[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
Volusia County October 9th Election Material Failure Report
- To: "Global Support \(E-mail\)" <support@gesn.com>
 
- Subject: Volusia County October 9th Election Material Failure Report
 
- From: "Ian S. Piper" <ian@gesn.com>
 
- Date: Thu, 8 Nov 2001 12:58:03 -0600
 
- Cc: "Barney Lucas \(E-mail\)" <barney@gesn.com>
 
- Importance: Normal
 
While Charlie Conrad and I were in Volusia County for 
the Nov. 6th election, we examined the materials that had failed during the Oct. 
9th election.  Below is a report of those failures.  I'll email the Nov. 6th material failure list 
early next week.  A report containing the info below and the Nov. 6th 
info will also be faxed to the Volusia County Election Director 
(Deanie Lowe).
 
Warning:  The answers for the memory card errors 
experienced on Volusia County's infra-red machines do NOT apply for similar 
errors that you might have experienced with visible light machines.  
The conditions that existed in the infra-red units have not been proven 
to affect visible light units.  In other words, we still need to find 
the answers for visible light unit memory card 
errors.
 
BTW:  Word should go out through Support to 
existing infra-red customers, re-iterating the importance of have good charged 
power supply batteries in their units.  The recommended test for 
determining if a battery can hold its charge is to successfully print two to 
three diagnostic printer tests using only battery power and without a 
"battery low" message being displayed or, obviously, without the unit 
dying.  Of course, the battery charge would then have to be topped up after 
the test, but at least it will be known that the battery is capable of 
holding its charge.  Just plugging a unit in overnight and running a 
small test deck through it in the morning is not 
sufficient.
 
 
Ian 
 
 
1. October 9, 2001 
election.  
Precinct 815 wouldn't 
upload to VTS.  VTS generates 
"Already uploaded error."  
  a) VTS:  The audit log 
  reports that an upload started for Precinct 815 but it never reports that it 
  completed.  Examination of AV-OS unit (SN 70598) found that the modem 
  serial port test failed.  Obviously the modem worked well enough to allow 
  VTS to initialize communications and report that the upload had started, but 
  the modem must have failed during data transmission before the transmission was 
  completed.  The audit log also 
  showed that VTS rejected subsequent upload 
  attempts, reporting that the precinct had already been uploaded.  
   
  Tab has confirmed that 
  although VTS shouldn't flag the precinct as "uploaded" until it is actually complete, there is 
  a possibility that this could happen.  
   
  b) AV-OS Modem:  It was found 
  that the ribbon cable connector on the modem serial interface did not have 
  sufficient height to contact the locking levers on the CPU Board header.  
  The modem connectors were reseated and the 
  modem passed subsequent tests.
   
  The ribbon 
  cable connector for the modem serial interface requires a strain relief to be 
  added to the connector to increase the height of the connector so the 
  locking levers of the CPU board header make contact with the top of the 
  connector to hold it in place.  A modem serial interface using 
  ribbon cable was only required for the older "2400" modems.  The 
  14.4K modems required a different cable that has a full height 
  connector.
2. October 9, 2001 
election.  
On opening the polls, three 
(3) units printed the top half of the asterisks (first line of characters in a 
report header) and continued to do so approximately every half minute.  (AV-OS SN 70564, 70660, 
70673)
  The battery 
  was either dead or low 
  on these units.  Testing on each unit repeated the error 
  described.  When a unit 
  was powered up in the election mode, a zero totals report was supposed to be generated.  In an 
  infra-red AV-OS, the printer needs to draw power from a charged battery to 
  operate.  With a dead or low 
  battery, the printer only printed 
  three dot lines (the top half of the asterisks 
  that are printed at the beginning of a label header) before it 
  dragged the logic power down to a level 
  that triggered a reset (4.75Vdc.)  
  The AV-OS reset and tried to generate 
  another zero totals report, but the printer again dragged the logic power down to the reset trigger 
  level, starting this cycle again.  In 
  the midst of these resets, SN 70660 also experienced an "OK to reformat this card?" error which 
  occurred as a result of the above condition.
   
  The infra-red units require 
  a charged power supply 
  ("PS") battery to allow printing to occur.  
  Expected sealed lead acid battery 
  life is five years.  Most of the failed PS batteries are five to six 
  years old.  Others may have been damaged by discharging to a flat 
  condition.
3.  October 9, 
2001 election.  
On opening the polls, the 
AV-OS unit reports "Memory Card Removed - 
System Halted".  (AV-OS SN 70560)
  The battery was 
  dead on this unit.  Testing 
  the unit repeated the error described.  When the unit was powered 
  up in the election mode, a zero totals report was supposed to be generated.  In an 
  infra-red AV-OS, the printer needs to draw power from a charged battery 
  to operate.  With a dead battery, the printer dragged the logic power down to a level that 
  triggered a reset (4.75Vdc) and affected the signal lines on the memory card 
  interface.  In the midst of this 
  logic power brown-out, the AV-OS CPU lost the signals that indicates a memory card is installed and halted the system as it is meant to if 
  it believes that a memory card has been 
  removed.  
   
  The infra-red units require 
  a charged PS battery to allow 
  printing to occur.  Expected sealed 
  lead acid battery life is five years.  Most of the failed PS 
  batteries are five to six years old.  Others may have been damaged by 
  discharging to a flat condition.
4.  October 9, 2001 election.  
On opening the 
polls, the AV-OS unit prints "memory card error".  (AV-OS SN 70589)
  The battery was 
  dead on this unit.  When the unit was powered up in the 
  election mode, a zero totals report was 
  supposed to be generated.  In an infra-red AV-OS, the printer 
  needs to draw power from a charged battery to operate.  With a 
  dead battery, the printer dragged the logic power down to a level that 
  triggered a reset (4.75Vdc) and affected the signal lines on the memory card 
  interface.  In the midst of this 
  logic power brown-out, the AV-OS CPU determined there was a memory card failure and 
  printed that message to the paper tape.  
   
  The infra-red units require 
  a charged PS battery to allow 
  printing to occur.  Expected sealed 
  lead acid battery life is five years.  Most of the failed PS 
  batteries are five to six years old.  Others may have been damaged by 
  discharging to a flat condition.
5.  October 9, 2001 election.  
On opening the polls, the 
AV-OS unit reports "I Install memory card".  Note the extra "I" at the 
beginning of the message.  (AV_OS SN 70536)
  The battery was 
  dead on this unit.  When the unit was powered up in the 
  election mode, a zero totals report was 
  supposed to be generated.  In an infra-red AV-OS, the printer 
  needs to draw power from a charged battery to operate.  With a 
  dead battery, the printer dragged the logic power down to a level that 
  triggered a reset (4.75Vdc) and affected the signal lines on the memory card 
  interface.  In the midst of this 
  logic power brown-out, the AV-OS CPU lost the signals that indicate a memory 
  card was inserted and displayed the message "Install memory 
  card".  The extra "I" in the 
  displayed message indicates that the message data was also affected by this 
  logic power brown-out.  
   
  
  The infra-red units require 
  a charged PS battery to allow 
  printing to occur.  Expected sealed 
  lead acid battery life is five years.  Most of the failed PS 
  batteries are five to six years old.  Others may have been damaged by 
  discharging to a flat 
condition.
 
6.  October 9, 2001 
election.  
Memory card is difficult to 
insert.  (AV-OS SN 74439)
  The memory card cover 
  mounted on the CPU board was 
  squeezing the memory card connector.  This was due to interference of the memory card cover with tabs on the 
  sides of the memory card connector that were improperly trimmed from the 
  connector.  
  
   
  After trimming the tabs correctly, the interference 
  was no longer present.  
7.  October 9, 2001 
election.  
Printer prints squashed 
lines.  (AV-OS SN 73888)
  The front edge of the 
  paper roll holder was binding against the paper roll, stalling the paper feed in the printer 
  mechanism.  The paper roll was only 2-1/8" in outside diameter.  
  Inside diameter was 7/16".  The paper roll 
  was in spec.
   
  The paper roll holder needs to be 
  replaced.
8.  October 9, 2001 
election.  
(21) 
Memory card failures.  (10) new 128KB (non-shuttered) and (11) older 
32KB (shuttered).
  Dirty contacts on the 
  memory cards were the cause of the errors.  All of the errors were 
  repeatable.  The material on the contacts appeared to be thin films of 
  adhesive residue.  When the contacts were cleaned, the memory cards 
  passed all tests.  This was found on both 32KB cards (shuttered) and 
  128KB cards (non-shuttered.)  One 32KB card was found to have lost 
  the springs in its shutter mechanism which caused its failure.  Three (3) 
  of the new 128KB cards (non-shuttered) had the cover labels improperly 
  installed.  
  
   
  A method has to be determined for the customer to 
  clean the memory card contacts without compromising the memory card.  The 
  32KB memory card without the shutter springs will be replaced.  The three 
  (3) 128KB memory cards with the improper cover labeling will be 
  replaced.
9.  October 9, 2001 
election.  
Repeatedly started and 
stopped.  (AV-OS 74450)
  The unit printed non-uniform squashed lines of text.  The 
  front edge of the paper roll holder was 
  binding against the paper 
  roll, stalling the paper feed in the 
  printer mechanism.  It was found 
  that the paper roll was an 1/8" larger than the maximum paper roll outside 
  diameter specification of 2-1/4".  
   
  After reducing the diameter of the 
  paper roll, the system printed text in the correct format.