Speckles on original VGA Videl output ??

[alanh]

While the SuperVidel output looks o.k., anyone see green speckles on the original Videl output now ?

[instream]

Does this happen with the motherboard-clocks-cable soldered? When we developed it I remember soldering the ground wire for the vsync signal to what I thought was ground near the Videl connector. But the result was some kind of disturbance on the Videl output blue (not green IIRC). I removed the soldering of that ground wire eventually since the vsync signal is so low in frequency information anyway. So you should not solder the ground wire that belongs to the vsync signal. Just let it travel along the vsync signal.

[alanh]

Just tried with and without the soldered clocks cable attached to the SV and it's the same regardless.

[instream]

OK. But you haven't soldered the ground wire that belongs to the vsync onto the motherboard , right? I think its presence induces interference into the analog Videl output, wether you have connected the other end to the SV or not.

[alanh]

Yes, I did, because that's what the manual said. So I should unsolder it ?

[instream]

Nope, it says under the picture of the vsync wire that the ground wire should NOT be soldered, and not touch anything.

[alanh]

Nope, page 8 of the manual clearly shows the the resistor wire is soldered to the clock, and another cable soldered to ground.

Or are you talking about L29 ?

[alanh]

I see you probably are talking about L29, then yes. That's all I have soldered, exactly what the manual states. Nothing more, nothing less.

[instream]

Yes this has been about the Vsync signal at L29 the whole time. The clock signal is not the problem, and its ground is supposed to be soldered. But the Vsync ground wire must not be soldered, like the last picture in the manual says.

[alanh]

Right, so I am as the manual states, but see speckles on the original Videl output.

[alanh]

Oh, and these speckles are not video glitches, these look like video RAM problems as they are fixed and not just green, but red and blue.

[instream]

Ok, maybe you can try to desolder one wire at a time until it disappears.

[alanh]

No cable connected to the SV, and then started to desolder...

Desoldered L29, powered up, still speckles.

Desoldered ground from clock, powered up, still speckles.

Desoldered clock line, powered up, still speckles.

Cable is completely removed, but no improvement.

[alanh]

With these tests, I don't have NVDI or anything installed. This is a straight boot to TOS with nothing loaded.

I'd expected the original Videl output to be unchanged.

[instream]

Can you take a photo of it?

[alanh]

Sure, give me a few minutes.

[alanh]

Here goes....

speckles.jpg

[alanh]

The interesting thing is if I move a new window with a white background, then the previous area will repaint mostly o.k.

I get more speckles as I move windows around, but as I say, repaints mostly repaint fine.

[instream]

These speckles look static once they have appeared until they are overwritten, right? And full pixels. Then we can rule out interference in the analog Videl output.
And they only happen in 060 mode?
What is your CT60 clock freq?

[alanh]

These speckles look static once they have appeared until they are overwritten, right?

Correct.

And they only happen in 060 mode?

Correct.

What is your CT60 clock freq?

I'd already tweaked that multiple times, but I've gone from my original 90MHz, down to 85MHz, and now at 78MHz, and it's always the same.

[alanh]

One thing to note is because I'm using the CTPCI underneath the SV, I removed the clock filter jumper.

I'm wondering if I should remove CTPCI and replace the jumper on the SV.

[alanh]

I'll try some experiments.....

[alanh]

O.k. Found the cause.

I was using 1.04 CT60TOS and found that I'd left the switch to bus boost on as I have an original CT60 with the F030 boost. Then flicked it off and the machine wouldn't boot without it. Just kept resetting.

I read in the manual that 1.04 had some problems, so I tried 1.03c and hey presto. We're back in business

Unfortunately 1.03c disables my CTPCI

So tried 2.01, and it worked for the first two resets, and I'm back to a black screen at bootup. I don't think this is related to the SV though, I've always had this problem with the 2.0x series.

So currently, I'm stuck with 1.03c with the SV and no CTPCI

[instream]

Good that you found the cause at least
I used to run 2.01 until it killed itself. Then I also thought the hardware was at fault until I tried 1.03c which made the computer boot again.
Therefore we recommend 1.03c over 2.01.
1.04 has always had strange behavior so we recommend against it.