Author Topic: EEVblog #1164 - Xbox Baptism Of Fire  (Read 5679 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29680
  • Country: au
    • EEVblog
EEVblog #1164 - Xbox Baptism Of Fire
« on: December 28, 2018, 03:20:10 am »
A retro look at the 2005 Xbox 14 million unit power cord replacement issue and a teardown and analysis of the supplied protection device from Microsoft.

 

Offline Tek_TDS220

  • Regular Contributor
  • *
  • Posts: 63
  • Country: 00
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #1 on: December 28, 2018, 03:55:34 am »
I'm not sure, but is the 'inverted' design used to keep the power low (and the relay off) during operation? 

Incidentally iFixit did a teardown of the new Xbox One X (https://ifixit.org/blog/9450/xbox-one-x-teardown/).  It's beautiful.
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 5846
  • Country: us
    • Personal site
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #2 on: December 28, 2018, 03:55:43 am »
I doubt they actually designed this thing on the spot. This is probably some existing product relabeled and trip current adjusted.
Alex
 

Offline johnlsenchak

  • Frequent Contributor
  • **
  • Posts: 280
  • Country: us
  • js@antihotmail.com
    • paypal.me/johnsenchak
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #3 on: December 28, 2018, 04:00:22 am »


Maybe the board  monitors    current leakage  to ground  for shock  hazard  in comparison  to  the current load of the XBOX

Electronic GFCI
https://home.howstuffworks.com/question117.htm
« Last Edit: December 28, 2018, 04:07:37 am by johnlsenchak »
John Senchak "Daytona  Beach  Florida "
 jls (at)  antihotmail.com   http://www.antihotmail.com
https://www.facebook.com/john.senchak.1
 

Offline sakujo7

  • Contributor
  • Posts: 35
  • Country: au
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #4 on: December 28, 2018, 04:32:33 am »
The solenoid coil placement could be to cover an unlikely case where the earth leakage circuit itself or the rectifier fails and becomes a short. They clearly wanted something that would reduce the risk of fire no matter what.

A youtube commenter suggested the micro could be looking for a particular wave shape that indicates arcing, which would allow it to act like a fast-blow fuse yet not trip on inrush current. If it's not something like that, then it's definitely overkill...
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29680
  • Country: au
    • EEVblog
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #5 on: December 28, 2018, 05:13:00 am »
A youtube commenter suggested the micro could be looking for a particular wave shape that indicates arcing, which would allow it to act like a fast-blow fuse yet not trip on inrush current. If it's not something like that, then it's definitely overkill...

That might make sense. The PIC 12F675 would have just enough grunt to do that crudely I guess.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29680
  • Country: au
    • EEVblog
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #6 on: December 28, 2018, 06:05:02 am »
Apparently this is an "Arc Fault Circuit Interrupter" (AFCI) that continually monitors the current for the presence of arcs caused by such an intermittently faulty power connection. Hence the complexity and micro that would be using the ADC to continually monitor the current. It's not just a simple 610mA electronic fuse, although it almost certainly has that function too, hence the label.

These seem standard in US outlets since 2014?

https://en.wikipedia.org/wiki/Arc-fault_circuit_interrupter
« Last Edit: December 28, 2018, 06:08:14 am by EEVblog »
 

Offline Whales

  • Frequent Contributor
  • **
  • Posts: 840
  • Country: au
    • Halestrom
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #7 on: December 28, 2018, 06:38:21 am »
Dave: what makes you think Microsoft designed this?  Wouldn't it make more sense that this is an off-the-shelf product with a custom trip level and a rebadge?  Esp given the custom relay construction.

https://duckduckgo.com/?q=inline+RCD&iar=images&iax=images&ia=images
https://duckduckgo.com/?q=inline+AFCI&iar=images&iax=images&ia=images
« Last Edit: December 28, 2018, 06:40:09 am by Whales »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 29680
  • Country: au
    • EEVblog
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #8 on: December 28, 2018, 06:41:37 am »
Dave: what makes you think Microsoft designed this? 

They have the clout to do this, and the Not Invented Here syndrome perhaps.

Quote
Wouldn't it make more sense that this is an off-the-shelf product with a custom trip level and a rebadge?  Esp given the custom relay construction.

With hindsight, yes.
 

Offline helius

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #9 on: December 28, 2018, 06:45:54 am »
@15:00:
There were power strips (APC was one brand) that were recalled for causing fires due to shorted, overheating MOVs*. The remedy was to redesign them with thermal fuses held against the MOVs.
MOVs are sacrificial components, but unfortunately their most common failure mode is to go permanently low resistance, and further heating in a death spiral. It's good to see that Microsoft designed them in correctly, even in 2005, several years before the APC recall happened.

* https://www.cpsc.gov/Recalls/2013/schneider-electric-recalls-apc-surge-protectors
https://recall.apc.com/en
« Last Edit: December 28, 2018, 06:48:02 am by helius »
 

Offline Whales

  • Frequent Contributor
  • **
  • Posts: 840
  • Country: au
    • Halestrom
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #10 on: December 28, 2018, 06:59:57 am »
Quote
Wouldn't it make more sense that this is an off-the-shelf product with a custom trip level and a rebadge?  Esp given the custom relay construction.

With hindsight, yes.

I've love to lookup the codes next to the tick of approval c-tick, see if that brings more info about who the OEM might be.



I think the place I'm supposed to search is ERAC.  Alas I'm having no luck -- I can't decipher what their terminology means and there are lots of server errors.  One error simply dead "Dead!" :D  The searches look like they need at least 5 fields to be filled exactly correctly -- quite a way of telling people to go away.

Anyone know how to search those codes next to the c-tick up?

EDIT: Turns out that's not the Australian tick of approval, it's a c-tick.  As a consumer: I had no idea, that's really bad symbolic language.
« Last Edit: December 28, 2018, 07:13:44 am by Whales »
 

Offline draconx

  • Newbie
  • Posts: 1
  • Country: ca
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #11 on: December 28, 2018, 08:02:55 am »
The separate test board has the RU mark on the silkscreen which means it was independently tested by Underwriters Laboratory, right?

So that perhaps justifies a two board design... so the test board can be certified separately from the rest of the product.  Could be a requirement to sell this in some jurisdictions?
 

Offline JoeMuc2013

  • Contributor
  • Posts: 21
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #12 on: December 28, 2018, 09:28:35 am »
Hey all,

all tech details aside, I still don't understand what is the approach here. So some Xbox models (0.01%, right) released the magic smoke? Why? And how is an inline fuse going to prevent that?
Do they try to save the precious Xbox by just switching it completely off once the current draw goes beyond expected levels? In that case, that's not a particularly good solution, or is it? Imagine the rage in MMO or RPG gamers if their console cuts out in the middle of a fight? The hardware might have survived the power issue then, but will it survive the destructive energy of mad users?
Maybe I'm completely wrong with this but afaik a fuse is just to cut power for safety. If the design fault is in the built-in power supply of the Xbox, wouldn't that have to be replaced instead of this makeshift solution? And what caused the power supplies to fail in the first place?

Thanks for any info on this!
I have never owned a console but this could just as well happen with any other consumer device, so it's a generally interesting matter.

Happy new year in advance to everyone  ;D
 

Offline riccardo.pittini

  • Contributor
  • Posts: 39
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #13 on: December 28, 2018, 11:03:34 am »
Dave: One comment MOV and Thermal fuse

You can buy directly MOV with integrated thermal fuses. If you make a decent power supply everything should be after the fuse, including MOVs (in some cases it's not possible, e.g. when you design a power supply to withstand large surge pulses).
MOVs with time they degrade, heating up, so it is a good design rule to place a MOV after the main fuse. If the MOV is directly in parallel to the AC line then use a thermal fuse.
 

Offline Whales

  • Frequent Contributor
  • **
  • Posts: 840
  • Country: au
    • Halestrom
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #14 on: December 28, 2018, 11:09:42 am »
all tech details aside, I still don't understand what is the approach here. So some Xbox models (0.01%, right) released the magic smoke? Why? And how is an inline fuse going to prevent that?
Do they try to save the precious Xbox by just switching it completely off once the current draw goes beyond expected levels?

Read the other posts in this topic and the comments on youtube: it's not a standard fuse-style circuit breaker, it's an AFCI. 

Quote
In that case, that's not a particularly good solution, or is it? Imagine the rage in MMO or RPG gamers if their console cuts out in the middle of a fight? The hardware might have survived the power issue then, but will it survive the destructive energy of mad users?
Maybe I'm completely wrong with this but afaik a fuse is just to cut power for safety. If the design fault is in the built-in power supply of the Xbox, wouldn't that have to be replaced instead of this makeshift solution? And what caused the power supplies to fail in the first place?

You're right it's not the best solution.  It can only try to prevent smoke/fire after a fault happens, but this might be enough in the eyes of Microsoft.  Even if it fails to do so 10% of the time it could still theoretically reduce the number of failures to '3' instead of '30'.   Additionally: if people keep turning it back on (because it keeps tripping) and then have a house fire, Microsoft can then say it's their fault for ignoring the clear warnings from a safety protection device.

The motivations and viewpoint of a company are different to the motivations and viewpoints of an ideal society.
« Last Edit: December 28, 2018, 11:11:15 am by Whales »
 

Offline AndyC_772

  • Super Contributor
  • ***
  • Posts: 3446
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #15 on: December 28, 2018, 12:20:35 pm »
Imagine the rage in MMO or RPG gamers if their console cuts out in the middle of a fight? The hardware might have survived the power issue then, but will it survive the destructive energy of mad users?

I'm imagining it right now, it's both hilarious, and hilariously unimportant by any standard.

If an AFCI makes the difference between "my Xbox cut out and lost me some game progress" vs "my Xbox cut out and set fire to my home", then it's done everything that was required.

Remember, a games console is a corner case in terms of safety and reliability, ie.

- it's made in huge numbers, and is sold to people who will take every possible level of care of it, from extreme mechanical sympathy to utterly reckless abuse - so it needs to be extraordinarily safe, but...

- if it simply fails to function, the real-world consequences are almost definitively trivial. "Someone can't play a video game" is about the single most unimportant consequence a failure could possibly have, so any need for reliability is all about the manufacturer's reputation. That's a commercial decision, and I can easily see how replacing a power brick made more sense than replacing entire consoles.
 
The following users thanked this post: rs20

Offline JoeMuc2013

  • Contributor
  • Posts: 21
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #16 on: December 28, 2018, 12:32:32 pm »
If an AFCI makes the difference between "my Xbox cut out and lost me some game progress" vs "my Xbox cut out and set fire to my home", then it's done everything that was required.

I agree, technically it prevents the worst. BUT: if this is a known design flaw in the Xbox, serious corporations would recall the units in question, and fix them, no matter the cost. This action is certainly more expensive than providing an ACFI to all users, no matter whether they are affected or not, but tackling a problem at its root would establish a good hardware manufacturer's reputation whereas this is a ridiculous cheat of a solution that I would never accept if I were an Xbox owner.
If you bought an electric vehicle that is under the known risk of fire in mid-travel, and the manufacturer sends you an AFCI for you to "fix" it yourself, which would cause the vehicle to do an emergency stop and power down once something odd was detected, this behavior would cause a class action lawsuit against the vehicle's manufacturer. I know, it's an odd comparison but anyway... what I want to say is that Microsoft gets out of this too easily. It's their fault, not the customer's.

So yeah, I'm happily not an Xbox owner, but after this story I'll consider twice before I ever buy any mains-powered hardware from Microsoft.
 

Offline AndyC_772

  • Super Contributor
  • ***
  • Posts: 3446
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #17 on: December 28, 2018, 01:35:16 pm »
I offer by way of counter-example the fact that this particular serious company, on this occasion, chose to do otherwise.

I also suggest that it would be a mistake to avoid buying mains powered equipment from a company which has clearly already had an expensive lesson in reliability, and which should undoubtedly have a check for this type of design flaw as part of its QA procedure now. Holding a grudge beyond the point at which a company has released new designs that are free of this flaw doesn't help anybody.

For an automotive product, the consequences of failure can be incomparably different. If a fault in my car causes it to shut down on the highway, then that's a big deal - but if it just means the CD player doesn't work any more, that's inconsequential unless the fault also starts a fire.
 
The following users thanked this post: thm_w

Offline Ferroto

  • Frequent Contributor
  • **
  • Posts: 287
  • Country: ca
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #18 on: December 28, 2018, 07:58:17 pm »
AFCI's are a huge pain in the ass and will trip if you have any inductive loads such as motors or ballasts on the same branch circuit. They are also highly susceptible to RF interference as seen here.

 

Offline Rasz

  • Super Contributor
  • ***
  • Posts: 2277
  • Country: 00
    • My random blog.
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #19 on: December 28, 2018, 08:45:38 pm »
Dell and HP had similar problems, lucky for them laptop power supplies are easier to recall avoiding half ass bodges like one demonstrated here.

https://www.pcworld.com/article/2599100/hp-recalls-6-million-laptop-power-cords-that-can-pose-fire-and-burn-hazards.html
https://www.cnet.com/news/millions-of-dell-power-adapters-recalled/

Apple is clueless when it comes to power delivery, and got almost EVERY SINGLE laptop charger wrong, starting with powerPC laptops, and ending at USB-C cables  :-DD |O most catching fire too.
https://discussions.apple.com/thread/551178
https://www.macworld.com/article/1140453/Apple_sued_over_alleged_MagSafe_fire_hazard.html
https://discussions.apple.com/thread/6037659
https://www.apple.com/support/ac-wallplug-adapter/
https://www.apple.com/ie/support/usbc-chargecable/
Who logs in to gdm? Not I, said the duck.
My fireplace is on fire, but in all the wrong places.
 

Offline orion242

  • Supporter
  • ****
  • Posts: 697
  • Country: us
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #20 on: December 28, 2018, 08:55:19 pm »
AFCI's are a huge pain in the ass and will trip if you have any inductive loads such as motors or ballasts on the same branch circuit. They are also highly susceptible to RF interference as seen here.


Guy I work with just finished building a new house.  First power outage he fired up his generator, only to find out none of the AFCI's would hold under gen power.  Brilliant!
« Last Edit: December 28, 2018, 08:56:51 pm by orion242 »
 

Offline Barny

  • Frequent Contributor
  • **
  • Posts: 283
  • Country: at
  • I'm from Austria, not Australia ;)
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #21 on: December 28, 2018, 09:54:08 pm »
Microsoft continued the tradition of bad power suplys.
I've the first generation of Surface Pro.
The power suply is completly without bend protection.
(Primary and secondary)

I replaced the ca. 300mm (ca. 12 inch)  long power cord after it started to arc & smell with an new one after one week of usage.
I planed to replace the cable from the beginning, because the original was next to unusable short.
Ca. 3 years later I got a new power cord because of an recall.

The low voltage side got an bend-protection made of multible layers of heat shrink and hot glue.

Here is the recall-site from microsoft:
https://support.microsoft.com/en-us/help/4040585/surface-ac-power-cord-recall-surface-pro-surface-pro2-surface-pro3
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 6221
  • Country: de
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #22 on: December 28, 2018, 10:25:17 pm »
The design of the connector mainly hold by the solder joints could be the problem and reason for the recall.  So it would be a fault waiting to happen in units that see a lot of stress on the connector and maybe have a not so good solder quality.

The protection device might very well be a separate product already designed before. Still the construction looks a little odd - not made to be sold in the millions. I can under stand that this could be cheaper than a full recall of the XBOX itself. They likely would have to replaced the PS board and organize twice the shipping.  When in a hurry they may not have the time to design a cheaper protection box and get it tested.

An really interesting tear-down.
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 5846
  • Country: us
    • Personal site
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #23 on: December 28, 2018, 10:27:46 pm »
But I still don't see how the crack on the connector will be remedied by the external device. And I don't see how anything more serious than a crack can happen here.
Alex
 

Offline helius

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: EEVblog #1164 - Xbox Baptism Of Fire
« Reply #24 on: December 28, 2018, 10:42:01 pm »
The failure mode is that the high contact resistance makes the connector pins red hot. There is a chance the PCB material begins to smolder. The hope is that arcing inside the joint can be detected using that device and shut off power before it gets hot enough to ignite.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf