My Mega Drive doesn't like to be overclocked :(

I don't know how people can make this happen!

I overclocked my mega drive with a 10 MHz oscillator, I kept the wires short as possible, as in right next to the CPU (I put the switch on that hole that almost looks like it was made to take a switch there.

There isn't a single game that doesn't glitch up.

For example, tried sonic 1, sprites started to glitch eventually the game crashes.
Sonic 2, tried the 2 player thing with all those rings, all looks ok, suddenly game crashes garbled screen.

Tried original carts, everdrive, all glitch up, booting directly at 10MHz or booting at 7 then halting the cpu and switching to 10 MHz same results, what gives?

My Mega Drive is a VA6 with a Hitachi CPU.

Can we see photos of your work? What is your 68000 processor named under? How often did you use your soldering iron to the legs of your oscillator?

maybe the oscillator is bad or the switch is bad, or the wiring from the oscillator to the switch is bad.

not many ways it can go wrong.

Camera doesn't do macro very well but:

http://img825.imageshack.us/img825/868/tbre.jpg

Sonic 2 (everdrive) booted at 10 MHz

http://www.youtube.com/watch?v=h4UMIwV76c0

Same but booted at 7 MHz halted then switched to 10 MHz:

http://www.youtube.com/watch?v=AQnrHOxC4wA

Tried this switch, tried a slide switch, same results, so I'm guessing it's not the switch.
Processor is an Hitachi HD68HC000P8

I didn't leave the iron on the oscillator for too long just enough to do the job. Could the wires be too thin? should I try to use wires with shielding?

Thin wires usually work better.

I think maybe the oscillator is just not working good. That Genesis model usually overclocks just fine.

Am i spinning out or is the standard clock wires arse about. I am on my phone so bit hard to see, but it looks like the common of the switch goes to the old clock signal and the "on" goes to the cpu? Sorry if i am wrong but that is how it looks on the phone.

No, the blue wire is going from pin 15 to the center of the switch, a red wire is on the original signal under the pin and the other to the oscillator to each side of the switch.

Sorry for the bad angle, I took a few photos but this one turned out to be the less blurred...

Where are you getting your vcc and gnd for the oscillator?

http://i.imgur.com/2aZps.png

^

I never thought about tapping the stock signal from under pin 15 - I've always pulled it from the cart slot. Gonna have to try this out.

Quote:

---

Originally Posted by Lastcallhall

^

I never thought about tapping the stock signal from under pin 15 - I've always pulled it from the cart slot. Gonna have to try this out.

---

Thats what I did In my master system. Has the advantage of allowing more compact installs.

Quote:

---

Originally Posted by omp

Am i spinning out or is the standard clock wires arse about. I am on my phone so bit hard to see, but it looks like the common of the switch goes to the old clock signal and the "on" goes to the cpu? Sorry if i am wrong but that is how it looks on the phone.

---

I think I'm seeing this too, but he says no.. :/

You guys are right - the clock pin is connected to one of the on pins of the switch. Also, the halt wire appears to be bridging onto pin 18, but that could just be the camera angle.

I know the pic is misleading but this is how it is:

http://img203.imageshack.us/img203/4500/vlqk.jpg

The oscillator power is connected parallel to a nearby capacitor I checked for 5V with the multimeter like this tutorial shows:

http://www.youtube.com/watch?v=oo8qlAIID_M

I really can't think of anything else other than the oscillator might be corrupted or the capacitor your using is ready to go bad. You said your using thin wires too. When I used thick wires on my mods before, I kept getting some negitive issues with them.

I recently did a tutorial on how to do this mod pretty much for beginners and to people who did this before but forgot, that clearly explains what to do.

Yep, that's pretty much what I did, too bad I don't have any other oscillator to try, and I'm afraid if I order another one it will happen the same. Perhaps I should try getting the power from somewhere else. I'm really frustrated as all I read are success stories on their OC'd MDs everyone says that 10 MHz is a good compromise and that the 13 MHz from the cartridge is really unstable, I ordered the oscillator for stability and look where it got me, Sonic 2 won't last 30 seconds LOL!

So, since it was suggested here that the capacitor could be going bad I powered the oscillator directly from the 5V rectifier, also I used a shorter length of wire from the switch to the clock, still no go. Sonic 2 for instance, crashes randomly from 5 seconds to a minute.

I've tried different switches with the same results, only test remaining is soldering the oscillator clock pin directly to the switch, :P.

Here's a better view of the wiring, for those in doubt :cool:

http://img839.imageshack.us/img839/3544/gee6.jpg

I'm thinking that, either the oscillator sucks or the CPU is VERY sensitive. But I read that Hitachi chips are as good or even better to overclock than Motorola chips.

So I'm guessing it's a bad oscillator, talk about bad luck, huh?

If you want to try a new oscillator, you could try using a socket instead of soldering directly to it. To eliminate the possibility of heat damage.

that a good idea, are oscillators that sensitive to heat?

Quote:

---

Originally Posted by t0nito

that a good idea, are oscillators that sensitive to heat?

---

I do not know honestly. But they are made to be soldered so my guess is you should not have any problem with normal soldering heat. However when you have problems to solve, its good to remove as much potential factors as possible.

I think I know what may have caused the problem though, I soldered everything with a soldering gun. Don't know if it's true but according to wikipedia about soldering guns:

"The soldering gun generates an electromagnetic spike when the button is released, which can be a problem for electromagnetically-sensitive devices."

If this is true, I guess I'm lucky I didn't screw up the CPU as it's working fine at stock speed. I'll keep that in mind next time.

Those white wires look REALLY thick, I recommend using the same as the blue and red.

Quote:

---

Originally Posted by Thunderblaze16

Those white wires look REALLY thick, I recommend using the same as the blue and red.

---

Yeah I did have the same wires as the blue and red, I just used that wire to experiment, same results... they are not that thick, they are thin speaker wires. The blue and red are rally thin, to have an idea, they are the inner wire of headphone cables.

Speaker wire is still too think tbh.

Can someone explain me why the thinness of wires would be so important for a clock signal?

Personally I used small multistrand wires I scavenged from old ethernet cables. :mrgreen:

Did not have any problem so far with my 10mhz overclock. Did not put much runtime to it though..

Hmm I think were on something here.

I've heard of the Motorola Chip,
http://cdn.cpu-world.com/CPUs/68000/...-MC68000P8.jpg

The Signetics Chip
http://markfixesstuff.co.uk/wp-conte...ng-genesis.jpg

But never heard of the Hitachi-HD68HC000PS10 Chip
http://cdn.cpu-world.com/CPUs/68000/...8HC000PS10.jpg

Now from what this guy says in 4:28
http://www.youtube.com/watch?v=CGqGqTgq-Tw

And what this link says in the beginning,
http://markfixesstuff.co.uk/tag/sega-megadrive/

It's possibly that your Genesis MIGHT carry an inferior processors and thus, isn't able to be overclocked.

I guess I'll order a new oscillator, and use a soldering iron instead of a soldering gun. I'll reapply the thin wire for the clock signal, I don't think the power supply wire makes any difference so maybe I'll leave those as is...

If after installing the new oscillator the problem remains the same, I guess I can conclude that this particular CPU doesn't like to be messed with... :confused:

Quote:

---

Originally Posted by Thunderblaze16

Hmm I think were on something here.

I've heard of the Motorola Chip,
http://cdn.cpu-world.com/CPUs/68000/...-MC68000P8.jpg

The Signetics Chip
http://markfixesstuff.co.uk/wp-conte...ng-genesis.jpg

But never heard of the Hitachi-HD68HC000PS10 Chip
http://cdn.cpu-world.com/CPUs/68000/...8HC000PS10.jpg

Now from what this guy says in 4:28
http://www.youtube.com/watch?v=CGqGqTgq-Tw

And what this link says in the beginning,
http://markfixesstuff.co.uk/tag/sega-megadrive/

It's possibly that your Genesis MIGHT carry an inferior processors and thus, isn't able to be overclocked.

---

Actually it's the other way around, the guy on the video is talking about the Signetics chip, and on the second link is actually says:

Quote:

---

"Undeterred I set about looking for a replacement CPU, and through reading a lot of internet sources I learned that the best CPU in peoples (admittedly unscientific) experiences was the Hitachi variant.

Well the catchilly named “HD68000P8″ appeared to be the CPU of choice. Apparently this CPU is the best unit when it comes to over-clocking."

---

Since my CPU is supposedly the best one to overclock, I'm finding it very strange that it get so unstable at just 10 MHz.

Yeah I notice that after reading it a bit more.

I guess it really is just your oscillator.

And once again, the Signetics actually overclocks fine. It's just a problem with the early motherboard themselves.

Quote:

---

Originally Posted by evildragon

And once again, the Signetics actually overclocks fine. It's just a problem with the early motherboard themselves.

---

What part of these early boards tho make it a pain? Was the quality of them lacking or something?

Damn, only 10 MHz oscillator at a reasonable price on ebay is from the same supplier, I'm not sure I wan't to order another one from there... I guess I'll have to put this project in stand-by for now...

Anyone got a spare oscillator? :p

Earlier IO controller. On the older signetics based systems they were two chips, on the later systems it was an all-in-one IC. Wasn't quality, it was just older and less tolerable.

Quote:

---

Originally Posted by evildragon

And once again, the Signetics actually overclocks fine. It's just a problem with the early motherboard themselves.

---

I also got the impression that the boards with Signetics CPUs were specifically problematic compared to other early boards sporting Motorola chips. (again, not due to CPU)
So either it's coincidence, or something specific to motherboards used in manufacturing plants also using that CPU.

AFIK noone has disproved the "boards with Signetics CPUs are bad" thing, though I suppose I could try it out with my 2 Signetics VA3 boards. ;)
I actually did do that for one of them, but at 12.1 MHz and without a switch. It was buggy and crashed, but actually worked better than my VA6.8 in the same set-up. (I have yet to go back and do a proper overclock with speed switch and halt switch -or button)


That said, not all games are going to overclock well in any case as the ROMs may be too slow to cope. (especially bad for early releases) I'd assume flash carts would be among the most stable.

Quote:

---

Originally Posted by kool kitty89

That said, not all games are going to overclock well in any case as the ROMs may be too slow to cope. (especially bad for early releases) I'd assume flash carts would be among the most stable.

---

Funny though, I have an everdrive and my mega drive crashes much faster with it than original roms, with Sonic 2 on the everdrive it crashes anywhere from 1 to 30 seconds.

I had sonic 2 crash at the SEGA logo, so there has to be something very wrong with my oscillator.

You have to keep your wires as short as possible. Preferably less than 10cm. If you cannot use shielded cables.
Ground wire has to be thick and clock line should be thin. You may want to add 10µF and 0.1µF caps on the oscillator power pins.

Thanks for the tip, if you look at the photos a few posts back you'll see I tried thin and thick wire and they're very short, only thing left is to try those capacitors! I just can't understand what the problem is, I've searched everywhere and can't find anyone who had the same problem.

How do I connect those capacitors?

One end to GND pin, other end to VCC pin.

It is also likely that the machine just won't overclock. The parts are not created equal. During manufacture all parts are tested and those that match or exceed the intended operating conditions get packaged and go into their intended designs. You may just have hit a case where the part just matches operating conditions.
MD overclocking is primarily limited by the bus management chip, less by CPU, VDP and IO chips.

Quote:

---

Originally Posted by TmEE

It is also likely that the machine just won't overclock.

---

That's what I'm starting to believe...

About the capacitor connection to mean in parallel like this?

(I just realize you said one 10uF and 0.1uF so one is electrolyte and the other non polarized ceramic both in parallel?

http://img138.imageshack.us/img138/5859/cja3.jpg
Also what's the purpouse of the capacitor? to filter out noise?

Just a small update (if you can call that).

So I soldered a 10uF capacitor in parallel with the power supply, I soldered the clock pin directly to the switch, so no wire type/length issues there.

http://img547.imageshack.us/img547/7849/8m32.jpg

Still no go :'( game crashes a few seconds after switching on the MD.

Because of the capacitor and power noises and such I just remembered that my power brick is a universal power supply, it is an electronic supply, not those heavy transformers.

Could that be inserting noise to the oscillator making the clock unstable?

I guess it's the only thing to rule out other than:

a) The oscillator is faulty.
b) Probably the first known unoverclockable MD.