#1050 HIGH BTest-4: Need new board ID mechanism

[Zarro Boogs per Child]

#1050: Need new board ID mechanism
--------------------------------+-------------------------------------------
  Reporter:  wmb at firmworks.com  |       Owner:  wad     
      Type:  defect             |      Status:  reopened
  Priority:  high               |   Milestone:  BTest-4 
 Component:  hardware           |     Version:          
Resolution:                     |    Keywords:          
  Verified:  0                  |  
--------------------------------+-------------------------------------------
Changes (by frief):

  * status:  closed => reopened
  * resolution:  fixed =>

Comment:

 Sorry, I object to this being set to fixed with just a single line
 comment:
 "At the circuit level, this is specified by the ratio of the values of two
 resistors (R535 and R534), providing a voltage on the EC_ID signal input
 to the EC."
 at http://wiki.laptop.org/go/Ec_specification

 It's probably not as straight forward as the comment suggests. The link
 does not point to a related schematic and the text does not mention the
 values of R535 and R534 or additional circuitry besides R535 and R534 so I
 think I'd better ask.

 Could you specify whether a) an additional capacitor buffering the voltage
 is used (so that the current through a high impedance voltage divider can
 be negligible) or b) a FET or an otherwise switched 3V3 voltage and a low
 impedance voltage divider are used? The size of capacitor would have to be
 matched with the resistors for an appropriate time constant.

 Off-the-shelf resistors with 5% tolerance won't do to reliably
 discriminate 0..0xff into 16 intervals (indeed if worst case error
 calculation takes into account: conversion noise, nonlinearity, bias
 current @Tmax, ADC input current/capacity, read repetition rate, power
 spikes, resistor tolerance...  this might be hard to meet with a 2%
 tolerance resistor pair).

 Please add meat or give an ack that these concerns have been addressed.
 Sorry for reopening the ticket. (The 8051 instruction set of the embedded
 controller would allow for non power of 2 intervals at almost no overhead
 so if needed there is some flexibility there)

 Greetings,

 Frieder

-- 
Ticket URL: <http://dev.laptop.org/ticket/1050#comment:7>
One Laptop Per Child <http://laptop.org/>