Return to BSD News archive
Xref: sserve comp.sys.ibm.pc.hardware:33902 comp.unix.bsd:6825
Path: sserve!manuel.anu.edu.au!munnari.oz.au!uunet!mcsun!chsun!Germany.EU.net!unidui!du9ds3!veit
From: veit@du9ds3.uni-duisburg.de (Holger Veit)
Newsgroups: comp.sys.ibm.pc.hardware,comp.unix.bsd
Subject: Re: Question on Diamond Clock Synthesizer
Date: 20 Oct 92 12:06:13 GMT
Organization: Uni-Duisburg FB9 Datenverarbeitung
Lines: 78
Message-ID: <veit.719582773@du9ds3>
References: <1b7tmgINNi06@agate.berkeley.edu> <1992Oct19.082420.16353@Informatik.TU-Muenchen.DE> <1992Oct19.151409.24581@osf.org>
Reply-To: veit@du9ds3.uni-duisburg.de
NNTP-Posting-Host: du9ds3.uni-duisburg.de
In <1992Oct19.151409.24581@osf.org> kenny@osf.org (Kenneth Crudup) writes:
>In article <1992Oct19.082420.16353@Informatik.TU-Muenchen.DE>
>roell@informatik.tu-muenchen.de (Thomas Roell) writes:
[...Thomas Roell, formerly THE main developer of X11 on 386 platform
on the possibility of destroying diamond hardware by improper
PLL programming...]
>ONCE MORE- explain to me exactly how programming a divider value into a
>chip WILL BLOW IT!!!
>If you run some chips (the i486 comes to mind) at higher speeds, they will
>dissipate more power, and could (BIG emphasis on could) overheat to the
>point where the chip becomes unreliable, but NOT fail. No chips on my
>SS24x run anywhere near warm, and definately less than the heat generated
>by my AMI 386-40. In order to get any type of useful divider resolution,
>the clock freq in has to be quite high. If you "program the PLL wrong"
>so that you generate one of these "chip-blowing high frequencies" YOUR
>MONITOR WILL SHOW IT (or I guess you'll tell me, despite my kicking holes
>in that theory, that it will blow your monitor too?)
>Either give me a full technical explanation, or quit talking shit! How much
>*is* Diamond paying you?
> -Kenny
>--
>Kenneth R. Crudup, Contractor, OSF DCE QA
>OSF, 11 Cambridge Center, Cambridge, MA 02142 +1 617 621 7306
>kenny@osf.osf.org OSF has nothing to do with this post.
> Religion: The longest-running gag ever played on Mankind.
Shut up! Before you post such an unqualified attack, you yourself should
show any basic qualification in either CS (in particular low level programming
of video hardware) or EE (chip design). Thomas at least has demonstrated his
competence in the first area.
If don't accept a well-meant warning of someone who is much more inside the
job of video programming than you, then DO WHAT YOU THINK IS RIGHT, but
don't flame anyone if something bad is happening then. Read any basic
book on chip design to see that overheating is not the only reason why
some hardware can become inoperable, and testing with a wet finger on some
chip whether it is too hot is certainly not a professional way of decision.
Why do you think are there chip environmental specifications? Just to
allow chip vendors sell a 40 MHz version for more bucks than a 25?
It is well known that operating a chip outside the validated and guaranteed
specifications *will* reduce its lifetime considerably. Can you imagine that
internally some chips use the clock oscillator to push up internal voltages
(any switch power supply works this way)? Ever found out that the parameters
of switched capacitors depend on their frequency? Also heard something about
aging effects of MOS semiconductors that are a function of not only the
temperature but also directly of frequency? There are many more irreversable
effects there than just overheating. Have you ever programmed a
fuse-programable chip, like a PAL, or PROM? You may do this bit by bit manually
by toggling some switches at the inputs in correct order, and the chip
does not heat up at all. But the result is quite obvious.
To be accurate: non of the above may be relevant for the special video
board in question, and you might really see an effect of misprogramming
faster on your monitor than by a failing chip.
But as long as Diamond is refusing to publish details on their programming,
I would not spend any thought on buying any of their hardware. It must be
assumed that there are not only "trade secrets", but as already mentioned,
different versions, which need special handling. This would demonstrate that
they have trouble to compile a working design at all. I would not trust such
hardware: there may any kind of dirty hack (of which I have seen may in
chip design) in it, which indeed may cause destruction of hardware by
improper handling.
Holger
--
| | / Dr. Holger Veit | INTERNET: veit@du9ds3.uni-duisburg.de
|__| / University of Duisburg | "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
| | / Dept. of Electr. Eng. | Sorry, the above really good fortune has
| |/ Inst. f. Dataprocessing | been CENSORED because of obscenity"