Overclocked NES. Yes...I actually did it...

Discuss hardware-related topics, such as development cartridges, CopyNES, PowerPak, EPROMs, or whatever.

Moderators: B00daW, Moderators

Post Reply
User avatar
AWal
Posts: 132
Joined: Wed Oct 06, 2004 8:00 am
Location: Northern New York
Contact:

Overclocked NES. Yes...I actually did it...

Post by AWal » Thu Mar 08, 2007 4:12 pm

the instructions are available elsewhere, but I brought up the discussion of Kirby's Adventure causing slowdown, with a fix of using an overclocked CPU to eliminate it.

Image

The image pretty much shows everything I did. Although pulling up the ppu's clock was not necessary, I did it so I could "see" a changed clock speed (since the speedup only effects the sound of games timing from vblanks).

I don't actually have a copy of Kirby's Adventure, I'll be taking care of that soon enough.

I'll post some videos, just give me a bit...

P.S.:Don't ask about the CIC...

P.S.S.:The other crystal is on a seperate pcb because the only oscillator I had lying around ran at 48MHz, so I constructed an oscillator myself:

Image

User avatar
No Carrier
Posts: 290
Joined: Tue Dec 13, 2005 4:19 pm
Location: Gainesville, FL - USA
Contact:

Post by No Carrier » Fri Mar 09, 2007 4:43 am

What about Dodgeball? I remember that game being annoyingly slow..

NC

User avatar
AWal
Posts: 132
Joined: Wed Oct 06, 2004 8:00 am
Location: Northern New York
Contact:

Post by AWal » Fri Mar 09, 2007 8:35 am

No Carrier wrote:What about Dodgeball? I remember that game being annoyingly slow..
I'll add that to my most wanted list. That game feels like it's programmed to run at a lower rate though, but it's nothing that can't be tested.

[man, who disables embed code...]...anyways:
http://www.youtube.com/watch?v=fgkKftUQy-8

User avatar
never-obsolete
Posts: 381
Joined: Wed Sep 07, 2005 9:55 am
Location: Phoenix, AZ

Post by never-obsolete » Mon Mar 12, 2007 10:33 am

AWal wrote:
No Carrier wrote: What about Dodgeball? I remember that game being annoyingly slow..
I'll add that to my most wanted list. That game feels like it's programmed to run at a lower rate though, but it's nothing that can't be tested.
i'd like to know if river city ransom falls under this category as well
. That's just like, your opinion, man .

User avatar
AWal
Posts: 132
Joined: Wed Oct 06, 2004 8:00 am
Location: Northern New York
Contact:

Post by AWal » Tue Mar 13, 2007 8:53 pm

No Carrier wrote:Dodgeball
never-obsolete wrote:River City Ransom
There's a good chance these both use the same engine...just a hunch....

Epicenter
Posts: 64
Joined: Mon Mar 26, 2007 7:30 pm
Location: USA
Contact:

Post by Epicenter » Mon Mar 26, 2007 7:33 pm

I'm glad to see you're getting good use out of my guide. I'm putting together a more advanced version of it some people might enjoy .. right now, I'm doing some work to try and exceed my current maximum, 4.2 MHz (>230% original CPU speed but I'm a crazy bastard like that.)

What is the frequency you ended up with? Good job on the homemade oscillator circuit.

User avatar
AWal
Posts: 132
Joined: Wed Oct 06, 2004 8:00 am
Location: Northern New York
Contact:

Post by AWal » Tue Apr 10, 2007 4:37 pm

Wow, I didn't expect the author of that guide to reply, lol.

The crystal runs at 28.224MHz (which /12=2.352MHz). Like I said, the only actual oscillator I have runs at 48MHz, and even though that worked, most games locked up with timing errors, irq lockouts, etc., although I was suprised that mario 3 showed the title screen quite stable (with the IRQs disabled).

Recently I shorted out somthing on my NES that prevented any kind of operation (although bus pins were still functioning apparently), so I ended up eBay'n a new one ($25 w/Free shipping, bare bones and I got a free game + controller, not like I needed the psu or rf, lol).

I am now installing a switch to roll back to the original frequency, and it's really creeping me out that some games (especially Ultimate Stuntman) are allowing clock changes in action.

Hopefully I'll be getting a copy of Kirby soon to see if this actually helps.

I guess I can make a new video as well, lol, since not many ppl are doing this and it's seems like one of the most off-beat overclocks (...to me anyways).

Epicenter
Posts: 64
Joined: Mon Mar 26, 2007 7:30 pm
Location: USA
Contact:

Post by Epicenter » Tue Apr 10, 2007 7:16 pm

Kirby had the most dramatic improvement. Most lag was gone at 2.3 MHz, pretty much all of it gone at 2.6, and aside from the status bar twitching a bit (CPU missing IRQs from the MMC I'm sure, as you mentioned) it was perfectly stable up to 4 MHz, I think 4.2 actually, where my CPU maxed out at.

You're not the first person to find on-the-fly clock changes work without crashes-- I used a variable oscillator and switching the multiplier settings on its PCB would not crash the system, ever. My friend Rob made a switch to toggle between onboard clock and an oscillator and it doesn't cause crashes, either. I was surprised because when overclocking the MegaDrive/Genesis I found you need to halt the 68000 before switching its clock most of the time, or it'll lock up.

User avatar
AWal
Posts: 132
Joined: Wed Oct 06, 2004 8:00 am
Location: Northern New York
Contact:

Post by AWal » Fri Apr 13, 2007 4:53 pm

My switch is sucessfull most of the time, it's a dated switch (from an old A/B select from some wireless crap that didn't work anyways), so it crosses paths inbetween switches, and that will lock it up somtimes.

Post Reply