Help! RGB Color weirdness - Apple IIGS/Pacman/Pat9000

kiphartzell

New member
Joined
Dec 3, 2006
Messages
201
Reaction score
1
Location
Raleigh, North Carolina
Help! RGB Color weirdness - Apple IIGS/Pacman/Pat9000

All- I could use some help understanding this RGB signal distortion issue.

Background: I have a working Pacman board running through my PAT9000 with a Apple IIGS (A2M6014) monitor attached as the Raster display. I am seeing distortion/weirdness with the displayed colors using this testbench setup. I traced the problem to a specific circuit on the Video Interface Card within the PAT9000 and would like some help with what the intent of this circuit.

When I drive the Apple monitor directly from the Pacman board (not through the PAT9000), the colors are crisp and correct (as seen in the first to pictures attached). Also, when I attach a Arcadeshop Programmable Jamma card in this setup using the PAT9000, the colors are correct so I know this general setup through the PAT9000 (and the Video Interface card) does work for some boards.

This problem seems to be something with how Pacman gnerates/drives RGB and how this mystery circuit (diagram attached) in the PAT9000 affects these RGB signals. The Circuit in question looks like it might be a simple 2 stage emitter follower buffer circuit for re-driving the RGB signals, but I also see an increased amplitude on these output signals when scoped. I am rusty with my analog design knowledge and I haven't done much with RGB signal processing before so any help from more experienced people would be appreciated. Thanks.

One other thing, when I attach a standard 19" arcade monitor instead of the Apple monitor, colors appeared to be more accurate and generally correct. So.... it seems to be a combination of what the RGB signals are like coming from the Pacman board and the input tolerance of the RGB signals on the Apple Monitor (once conditioned by the PAT9000 mystery circuit).

First two pictures are the good colors directly from the Pacman card, the last two pictures are what I see when the RGB signals are passed through the PAT9000 Video Interface card.
 

Attachments

  • RGB Pat9000 circuit corruption.jpg
    RGB Pat9000 circuit corruption.jpg
    51.6 KB · Views: 23
  • GoodColor.jpg
    GoodColor.jpg
    67.4 KB · Views: 32
  • GoodColor2.jpg
    GoodColor2.jpg
    62 KB · Views: 27
  • DistortedColor2.jpg
    DistortedColor2.jpg
    80.3 KB · Views: 35
  • DistortedColor3.jpg
    DistortedColor3.jpg
    61.2 KB · Views: 29
It's really hard for me to read that schematic, since the picture is so small, but the image almost looks like what you get if you hook something up without terminating resistors... Try adding a 75 ohm (or thereabouts) resistor to ground on the RGB lines coming out of the buffer.

-Ian
 
Sorry about the schematic. By the time I re-sized it for meeting the upload requirements, it was pretty crappy to view. It is from the PAT9000 manual (sheet 147) and the copy I have is not very good. I reattached the same circuit, but blown up a little more. The circuits consist of 2N3906(PNP),2N3904(NPN), 1K and 470 ohm resistors.

I can try some ~75ohm resistors as you suggest. Thanks.

BTW, I already tried putting some caps on the RGB signals based on a document I found for hooking up a Apple monitor for arcade testing. See note 3 on the pdf attached. One questions on trying the caps.... I ran each RGB signal "lead through 220uf caps", but I am wondering if the caps should be across the RGB signals to Vid GND instead of sending the signals just through the caps.
 

Attachments

  • RGB Pat9000 blowup circuit.jpg
    RGB Pat9000 blowup circuit.jpg
    41.5 KB · Views: 13
  • Apple_monitor_notes.pdf
    27.5 KB · Views: 3
Last edited:
No. The capacitors are there for DC blocking, to remove any DC offset the signal may have. I've not personally found them necessary for arcade boards, but they are definitely required on the Super Nintendo. The positive lead of the capacitor goes toward the game, negative to the monitor.

-Ian
 
Ian, Thanks for clearing up the capacitor usage and how to properly install them. I tried both directions and didn't see really any effect either way.

So, RGB signals are AC in nature. I drove a White grid from the Pacman and measure ~.200 volts using a DMM when the board was directly attached to the Apple Monitor. I measured a high voltage ~ .600 went it went thorugh the PAT9000 circuit, but I don't recall if I was measuring AC or DC. BTW, this did seem to indicate that the mystery buffer circuit was also amplifying the signals, not just buffering or acting as a protection circuit?!?
 
Okay Ian, I tried some ~75 ohm termination resistors on the output of the "buffers". No effect noticed and in affect these outputs already have 470 ohm resistors from the R/G/B output signals to GND per the circuit diagram provided.

I am back to wondering what this interal PAT9000 circuit is really trying to accomplish?!? It operates okay with the ArcadeShop Programmable Jamma card attached, but not the Pacman board. BTW, these are the only two raster games that I currently have PAT9000 plus wired up so my observations are limited to these two games.
 
Somebody must have an idea with what the intention of this circuit is? For whatever reason, it modifies the RGB signals so they are now distorting the colors on my Apple testbench monitor. Can this circuit be bypassed on the PAT9000 card and just drive out the same RGB signals just as the came in?
attachment.php
 

Attachments

  • RGB Pat9000 blowup circuit.jpg
    RGB Pat9000 blowup circuit.jpg
    41.5 KB · Views: 9
Back
Top Bottom