Device Capacitance balance

Paul D. Aloisi 603-429-8687 aloisi@unitrode.com ALOISI at unitrode.com
Wed Sep 3 05:43:02 PDT 1997


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* "Paul D. Aloisi 603-429-8687 aloisi at unitrode.com" <ALOISI at unitrode.com>
*

There has been a lot of off line an on line discussion about the specification.
0.5 pF in the standard means manufacturers of adapters, disk drives and other
SCSI devices need to guarantee 0.5 pF difference between pins that are
on two sides of a PC board. 

What is around the PC board has a drastic effect on the capacitance. The
components are only part of the issue. Measuring a component in a test socket
gives one reading, mounting it on a multilayer board gives another reading.
Etch routing under the device changes the readings. The module in the next
slot even has effects for the adapters. 

A disk drive controller board measured off the drive gives one reading, 
mounting it on the drive adds pFs to the reading on one side of the module.
Mounting the drive in a box with metal underneath the drive adds pFs to the
other side of the module.

Bill Ham's meaning of capacitance balance is the device measured in the 
system. If you a designing an adapter or disk drive, you usually have 
very little control on where the device is mounted and what is mounted
next to it. The only way to guarantee the capacitance stays constant 
between the pairs that are usually on different sides of the module is 
to enclose the device. 

The 2 pF capacitance between pairs came from models, the 0.5 pF between
pairs was a choosen number without modeling to set the requirement. The 
discussion at the time was that it would stay 0.5 pF unless proven that
it could not be met. George is completing the edits to SPI-2, we need to
make sure all the specifications are realistic. 0.5 pF with an undefined
environment over 3-4 components; connector, PCB, Controller and possiblay
the terminator is too tight. A disk drive tested on a plastic hots swap
guide versus direct mount in a metal rack would have two different reading,
pFs appart, not 0.5 pF. The mounting tolerance of the screws can change 
the readings more than 0.5 pF. Some systems the drive does not have anything
under it except the next drive, with no guaranteed spacing to the next
drive.

Bill would have to spec his system, is the cable folded beside the module
between it and the next module and the modules on both sides of the 
SCSI Controller. Right angle connectors, the pins furtherest away from the
module will change capacitance with the proximity of the cables and modules.
Strattle mount connectors have etch and metal different sides of the 
module, what is mounted on each side is important to balance. Cables 
between modules have an effect. 

Gene, silicon is only a small part of the total device. The drive to reduce
capacitance 10 pF single ended measurement for a device pin, 5 pF is
normally the etch and connector; 5 pF for the controller and 
possiblay a terminator. 25 pF devices in the past, 20 pF was for components
which silicon dominated the measurement. 5 pF for the components, leadframe
and bonding wire capacitance is a major factor. Ground planes, etch routing,
and the metal housing over the circuit effects the device capacitance
more than the silicon.

The level of the readings requires a definition of the test conditions and
realistic specifications.

Thank you,
Paul Aloisi
Unitrode 
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com




More information about the T10 mailing list