View Single Post
  #3  
Old June 23rd 18, 04:20 AM posted to alt.comp.hardware.pc-homebuilt
Paul[_28_]
external usenet poster
 
Posts: 1,467
Default packaging of DIMMs (new componentry)

Paul wrote:
wrote:
I notice that some brands of memory come with the DIMM stuck in a black
platic tray (presumably the conductive anti-static stuff) but others are
totally in transparent plastic.
So is the protection important.


The plastic just has to conduct a little bit.

It can't be a complete insulator.

And you may not be able to measure the resistance properly,
without additional materials. You might have to fit electrodes
rather than make point contact with a regular ohmmeter.

If the material is conductive, charge shouldn't build up on it.

Paul


There are some definitions here.

"Difference between Anti-Static, Dissipative, Conductive, and Insulative"

https://www.gotopac.com/art-esd-resistivity

"Dissipative:

Dissipative materials have a surface resistivity equal to or
greater than 1 x 10^5 Ω/sq but less than 1 x 10^12 Ω/sq

The charges flow to ground more slowly and in a somewhat more
controlled manner than with conductive materials.
"

The purpose of that, is so when you touch the package, you
and the materials (package and DIMM) are brought to the same
electrostatic potential as your body, in a matter of a second or two.

With a "highly conductive" solution, like wrapping an IC
in tin foil, thousands of amps can flow into a signal pin,
from a brief electrostatic discharge. This blows a hole
in the clamp diode junction. And that's to be avoided.

A "softer" conduction characteristic, keeps the "spark" current
flow level much lower in value.

The higher resistance (but still conductive) plastic used
for the DIMM tray, seeks to avoid large discharges into
any signal pins.

ESD damage can go unnoticed. A part can fail some time
after an ESD event. An instantaneous effect on the other hand,
is "latch-up" which drives a totem pole structure into
a conducting path between VCC and GND and "heats it up".
I had a chip once, which went into latchup, and the
manufacturer had designed it so the chip just "stops"
when that happens. Scared the crap out of me :-)
I thought the chip was dead, but it wasn't.

In the old days, a latch-up event could destroy a chip.

The Intel ICH5 had some sort of issue, that looked
similar to the phenomenon, and it used to burn
the bonding wires inside the chip, causing the USB
I/O pads to lose power (permanently). There were quite
a number of reports of that happening, back in the day.
While some ESD events punch holes in front end diodes,
ESD can also cause chip-destroying latchup. While
some ICH5 chips survived, others were ruined such that
a system would no longer POST. The burn mark on the
chip, told you why it could no longer work. Intel never
released public details of the root cause on that one.
There wasn't a big stink about it.

Paul