1 |
Roy Wright posted <4401999F.9080507@×××××.com>, excerpted below, on Sun, |
2 |
26 Feb 2006 06:05:51 -0600: |
3 |
|
4 |
> Peter Humphrey wrote: |
5 |
> |
6 |
>>Which reminds me of a problem I had a few years ago which was solved by |
7 |
>>moving a card to another slot. When clutching at straws ... |
8 |
>> |
9 |
> I had a similar problem last year, turned out the 6800GT wasn't all the |
10 |
> way into the connector. When I would tighten the hold down screw, the |
11 |
> opposite end would raise just a little (~1/16"). Removed and reinstalled |
12 |
> the board three times before I caught it... |
13 |
|
14 |
It's observed that the following discussion probably isn't new to many, |
15 |
but perhaps some haven't thought about all the complex cross-effects in |
16 |
this way before... |
17 |
|
18 |
Actually, if you read the troubleshooting info with quite a number of |
19 |
items, both reseating and trans-slotting are officially recommended |
20 |
troubleshooting methods. |
21 |
|
22 |
Over months or years of on/off cycling and other hot/cold cycles, not only |
23 |
is it possible for cards (and connectors) to work loose, but an oxidation |
24 |
layer may form preventing a decent connection. Again, at the frequencies |
25 |
many computer components operate at, the problem may not be simply lack of |
26 |
a good DC electronic connection, but the fact that the changing |
27 |
conditions takes a connection or set of connections out of capacitance |
28 |
spec, sometimes causing an entirely separate resonance at an unintended |
29 |
frequency. Capacitance of course allows current to jump to parallel but |
30 |
not directly connected circuit traces, and there are often lots of those |
31 |
on a circuit board. The circuit of a specified frequency intended to be |
32 |
completed over one set of physically connected leads may end up entirely |
33 |
transmutated as a parallel circuit develops resonating frequency |
34 |
capacitance, changing the character and details of the transmitted data |
35 |
entirely and wrecking havoc. |
36 |
|
37 |
Note that as memory speeds have increased, mobo makers have had to go to |
38 |
increasing lengths to ensure such interference and cross-echo patterns |
39 |
don't develop. Properly terminated data circuits aren't just a SCSI |
40 |
thing, and many a board design has been delayed or even ultimately junked |
41 |
as it simply couldn't handle the memory and other signal timings, and keep |
42 |
trans-circuit resonance and feedback, to within acceptable tolerance |
43 |
levels. |
44 |
|
45 |
Anyway... something as simple as reseating a card, due to the friction of |
46 |
the connector spring pins against the exposed board traces, will often |
47 |
scratch away that oxidation and return the various circuits to designed |
48 |
operating parameters in both DC resistance and RF capacitance, within and |
49 |
across circuits. That's often all that's needed, and the reason just |
50 |
taking something apart, finding nothing visibly wrong, giving up and |
51 |
putting it back together "without changing anything", actually allows some |
52 |
things to work again. |
53 |
|
54 |
As for trans-slotting, it's less common in modern boards, but it was quite |
55 |
common a few years ago for only 1-2 slots out of 3-5 to be fully |
56 |
"pci-master" capable. Anything using heavy data transfers could often |
57 |
make use of pci-master, offloading work from the CPU under certain |
58 |
conditions, altho the most common use was data transfer between harddrives |
59 |
-- if both were in PCI-master slots, or one was and the other was onboard, |
60 |
the chipset could often handle data transfer between them without direct |
61 |
CPU intervention, once the transfer was initially setup. |
62 |
|
63 |
Additionally, the PCI spec has four interrupt channels, normally staggered |
64 |
so a different one will be in each slot of the first four -- a fifth slot |
65 |
shares placement with the fourth, in most cases, and will work best when |
66 |
purposed to something (like a video card) not normally requiring |
67 |
interrupts. This is because while all four interrupts are available and |
68 |
exposed to all slots, most cards only use one primary interrupt pin, in a |
69 |
standardized location, altho some will allow use of a second for |
70 |
flexibility and many multifunction cards use two or more if all functions |
71 |
are enabled. Most new mobos come with a diagram of their PCI interrupt |
72 |
layout per slot, and specifically advise trying a different slot if a card |
73 |
doesn't work as expected, thereby causing it to utilize a different one of |
74 |
the four PCI interrupts. |
75 |
|
76 |
Finally, there are intercard proximity factors as well. Of course, |
77 |
there's pure physical space requirements -- some cards simply don't fit |
78 |
next to each other or for sound cards and the like, plus PATA/SATA drives |
79 |
and even interface cards with detached plugs, attached cabling simply |
80 |
won't reach to all possible slots. Heat and ventilation are another major |
81 |
factor. However, the same cross-circuit inductive and capacitive effects |
82 |
outlined above can play a factor here, as well. This can be as simple and |
83 |
easily recognized as a buzz on your sound card when it's too close to |
84 |
another specific card, or as difficult and unintuitive to recognize as |
85 |
memory issues or disk storage errors. One particular thing that we've not |
86 |
had to worry about for a few years (due to dedicated AGP slots) that we'll |
87 |
have to start worrying about again as PCI-Express becomes more common is |
88 |
video card placement, as those have all sorts of RF things going on, with |
89 |
both high data rate memory access, and high frequency video signalling. |
90 |
Get those signals cross-interfering with something like memory/CPU |
91 |
channels or CPU/CPU interconnects (AMD's hypertransport, and Via has a |
92 |
similar technology for northbridge/southbridge interconnects, among |
93 |
others), or an add-on SATA/PATA/SCSI interface card, and there could be |
94 |
/major/ data integrity issues to worry about! |
95 |
|
96 |
It should be quite obvious by now that mobo designers certainly don't have |
97 |
it easy, these days! Sure, they deal with lower frequencies than CPUs and |
98 |
graphics chips deal with, and don't have the intricate logic and heat |
99 |
dissipation layout issues that those guys do. However, the frequencies |
100 |
they do deal with are quite high enough to cause **MAJOR** cross-circuit |
101 |
interference design issues, particularly over the trace length of |
102 |
centimetres they deal with, as opposed to the nanometre/micrometre scales |
103 |
a chip designer deals with. As I said, it's not unusual at all for a mobo |
104 |
design to simply not be able to handle the necessary timing demands placed |
105 |
upon it, and either be limited to the low end of an offered range (those |
106 |
mobos that are certified for PC2700 memory or less, because they simply |
107 |
can't handle PC3200), or end up entirely rejected due to stability issues. |
108 |
|
109 |
Under this sort of tight timing requirements, it's actually a wonder we |
110 |
don't have more cards that simply won't work right in certain boards, |
111 |
while they work just fine in others. That alone is tribute in itself to |
112 |
the wonders of human engineering genius, both in design, and in the |
113 |
ability to actually implement said design in an assembly-line environment |
114 |
to within the required tolerances, without costing us several thousand |
115 |
dollars for the mother-board alone! |
116 |
|
117 |
-- |
118 |
Duncan - List replies preferred. No HTML msgs. |
119 |
"Every nonfree program has a lord, a master -- |
120 |
and if you use the program, he is your master." Richard Stallman in |
121 |
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html |
122 |
|
123 |
|
124 |
-- |
125 |
gentoo-amd64@g.o mailing list |