Re: Apple //e RGB Card to CGA pinout?

Knut Roll-Lund wrote:
Bryan Parkoff wrote:
"Knut Roll-Lund" wrote:
Knut Roll-Lund wrote:
Bill Garber wrote:
"Knut Roll-Lund" wrote:
Mark McDougall wrote:
Bill Garber wrote:
IIRC, CGA monitors accept Composite Sync on the Horizontal Sync pin.
I'd have to research it again to be sure. Been a long time.

Yeah, that's exactly what I was thinking when I first read the post... so ditto from me.

I will try it... the CGA monitor too then and report back my result. The Mitsubishi is like that, it can use composite sync on the same terminal which is Horizontal when using separated syncs.

I read in the Mitsubishi manual that it should be possible to have the syncs composite on the green input too, but that doesn't work because then the normal monocrome video output should work on the green input. It just swims no sync. 

Well, you can't just attach the sync signal to the green,
it has to be incorporated digitally.


It wasn't the sync, it was mono composite video.

And, yes I know, and probably not just digitally, it is also a special level. I'm not sure but I though a normal mono composite video would be similar to what that green+sync input expects. It might of course be different so that is why it didn't work. Applying this mono composite signal to the blue or red inputs would produce exactly the same result.

I have the RGB card now so I'll soon be back with some more test results (it at least works as a normal 80columns 128k card...)


I first made a straigt converter from DB-9 to DB-15 for the CGA monitor (IBM 5153) I put the composite sync signal to the horizontal input. It was no success the picture came though clear but wouldn't stop rolling. I think vertical sync is needed unfortunately.

I then made a quick adapter for the mitsubishi monitor without the resistors to display "8 colors". It wasn't a success either. The picture was stable but the screen was garbeled. It seemed to me that horizotal sync wasn't working, that the lines were too long or too short. making each scanline not in sync with the next. Either there is a problem with the levels or the monitor might be broken (I will open it and see if there is anything to adjust.

I am concerned that CGA is like analog RGB, but not digital RGB. Michael M. claims that most CGA monitors connect to ISA card for PC should be digital RGB. If you already connected CGA monitor to PC and it is working fine with no problem, then the configuration from Apple //e RGB card to CGA monitor is wrong according to my theory.
I encourage to compare the schematic from Apple //c's video port to Apple //e RGB card so it might give you an idea how to build an adapter. An adapter should contain one or more chips to manipulate H sync and V sync. Please comment what you have made some attempts when progress is going on. Thanks...

Well, you can see I have posted a lot of messages in this thread now. Maybe all of these tidbits aren't of interest to all so you might email me remove 'nogarbage.' from the email address to reply directly to me. It might be better to present the results to the group. We are probably trying to do something similar.

I traced the composite sync signal back in the 2e. On the RGB card it goes through two inverters (20ns delay...) in an LS04 and comes from the mainboard. Here it is generated by a 40 pin ic next to the aux connector (p.a.l. version European motherboard) so I think if one needs separate vsync and hsync they have to be made by a circuit like the LM1881.

The CGA monitor I have is an IBM 5153 and it is definitely a digital monitor. The colors come out just nicely, but I can't synchronize, the picture rolls. With the LM1881 I could get it to hold the picture only if I enlarged the vertical so that 1/3 of the screen was not visible and there was a noticeable gap between the lines. The LM1881 takes 5 to 12V supply so I tried supplying both from the apple but there was no significant difference.

I will proceed later today to investigate with an oscilloscope and compare the signals from the apple with the signals from the cga card. It seems that the cga monitor is a good choice if I only can get it to sync.

The other monitor, the Mitsubishi is an analog monitor but not a multisync... it should work with an adaptation like the one suggested by Marco Orlandi. It should according to the marking on the back be able to use a composite sync. I have made a setup for 8 colors though intending to use Marco Orlandi's setup to generate 16 colors. But it won't sync neither with the sync signal from the RGB card, nor with the regenerated one of the LM1881.

So I hope the oscilloscope helps me to understand this better. 

Again replying to my own post, sorry...

I have now looked at the signals with an oscilloscope and can find nothing wrong. I think my problem is that the Apple's output is interlaced, so the signal is unsuitable for the IBM 5153. I get the colors and horizontal sync ok but the picture rolls. The IBM 5153 even makes a noise back into the LM1881.

The Mitsubishi HF2400 now syncs to the vertical, with the cleaned composite sync generated from the LM1881 but horizontally it is completely wrong. It seems to do 2 lines after each other per line and loosing one before the next. This happens whatever I do. I have tried to adjust a lot but it seems that the line from the apple is just too short. I think this monitor was used in a 1024 by 768 setup so it might be that the line generated from the Apple is just too short or something to that effect. BTW I could get this monitor to do sync on green there was a switch accessible with the cover removed.

So, for now, I give up. I will see if I can get hold of a suitable monitor that was meant for the Apple. Analog or digital. If I get hold of an analog Applecolor GS type monitor I will use Marco Orlandi's setup for the RGB and the LM1881 to boost the sync.

It is a pity that CGA monitor wouldn't work, the colors were nice, 80 columns text were very readable, I just needed it to not to roll.

It is possible that I can study the RGB card and make some schematics, how digital RGB is generated from the aux port. I'm interested to see if it had been fairly easy to make this extension (piggyback etc.).

Knut
.