PCBUILD Archives

Personal Computer Hardware discussion List

PCBUILD@LISTSERV.ICORS.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Dave Gillett <[log in to unmask]>
Reply To:
PCBUILD - Personal Computer Hardware discussion List <[log in to unmask]>
Date:
Mon, 12 Jul 1999 11:21:22 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (154 lines)
Someone wrote to me privately:

> Dear Master,

  While I appreciate the compliment, I'm afraid that I know answers to only some
of your questions -- and I have only now returned from a family vacation.  I
believe that if you were to send your questions to the PCBUILD mailing list,
you would receive a more complete set of answers, and sooner -- and others
who have the same questions would get a chance to see the answers.

> I have some probs with these terms:
>
> EPP/ECP capable ports,

  These are standards for additional functionality added to a PC "parallel"
port, primarily to support more sophisticated external devices such as drives
and scanners.  The original port design was intended only for printers, which
generally work at relatively low speed and almost entirely for output; these
other devices provide input as well and can take advantage of higher speeds.

> IR ports,

  Infrared (IR) provides a short-range wireless connection, like that between
a remote control and a TV, VCR, or stereo system.  Many modern PCs provide
for IR hardware to be addressed as if it was a serial port, and used to
connect a keyborad, printer, or for connection between two PCs.

> UART,

  The first Univeral Asynchronous Receiver/Transmitter chips became common in
the late 1970s; they provided virtually all of the logic for an asynchronous
(serial) port on a single chip.  [In most cases, it was necessary to add two
very small devices to convert between internal TTL and external RS-232
voltages.]  Later UART models could handle higher speeds, and included FIFO
buffers to allow for system latency.
  On most modern PCs, a couple of UART chips are emulated by circuitry
included in the motherboard chipset, and it has become less common to
encounter them as separate devices.  Most software written to communicate
directly with UARTs will work just fine with these emulations.

> PBSRAM, BEDO,

  These are a couple of RAM technologies I've never worked with.  Maybe
someone else on the list can provide some information.

> DRAM,

  Dynamic RAM is composed of bit-storage cells which are relatively simple
and cheap, but when a bit signal is stored in a cell, it gradually decays.  A
system design using dynamic RAM will include a "background" process that
cycles through the memory space, retrieving and re-writing each datum to
"refresh" the signal retained by the memory so that it is still available
when needed.
  Because dynamic RAM technologies tend to be inexpensive, they are the most
popular choice for main system RAM.

> SRAM

  Static RAM, on the other hand, retains signals at least as long as power is
available, without requiring a "refresh" process.  The memory elements are
more expensive, but also tend to respond more quickly.  SRAM is routinely
used for caching, where a small amount of expensive but fast RAM can make a
big difference to system performance.

> and SDRAM?

  Synchronous DRAM is the most common current technology for main RAM.  It
uses various techniques to try to mitigate the normal slow response of DRAM
in order to try to keep up with ever-faster CPUs without making system prices
prohibitive.

> What does "synchronous 3.3V pipeline burst SRAM" mean?

  This is a cache memory (SRAM); the "pipeline burst" is a method to improve
performance, based on the realization that whenever the CPU retrieves a value
from memory location N, odds are very high that it will want the value from
location N+1 in just a moment -- a cache implementation that anticipates this
will be right often enough to make a difference.

> 32x32/64x32 PBSRAM?

  The number after the 'x' is probably the "width" of the memory component;
the most common widths of individual RAM chips are 1, 4 or 8 bits, but these
may be combined on SIMMs, DIMMs or other modules to build wider (and longer)
memory arrays.  32 bits is four bytes, so 'x32' components are 4 bytes wide
by however long is indicated; depending on the application, the length will
often be in K (2^10, 1024) or M (2^20).

> The speeds (ns) for those kind?

  Individual memory chips often carry a suffix to the part number indicating
speed.

> Can we determine the speed (xxns) from the frequence(MHz)?

  Only roughly.  While in theory a 10ns RAM part can function in a 100 MHz
system, this is right on the edge of its rated capability -- parts certified
for operation at 100 MHz will be rated at 7 or 8 ns, leaving a margin that
greatly increases reliability.

> Also, the unit of RPM, ms, M/sec use in HDD? What is the concern between them?

  RPM: rotations per minute.  This measures how fast the disk platters spin.
Since the read-write heads are at a single place, a faster rotational speed is
likely to reduce the time from when a particular sector is requested until it
passes the heads and can be read.
  ms: milliseconds.  This measures typical time to access any sector on the
disk surface.  It's going to reflect both RPM (above) and travel time
positioning the heads to the correct track.  A lower number indicates a drive
that provides faster responses to random requests.
  M/sec: Megabytes per second.  This is a measure of transfer speed -- having
located the correct sector, how quickly can a substantial volume of
information be read or written.  High RPMs can help with this -- or high
numbers of sectors per track.  This is also often cited as a characteristic
of the drive interface (SCSI, IDE, etc) rather than of the drive itself.

> And, BTW, I dont know the different between CASEs for x86, Pentium, P2, or
> AT, ATX, I mean the power units(voltage different)? Can I change the AT
> case of a VT82C580VPX mobon with P233MMX to ATX?

  Not *generally*.  AT and ATX power supplies use different connectors to
deliver power to the motherboard.  Some recent motherboards, including some
using that VPX chipset, provide both sorts of connector.
  But where an AT case provides a round hole for a keyboard connector, an ATX
case will provide a panel with cutouts for various ports built into the
motherboard (including a smaller PS/2 keyboard connector).  And so even if
you can use an ATX power supply with your AT board, you probably can't use an
ATX case.

> There're two level of voltage, 3.3V and 5V for RAM, I only know DIMM SDRAM
> use 3.3 and EDO/FPM DIMM use 5V, don't know anything about others!

  The power supply is going to provide power to the motherboard at 12v and 5v.
If components on the motherboard need any other voltage, the motherboard will
include voltage regulation to provide it.

> When I used a case of a 486(sure no prob with it) with a VXpro, it could
> not boot up, even not a beep from the spk. Then I tried a generic AT case,
> it ran well, why? Is it a voltage problem?

  It's not likely to be a voltage problem.  More likely causes would include:
(a) Your VxPro system needed more power than the 486 did, and than the 486's
power supply was rated to provide, and/or (b) over time, the 486 power supply
is likely to have degraded, and might no longer be delivering its full rated
power.  It might still have been enough for the 486 without being enough for
the VxPro.


David G

         PCBUILD maintains hundreds of useful files for download
                     visit our download web page at:
                     http://nospin.com/pc/files.html

ATOM RSS1 RSS2