Bob Eager wrote:
> On Thu, 21 Aug 2008 00:46:07 UTC, Warren Block
> wrote:
>
>> Bob Eager wrote:
>>> That doesn't solve it for me, though. It turns out that the card
>>> (chipset) is bad at detecting monitors if connected via a KVM (as this
>>> one is, on my workbench). Plug it in directly and it works.
>>>
>>> Unfortunately the final destination for this system requires a KVM :-(
>> When you say "detecting", do you mean DDC data, or just enabling
>> outputs?
>>
>> Maybe you can just set NoDDC, horizontal and vertical, and PreferredMode
>> for the correct connector.
>
> No, the hardware actually checks for the right impedance at the other
> end of the cable. It never gets as far as DDC - it doesn't believe
> anything is even connected to the VGA socket.
>
> I have the connector selection sorted - the Xorg.0.log tells me that.
> And if I leave out the KVM, it immediately detects the cable as plugged
> in, uses DDC and it all works.
>
> The failure mode is actually flaky - while testing, it managed to see
> the monitor once or twice without me changing anything. Then it would
> fail again, seconds later.
>
> This monitor, and this KVM, have worked on several card/monitor
> combinations on X in the past.
>
> I'm going to try a different KVM, and if that fails I might start
> modifying one...
>

Monitor detect: All seems a bit like the ATI cards we had trouble with
under Vista, XP.
I think all the computer manufacturer did was supply the same ATI card
from a different source.
One was a red pc-board and the other green. Damned if I can remember
which it was that worked now.
- Could mess with some terminating resistors or something I guess.

- Standards compliance is a beast from the dark ages.
Maybe another model ATI would be the simplest...


--

Bill
FreeBSD 7 on AMD-64
"The road less travelled"