 am

This is a blunder

https://www.space.com/boeing-starliner- ... pdate.html

Missing out a resistor that affect almost no-one is a minor mistake.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I own the world’s worst thesaurus. Not only is it awful, it’s awful."

User avatar
Gavinmc42
Posts: 4292
Joined: Wed Aug 28, 2013 3:31 am

Re: New Pi 4B v1.2?

Mon Dec 23, 2019 11:03 am

https://www.space.com/boeing-starliner- ... pdate.html
It will be interesting when space craft are as reliable as autonomous cars?
India and Israel crashing onto the Moon.
Will the next generation of space vehicles have Tesla autopilot systems on board?

Some of this stuff makes The Expanse Technology look dated.
Stainless steel rockets :o
Which Pi generation will be able to do what Tesla CPU's can do now?

Probably not 1.2?
Still trying to get Yolov3 to work ;)
I'm dancing on Rainbows.
Raspberries are not Apples or Oranges

Return to “General discussion”