Intel burnt ICH4 ICH5 South Bridges!

[Gnu_Raiz]

Just when you thought it was safe to buy Intel again it turns out that
people are having problems with USB and these ICH4 ICH5 Bridges.
<Warning page is not in english>

http://www.ocworkbench.com/ocwbcgi/newsprochinese/viewnews.cgi?id=EEpFuykVFuShiCjTDc
http://www.ocworkbench.com/index.stm

Gnu_Raiz

 

[George Macdonald]

Just when you thought it was safe to buy Intel again it turns out that
people are having problems with USB and these ICH4 ICH5 Bridges.
<Warning page is not in english>

http://www.ocworkbench.com/ocwbcgi/newsprochinese/viewnews.cgi?id=EEpFuykVFuShiCjTDc
http://www.ocworkbench.com/index.stm

USB and ESD is a bad mixture... though I'd hardly think it'd be limited to
Intel chipsets.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[Gnu_Raiz]

I would of thought that the motherboard would help disipate the ESD, I
could understand if the motherboard was not ground right. This makes
you think that the average person just plugging in a device frys the
south bridge.

Gnu_Raiz

 

[John Smithe]

I would of thought that the motherboard would help disipate the ESD, I
could understand if the motherboard was not ground right.

The motherboard is not going to protect the IC's from ESD even with good
grounding.

This makes
you think that the average person just plugging in a device frys the
south bridge.

Gnu_Raiz

The way to prevent ESD damage is to not let ESD occur. Ground yourself
before getting your fingers, or whatever, near connectors or other
electronic devices. 'Hot Swap' and 'Hot Plug' is not a gaurantee of
100% protection. Sure, lots of electronic devices have protection
circuitry, but that circuitry has its limitations. You'd be
surprised at the limitations of ESD protection circuits.

P.S. Just because you didn't see or feel an electrostatic discharge, does
not mean that a discharge did not happen.

Just trying to offer some friendly advice.

 

[George Macdonald]

I would of thought that the motherboard would help disipate the ESD, I
could understand if the motherboard was not ground right. This makes
you think that the average person just plugging in a device frys the
south bridge.

I'm surprised there aren't more frequent indidents with USB. I haven't
read the details on USB's safety precautions or any staggering of
connection strips to allow hot plug in the first place but it doesn't take
much to blow an IC. Just a few years ago there was a power supply which
was blowing VIA South Bridge chips just because the power-on voltage
ramping wasn't quite right.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[alexi]

I'm surprised there aren't more frequent indidents with USB. I haven't
read the details on USB's safety precautions or any staggering of
connection strips to allow hot plug in the first place but it doesn't take
much to blow an IC.

The USB connectors are staggered by design. All USB connectors
are required to have a metal shield-housing. The shield is connected
via a separate wire inside the cable. The shield makes the contact first
and is supposed to discharge whatever the charge is into protective
ground. Then signal ground and power make the contact. These nets are
usually
pretty robust with regard to ESD. D+ and D- make connections last,
so technically they are not expected to carry any charge. Even given
this design, some manufacturers break the integrity of USB
cables/connectors and test naked wires directly, and all USB inputs
are expected to pass 4kV direct discharge, and 12-16kV air-gap
discharge. So, given many hundreds of millions PC all with built-in
USB ports and even bigger number of USB hubs and devices,
I found the report of ESD-broken bridges pretty strange.
Of course, there are some devices (I've seen some flash memory
keychains) without the metal shield. These devices are in violation
of USB specifications and technically should not be used. However,
even without the shield, power and ground should take the discharge
pretty well and no damage should happen.

Just a few years ago there was a power supply which
was blowing VIA South Bridge chips just because the power-on voltage
ramping wasn't quite right.

Yep, on some poorly-designed multi-rail chips the dynamics of voltage
ramp at start-up may be critically important.

- aap

 

[George Macdonald]

The USB connectors are staggered by design. All USB connectors
are required to have a metal shield-housing. The shield is connected
via a separate wire inside the cable. The shield makes the contact first
and is supposed to discharge whatever the charge is into protective
ground. Then signal ground and power make the contact. These nets are
usually
pretty robust with regard to ESD. D+ and D- make connections last,
so technically they are not expected to carry any charge. Even given
this design, some manufacturers break the integrity of USB
cables/connectors and test naked wires directly, and all USB inputs
are expected to pass 4kV direct discharge, and 12-16kV air-gap
discharge. So, given many hundreds of millions PC all with built-in
USB ports and even bigger number of USB hubs and devices,
I found the report of ESD-broken bridges pretty strange.
Of course, there are some devices (I've seen some flash memory
keychains) without the metal shield. These devices are in violation
of USB specifications and technically should not be used. However,
even without the shield, power and ground should take the discharge
pretty well and no damage should happen.

Thanks for the run-down. I've been wondering if it might have something to
so with less well designed case front panel USB connectors - something the
mbrd mfr has no control over. I've seen where some case mfrs, to allow USB
2.0 operation, just eliminated the resistance in the USB 1.1 RC low-pass
filter. Possibly the shield is not connected to any "protective ground"
with them too?

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??