Opened 10 years ago

Closed 10 years ago

Last modified 20 months ago

#473 closed defect (fixed)

Keyboard keys occasionally stop working

Reported by: mfoster Owned by: wad
Priority: blocker Milestone:
Component: hardware Version:
Keywords: relnote Cc:
Blocked By: Blocking:
Deployments affected: Action Needed:
Verified: yes


Hi, Folks,

On 11/16, when B1 units first entered final functional testing, we were there to watch what failures occurred. Fairly quickly, a few units appeared which had keyboard failures where groups of keys would not work. Which keys would not work varied from machine to machine. Though the keyboard appeared to be at fault, continued investigation revealed that the keyboard scanning microprocessor on the touchpad assembly was sometimes seeing latchup on its pins. Latchup is a condition in a CMOS IC where a pin will be "stuck" at a high or low voltage. When this occurs, the stuck pin prevents the keyboard microprocessor from seeing keys change state. Note that latchup only happens when excessive voltages are present, as when ESD (Electro-Static Discharge - i.e. a "static zap") occurs.

The root cause of this problem is that the mechanical design in the B1 systems does not properly ground the keyboard baseplate. While an attempt was made to do this in the design, the metal components do not make proper contact with the grounded hinge assembly. As a consequence, any static buildup on the keyboard can only discharge through the keyboard controller's scanning pins, then back up the PS/2 cable to the motherboard. This leads to some keys on the keyboard not responding.

While this was found and fixed on the same day, the true resolution was identified just as the last few B1 units were being built on the line, so all B1 units are susceptible to this failure. The workaround is to power the system off and on again, which will reset the latchup condition in the keyboard microrocessor, and allow the keys to function once again.

This will obviously be a very high priority fix for B2.

Change History (10)

comment:1 Changed 10 years ago by mfoster

  • Component changed from distro to mechanical
  • Status changed from new to assigned

comment:2 Changed 10 years ago by mfoster

  • Priority changed from high to blocker

The specific corrections which will be made in the B2 systems to correct this include:

1) Modifying the metal stengthening plate which lies below the hinge assembly in the system's base assembly, so that it proper contacts the grounded hinge.

2) Extending the aluminum baseplate which surrounds the touchpad to provide proper electrical contact with the touchpad/keyboard controller PCB.

comment:3 Changed 10 years ago by jg

  • Keywords relnote added

comment:4 in reply to: ↑ description Changed 10 years ago by 90030501

Frank:Add grounding on Hinge position and TP. We will get sample on 12/10/06'

comment:5 Changed 10 years ago by 90030501

Frank:Add grounding on Hinge position and TP. We will get sample on 12/10/06'

comment:6 Changed 10 years ago by 90030501

Frank: Sorry we did not get sample on time. Cause we stop some tooling for 12/03/06' meeting. We will get sample on 12/25/06'.

comment:7 Changed 10 years ago by jg

Just checking: will this be fixed in BTest-2?

comment:8 Changed 10 years ago by jg

  • Component changed from mechanical to hardware
  • Owner changed from mfoster to wad
  • Status changed from assigned to new

It is believed fixed in B2.

I'm closing this one; if anyone sees it on BTest-2 hardware, please reopen this and mark it a blocker.

comment:9 Changed 10 years ago by wad

  • Resolution set to fixed
  • Status changed from new to closed

comment:10 Changed 20 months ago by Quozl

  • Milestone BTest-2 deleted

Milestone BTest-2 deleted

Note: See TracTickets for help on using tickets.