Forums

I had some problems with some games on a Mark III, garbled image (looking as corrupted sprites) or game not starting. I discussed this here. However, I noticed that both on the Master System and on another Mark III they work flawlessly.
Investigating the hardware, I found the following weird situation.

If the mark III is reset, the games start working. I reset by grounding pin 23 of the VDP.

I thought the problem was an unstable power source, so I replaced the 7805, but nothing changed. I tried several original AC adaptors, and the problem does not depend on it. I checked all capacitors, all have low ESR. I cleaned the board and resoldered every single pin.

I supposed the time for which reset is kept low after the power on
is too short. I then checked C18, which is exactly in the same configuration of C131 of the master system (see the schematic here. If I am not wrong, this RC is used to keep reset low for a bit after power on.
Capacitor and resistor were fine.I checked another mark III,it uses the same values, so no errors at assembly time.
I then tried increasing the capacitor, thinking that a longer reset could only help. To my surprise, in that case the problem persists. If, however, I remove it or replace it with a smaller one (now I put a 100n there) it works well.

Can someone guess why this happens?

The only reason that come to my mind is that with the slow slope of the capacitor, Z80 and vdp exit from reset at too different times, and so the problem happens, but that never happened to me elsewhere.

I'm curious about this too. There is no reset signal brought out to the Mark III cartridge port, so games using mapper chips that need a reset to select the correct banks don't work correctly. Does this mean that all Mark III games that have a mapper chip include a reset circuit in the cartridge, such as your Afterburner cart?

I posted earlier about a Mark III development cartridge addressing this problem by including an IC to handle resets for the mapper chip, so it looks like that was at least one solution to the problem by Sega.

Thank you for your comment. Could I ask you where you posted about a Mark III development cartridge? I am developing the same, so I'd be interested in seeing what are other people doing.

Afterburner does not have a reset chip, phantasy star (which uses the same mapper) does. It uses this
this chip.
I then wonder if that chip is missing in the export version of phantasy star!
If I can get my hands on a Penguin Land cart, I will check if it has the same resetter chip. Seeing after burner (absent) and phantasy star (present), I would bet it is required any time SRAM is actually used.

kamillebidan wrote

Seeing after burner (absent) and phantasy star (present), I would bet it is required any time SRAM is actually used.

Maybe so, and maybe that is why SMS didn't need the reset button warnings NES games did.

kamillebidan wrote

Thank you for your comment. Could I ask you where you posted about a Mark III development cartridge?

Oh I was just describing that a MB3771 reset circuit was used in a Mark III/SMS hybrid development cartridge from Sega. It's pretty much the same thing as the PST529F you linked to. I haven't made a homebrew devcart myself.

Quote

I then wonder if that chip is missing in the export version of phantasy star!

Many SMS games add a reset chip to drive CS2 on the SRAM such that it is disabled during a power-up or power-down event. This isn't used to reset the mapper chip which uses the regular system reset signal, only to disable the SRAM. The US SMS version of Phantasy Star does this too.

Author	Message
kamillebidan Joined: 25 Feb 2013 Posts: 391 Location: Osaka	Weird reset problem - need help from hardware oriented members - Posted: Sat Mar 29, 2014 8:51 pm
	I had some problems with some games on a Mark III, garbled image (looking as corrupted sprites) or game not starting. I discussed this here. However, I noticed that both on the Master System and on another Mark III they work flawlessly. Investigating the hardware, I found the following weird situation. If the mark III is reset, the games start working. I reset by grounding pin 23 of the VDP. I thought the problem was an unstable power source, so I replaced the 7805, but nothing changed. I tried several original AC adaptors, and the problem does not depend on it. I checked all capacitors, all have low ESR. I cleaned the board and resoldered every single pin. I supposed the time for which reset is kept low after the power on is too short. I then checked C18, which is exactly in the same configuration of C131 of the master system (see the schematic here. If I am not wrong, this RC is used to keep reset low for a bit after power on. Capacitor and resistor were fine.I checked another mark III,it uses the same values, so no errors at assembly time. I then tried increasing the capacitor, thinking that a longer reset could only help. To my surprise, in that case the problem persists. If, however, I remove it or replace it with a smaller one (now I put a 100n there) it works well. Can someone guess why this happens? The only reason that come to my mind is that with the slow slope of the capacitor, Z80 and vdp exit from reset at too different times, and so the problem happens, but that never happened to me elsewhere.

Charles MacDonald Joined: 28 Sep 1999 Posts: 1200	Posted: Sun Mar 30, 2014 4:42 pm
Charles MacDonald Joined: 28 Sep 1999 Posts: 1200	I'm curious about this too. There is no reset signal brought out to the Mark III cartridge port, so games using mapper chips that need a reset to select the correct banks don't work correctly. Does this mean that all Mark III games that have a mapper chip include a reset circuit in the cartridge, such as your Afterburner cart? I posted earlier about a Mark III development cartridge addressing this problem by including an IC to handle resets for the mapper chip, so it looks like that was at least one solution to the problem by Sega.

kamillebidan Joined: 25 Feb 2013 Posts: 391 Location: Osaka	Posted: Sun Mar 30, 2014 4:56 pm
	Thank you for your comment. Could I ask you where you posted about a Mark III development cartridge? I am developing the same, so I'd be interested in seeing what are other people doing. Afterburner does not have a reset chip, phantasy star (which uses the same mapper) does. It uses this this chip. I then wonder if that chip is missing in the export version of phantasy star! If I can get my hands on a Penguin Land cart, I will check if it has the same resetter chip. Seeing after burner (absent) and phantasy star (present), I would bet it is required any time SRAM is actually used.

KingMike Joined: 14 Oct 2008 Posts: 533	Posted: Sun Mar 30, 2014 5:20 pm
KingMike Joined: 14 Oct 2008 Posts: 533	kamillebidan wrote Seeing after burner (absent) and phantasy star (present), I would bet it is required any time SRAM is actually used. Maybe so, and maybe that is why SMS didn't need the reset button warnings NES games did.

Charles MacDonald Joined: 28 Sep 1999 Posts: 1200	Posted: Sun Mar 30, 2014 11:07 pm
Charles MacDonald Joined: 28 Sep 1999 Posts: 1200	kamillebidan wrote Thank you for your comment. Could I ask you where you posted about a Mark III development cartridge? Oh I was just describing that a MB3771 reset circuit was used in a Mark III/SMS hybrid development cartridge from Sega. It's pretty much the same thing as the PST529F you linked to. I haven't made a homebrew devcart myself. Quote I then wonder if that chip is missing in the export version of phantasy star! Many SMS games add a reset chip to drive CS2 on the SRAM such that it is disabled during a power-up or power-down event. This isn't used to reset the mapper chip which uses the regular system reset signal, only to disable the SRAM. The US SMS version of Phantasy Star does this too.

Forums

View topic - Weird reset problem - need help from hardware oriented members -