Quantcast

Results 1 to 9 of 9

Thread: Mega Drive 2 60hz Colour Fix - crystal oscillator RGB noise?

  1. #1
    Death Adder's minion
    Join Date
    Jan 2009
    Posts
    12
    Rep Power
    0

    Default Mega Drive 2 60hz Colour Fix - crystal oscillator RGB noise?

    Hi, I've just got questions for people with experience doing the 60hz colour fix mod for Mega Drive (particularly MD2), or people just with great knowledge of Mega Drive hardware in general.

    I'm looking at connecting a four-pin 4.43mhz crystal oscillator to Pin 6 of the Fujitsu MB3514 video encoder on my VA1.8 Mega Drive 2 and cutting the trace to system clock (which, when set to 60hz, generates a subcarrier frequency neither PAL nor NTSC) - to ensure that when the system is set to 60hz the picture will always be PAL in composite, that is, PAL60 when set to 60hz.

    However, when doing research on this modification, I've seen a few people comment (including in this blog post: http://www.retro-otaku.com/2010/07/p...e-and-s-video/) that while connecting the crystal oscillator to the video encoder may result in good composite output, it introduces noise into RGB picture, so it is better to put in a switch between 50hz/system clock and 60hz/crystal oscillator and just accept bad RGB in 60hz. I'm not at all experienced in soldering, so I'd like a good picture in my mind of what to do in advance...but preferably I'd like to avoid having a switch between crystal oscillator and system clock and just have the crystal oscillator permanently connected. I've done an equivalent crystal oscillator mod on PS1 with a CXA1645 and everything, including RGB, seemed to work fine...

    Here are my questions...

    1) What is the reason for RGB noise when the crystal oscillator is connected to the video encoder? For example, is it because of unclean power? If so, why does it only affect RGB and not composite?

    2) Virtually everybody that reported this issue was working on an original Mega Drive with a CXA1145 video encoder. Is it likely to affect work on a Mega Drive 2 with MB3514 video encoder?

    3) I have only seen one colour fix guide suggest that Pin 7 (NTSC/PAL switch) on the MB3514 should be wired to GND in order to permanently "lock in" PAL mode. Is this truly a necessary step? Could omission of this step introduce RGB noise?

    Thank you
    Last edited by bioform; 02-22-2017 at 01:57 PM.

  2. #2
    Wildside Expert
    Join Date
    Aug 2013
    Location
    Portugal
    Posts
    199
    Rep Power
    8

    Default

    Though I'm not familiar with the model 2 on my model 1 I have modded it with not one but two crystals, one for PAL 4.43 and another for NTSC 3.58 so I can choose what colour encoding I want. My Megadrive has 4 switches, one to choose between one crystal or another, another to choose PAL or NTSC encoding another is the region mod and last and not least, and the one with most interest to you is a chroma killer, the switch can either turn on the chroma carrier or pull the carrier to ground.

    In one position you have normal composite operation with all the dot crawl and RGB full of jailbars, on the other position you "kill" the carrier, you get a black and white image in composite and super clean RGB with no jailbars at all, I can make a video to prove it.

  3. #3
    Death Adder's minion
    Join Date
    Jan 2009
    Posts
    12
    Rep Power
    0

    Default

    Quote Originally Posted by t0nito View Post
    In one position you have normal composite operation with all the dot crawl and RGB full of jailbars, on the other position you "kill" the carrier, you get a black and white image in composite and super clean RGB with no jailbars at all, I can make a video to prove it.
    That's quite an interesting setup you have there. It suggests that (on the Model 1 at least) the video encoder's RGB signal must get some kind of interference from any connected subcarrier frequency. I'm assuming that using the PAL crystal makes RGB quality even worse? (perhaps because the frequency is higher?)

    Your encoder is a CXA1145 right? Still unsure if it's some kind of design flaw with this older Sony encoder or if the issue goes deeper within Mega Drive hardware...

  4. #4
    Wildside Expert
    Join Date
    Aug 2013
    Location
    Portugal
    Posts
    199
    Rep Power
    8

    Default

    Quote Originally Posted by bioform View Post
    That's quite an interesting setup you have there. It suggests that (on the Model 1 at least) the video encoder's RGB signal must get some kind of interference from any connected subcarrier frequency. I'm assuming that using the PAL crystal makes RGB quality even worse? (perhaps because the frequency is higher?)

    Your encoder is a CXA1145 right? Still unsure if it's some kind of design flaw with this older Sony encoder or if the issue goes deeper within Mega Drive hardware...
    Using a 4.43 or 3.58 crystal doesn't make it better or worse, just shows you a different pattern. Yes it's the CXA1145.

  5. #5
    Death Adder's minion
    Join Date
    Jan 2009
    Posts
    12
    Rep Power
    0

    Default

    Quote Originally Posted by t0nito View Post
    Using a 4.43 or 3.58 crystal doesn't make it better or worse, just shows you a different pattern. Yes it's the CXA1145.
    Thanks t0nito, your replies are helping narrow things down a bit. Do you know whether you had any similar RGB distortion patterns with the original system clock subcarrier or was it distortion-free like the zero crystal subcarrier RGB picture?

  6. #6
    Hedgehog-in-Training Hedgehog-in-TrainingSports Talker
    Join Date
    Sep 2015
    Posts
    40
    Rep Power
    0

    Default

    The signal you feed into the RGB encoder needs to be conditioned well. Amplitude, signal shape and absence of all noise and oscillation are important.
    I don't think any of us hobbyists can do a perfect job on all these parameters, but if you get it all right, then the RGB output won't be affected by your clock (at least not any more than the original circuit).

  7. #7
    Wildside Expert
    Join Date
    Aug 2013
    Location
    Portugal
    Posts
    199
    Rep Power
    8

    Default

    Quote Originally Posted by bioform View Post
    Thanks t0nito, your replies are helping narrow things down a bit. Do you know whether you had any similar RGB distortion patterns with the original system clock subcarrier or was it distortion-free like the zero crystal subcarrier RGB picture?

    Before the mod mine was a PAL unit, I used to have vertical stripes on the RGB output with the original clock but I only did the chroma killer mod after the crystal mod but I believe that pulling the chroma to ground is enough to eliminate or atleast greatly reduce the interference.

    To tell you the truth on my unit, if I connect it to a CRT I can't see any banding at all whether the chroma is on or off, on my HDTV's you can defenately see the difference.

    The mod is easy to do you use a 2-way switch, you lift the pin with carrier frequency from the board, you solder a wire from the encoder pin to the center pin of the switch, you then solder on side of the switch to the pad and the other side to ground.

    Simply detaching the pin from the pad is not enough because the pin will act as an antenna and may still pick up the floating signal from the pad.

    Here are a few photos so you can see the difference:

    Composite with chroma killer ON:

    Composite PAL 60:

    Composite NTSC:

    RGB Chroma killer OFF (Max shapness to make lines more obvious):

    RGB Chroma killer ON (Max sharpness):


    Last edited by t0nito; 02-23-2017 at 04:37 PM.

  8. #8
    Death Adder's minion
    Join Date
    Jan 2009
    Posts
    12
    Rep Power
    0

    Default

    Thanks for those images t0nito, definitely shows a clear picture of the differences.

    Before the mod mine was a PAL unit, I used to have vertical stripes on the RGB output
    Though I do have to ask, the vertical stripes that you've shown with RGB Chroma killer OFF, I'm assuming that's with one of the crystals (NTSC or PAL) connected rather than system clock. Was there any difference with the vertical stripes when system clock was connected, or was it about the same?

  9. #9
    Wildside Expert
    Join Date
    Aug 2013
    Location
    Portugal
    Posts
    199
    Rep Power
    8

    Default

    The only difference was with the stock clock the stripes were vertical in straight lines, with the crystal the lines are slanted.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •