MegaDrive 1 CXA1645 mod

Odd question here but would it be possible to solder my Scart cable directly to the board?

Just the cable I have needs some serious "waggling" to maintain a picture and sound, not sure why, just cheaply made I guess.

I'd be quite happy to have a cable permanently coming out of the system.

 

x

 

Somewhat related, I tried to use a CXA1645 to output S-Video and composite in a TurboGrafx 16. Thing is the circuit flat out doesn't work, could be something on my end (should've made a PCB ahead of time me thinks) but I noticed something.

The 7805 outputs 4.5v with the console on without any modifications installed. With the CXA1645 and associated circuitry the voltage drops to 3.4v. Seems a bit drastic a voltage drop and I'm guessing it is the main cause as to why I can't get this thing to work. Either that or the sync separater circuit doesn't want +5v. I know some voltage drop is normal, but could this chip actually function if fed 4.5v?

Offhand I don't recall which iteration I have but in the datasheet it is listed as having a much higher current draw maximum than the alternative model. The TG16 PSU isn't exactly a workhorse either.

EDIT: Tried powering it externally and drew RGB+sync from the expansion connector. Composite out changed picture a bit but nothing remotely considered usable but it seems my circuit is either built improperly (likely) or the diagram I was using is incorrect (very possible).

 

Hey all! Sorry to bump this thread but I found out about it by searching this forum.

I did this from page 1 to my model 1 Genesis:

http://mmmonkey.co.uk/console/sega/md-sync.htm

I get a beautiful, stable, non-flickering image...only it's completely red

It's like what you'd expect if I only plugged the red cable from a component device in, but it's not that because my SNES and RGB modded NES look wonderful. It's also not the cable because I actually own 2 different RGB Genesis model 1 scart cables made by 2 different people.

I made sure the cap and resistor are soldered to pin 11. I even went as far as trying different pins just to make sure I'm not counting wrong or something. I even tried different caps. I am getting a completely red image. It's a stable, jailbar-free image, but it's completely red. I even removed the s-video and composite mod thinking that might have been screwing with it. If someone could help, I'd appreciate it greatly. This is a US "High Definition Graphics" non TMSS Genesis. Thanks kindly.

 

What happens if you undo the sync mod and returns the Genesis to it's original state?

 

That would require me un-destroying the trace that guide told me to destroy. How would I do that?

 

Theres a difference between "destroying" a trace and cutting it.

If you just cut it, then scrape away to expose the copper on both sides of the break and solder some wire to bridge the gap - reconnecting the trace.

 

Follow the trace from where you cut (hopefully not literally destroyed) back to where to comes from as best you can. Solder wire there, solder wire to where sync used to go. Done (hopefully).

 

I would also check the traces going to, and from the CXA1145's green and blue In and Out-puts.

That would be pin 3, 4, 21 and 22.

 

Last week I tried adding a second CXA1145 encoder to my MegaDrive, to only run the blue color through. I got a picture but it was flickering like crazy and strangely there was still jailbars :/

In the CXA1145 datasheet there are schematics that explains what's going on inside the chip. What I'm interested in is what's happening between pin 2,3,4 and 21,22,23 (this is where the colors are passed through).



I was hoping that someone here with a fair knowledge of electronics can help draw up a schematic from the above so we can build this at a much larger scale. The end result should be a perfect RGB color amp.

CXA1145 schematics:
http://www.alldatasheet.com/datasheet-pdf/pdf/77343/SONY/CXA1145P.html

 

You need to feed SYNC to the buffer CXA1145 you just added because it has a clamping circuitry (clamp circuit shorts R, G and B to GND at a certain phase of the SYNC signal to restore DC level) and SYNC is what synchronizes it.

The fact that when you connect the Mega Drive (with original SYNC circuitry from SEGA) to a monitor with low impedance SYNC screws RGB output signal proves that. When you RUIN the SYNC signal (SONY chips are quite sensible to the sync voltage, it has to be something around 1.7v to 2.0v), it breaks all outputs of the chip including the RGB buffer circuit.


SO MAKE SURE you FEED SYNC to the new RGB buffer chip too. :thumbsup:

Edit: also, since you're having most of the new chip disconnected, use the RED buffer for your blue instead of the chip's BLUE buffer.

 

I did, both things But it didn't work for some reason. Also tried using the green buffer just in case :/

Now I'm back to the colorboost amp once again. I also tried the PNP transistor mod discussed earlier. That worked, but was much too bright and once I got the brightness to an acceptable level the colors were still changing like crazy when the screen was moving.

The colors of the sky in Tiny Toons could change from lightblue to a light greenish/blue when I jumped up high, using the old color amp. That's why I decided to try something new...
In the end it seems I could have solved that problem by replacing the first resistor with something larger. It does give a darker picture but nothing that can't be adjusted on the XRGB-3.



So, yeah. Until someone else finds and tests a better solution this is what I'll stick to

 

Awesome, wish the datasheet had mentioned that when I was building my encoder circuit. Probably why it doesn't work.

 

It does mention (indirectly or briefly) that the chip is doing clamping, but fails to mention that the clamping circuit actually needs the sync to work.

Just look any clamp circuit on the internet. Will probably be using an LM1881 to get the clamp timing out from the sync signal. :shrug:

 

They do. When I built my circuit (that is still sitting here on verboard) I found a way around needing a LM1881 for the clamp timing but I can't for the life of me find where I got the idea from.

If I'm lucky I can retrofit the appropriate components and get some life out of this thing. Otherwise I spent $15 on this 1645 for nothing. :crying:

 

Good/bad news for me here. I'd picked up a Genesis 1 recently and notice RGB bars even on CRT. As expected it takes extreme blueness to see clear with that type of TV, like the sky in Flying Battery Zone.

Which MD-based systems and revisions have we definitively ruled out from being affected by this or similar RGB issues?

 

For anyone that's interested I got a cxa1645 going in my va6 high definition graphics model 1 genesis. I built a kit that has the sync go through a lm1881 and then feeds rgb and cleaned up sync into a 1645m and has all the necessary parts. I desoldered the 1145 and threw in a precision socket and took screenshot comparisons between the 1145 and the 1645. When using the 1645 I remove the 1145 and just attach rgbs, 5v and ground to power the kit.

 

x

 

That is so strange. But after a while it corrects itself?

 

x