I'll have to try and get a higher rating power supply and try it out.
Printable View
Tried a 12V 2A power supply for a small period, I guess the source is noisy because I could see some dots on the screen.
Results with OC:
Everdrive booted to menu, when selecting play game, nothing, black screen.
Booting on 7 MHz halting then switching to 10 MHz, immediate crash.
So I'm guessing all this headache is due to a noisy power supply, maybe ripples from the 5v DC source is interfering with the 10 MHz clock.
I already put a 50V 10uF capacitor on the oscillator as suggested here but having the same results.
Any other suggestions as to filter out the noise?
I've looked for a power supply on ebay, I need one with a Euro plug because I'm in Portugal, on ebay UK I found a few but the plug isn't the same and a plug adapter will become too cumbersome. Other than that I found a few 3rd party ones that make me feel doubtful about how clean the source will be...
get a 1000µ+ capacitor and stick it on the DC input port like on the oscillator.
I think you really need an oscilloscope to see what's going on if what TmEE recommended doesn't work.
put in a 10v 1000uF parallel with the power supply, damn, still crashes a few seconds after turning it on...
Ah yea, some of those ebay power supplies....Shebus they are bad! When I find a good one, I quickly order more from the same supplier before they disappear! The bad one's go in the bin where they belong. Finding [decent] Aussie power supplies is pretty bloody tough.
Does the everdrive even run reliably when on an overclocked machine ?
Some youtube videos show that yes it does. But in my case even normal carts crash and glitch when overclocked.
It's funny that I also read another case that someone tried everything and couldn't get the Mega Drive to overclock properly, even at 8 MHz.
The only solution he found was to get another console, he overclocked it successfully to 12 MHz.
And guess what version of the console he managed to overclock, the exact same that I have a VA6 (IC BD M5 PAL) board with a Hitachi CPU which makes me even more frustrated!!!
I was slowly making preperations to do this mod (so far I got a 10Mhz and 13Mhz clock and wired up a switch) but it seems that there are quite a few potential problems / risks and the benefits few and far between. If I do decide to go ahead, for stablity reasons I trust that 10Mhz is sufficent? (I have the Motorola 68000 btw).
There is only one problem: either the mod works or it doesnt. The higher you go in clock speed, the greater chance you have of a crash. The 68000s inside the MD/GEN are all prone to crashing under variable conditions. The best thing to do is keep your wires short, simple and organized. 10MHz should be fine for most anyone and any game. In fact, I have yet to come across a game with my 10MHz system - Everdrive or not - that has slowdown or crashes.
I do recommend every one install a halt switch with the OC switch, though...
This is the only thread I have ever seen a 10mhz overclock that does not work. Mainly it should. 12mhz is not for most people. It has major issues and only the lucky few really get it to work and only in some games.
My 10mhz works fine, even with a everdrive. A few games are sensitive to it and work poorly or not at all, but these are the very few. Just switch it off if the game has issues.
Well, I have the same problem with the same board. A VA6 IC BD M5 PAL board with a Hitachi HD68HC000P8 CPU. Doesn't like to be overclocked either and crashes withing 60 seconds no matter what I do. In the end I cut out both the overclock select and halt switch and soldered everything directly to the board. Used multistrand 30AWG wire for the 10Mhz signal and 5V line, and multistrand 24AWG wire for the ground line. Wires were kept under 2 inch, and the 10Hhz signal line under 1 inch. Oscillator pins only touched for a second while soldering. Used a different oscillator after disconnecting the switches. Tried tapping 5V and ground both from the chip itself and the adjacent capacitor. Used 3 different official MD power supplies. All to no avail.
Attachment 10386