PDA

View Full Version : Overclocking the DC and where to get the Increase



Cross
June 25th, 2005, 04:31
Ok I have my DC apart and I am ready of OCing but of course I need a chip for increasing the mhz from 33.333 to something above 40. I want the highest I can get as its main use will be for SNES and NES emu's.

I have been reading up and found that it seems the use of a quartz crystal was pref but no longer as easy to find. What is a good sub? I do not plan on putting this DC back to the stock settings as I have another to play with. I am curious though as it says 40mhz was the fastest with no cooling mods, I have no problem upping the cooling but I would like to know what others have tried, like the peak of 276 (I have heard 286)basically I would like to know what they used. I would like to hit the same 276 since that seems to be the limit of the stock voltage. So I need a sub for a full time mod to my DC for this, if I can get it at radio shack all the better.

Thanks guys!!!

Cross
June 25th, 2005, 19:43
Aww, nothing yet and I am getting ready to go to town...anyone have anything for me on what to use for the sub?

Morph
June 26th, 2005, 07:33
You still need a crystal oscillator in order to overclock. A 45-46MHz one will do nicely in your case. If there was a way to increase the stock voltage, you COULD just sub in the gfx clock line and get 324MHz, but this is impossible until I can track down where the CPU gets it internal 1.8v power from. So until then, you NEED an oscillator, its not something you can bypass.

Cross
June 26th, 2005, 09:05
I think I found one its a 46mhz I will get it monday. Now to see what kind of cooling works. I think I can mod a laptop heatsink in.

Morph
June 26th, 2005, 21:19
Good luck. The highest AVERAGE overclock is about 270MHz. Simply because not many oscillators seem to offer successful LARGE overclocks. Only 1 person has hit 286, and he himself doesnt know why. I am nearly certain its voltage related, but no way to prove it.

ptr.exe
June 27th, 2005, 12:27
It's weird that matt doesnt know how he did it, is there any test for the CPU speed? such as a simple program that runs clock tests on GPU and CPU, would be useful for OCing DC's.

Cross
June 27th, 2005, 20:19
I guess I will find out if its any good once I try. I just want to push SNES Emulation further and since I can not code it this is the only other option.

ON another note, I asked a few friends of mine at another forum to look into the voltage source for the processor, they said they would take a look.

ptr.exe
June 27th, 2005, 22:00
ON another note, I asked a few friends of mine at another forum to look into the voltage source for the processor, they said they would take a look.
Dozens of Dreamcast fans have spent hours trawling over DC mobo's trying to figure out the point where the voltage supply to the Sh-4 is produced and no one as of yet has found anything... but your gonna ask your mates, so we can all sleep easy at night now :p

Just joking, but this info won't be found easily on the internet as no one knows of anyone who has done it yet.

Morph
June 28th, 2005, 06:26
I got referred to another place by Dan Potter. It might be possible that i find out. As far as matt and the 286MHz OC are concerned, he didn't know how he did it since he had tried an oscillator rated lower and got a no-go, but the higher one worked. He knows its at 286, but not sure WHY the DC accepted that co.

ptr.exe
June 28th, 2005, 16:42
Thats strange, lucky git must have had a DC that for some reason could do it.

Cross
June 28th, 2005, 17:40
Dozens of Dreamcast fans have spent hours trawling over DC mobo's trying to figure out the point where the voltage supply to the Sh-4 is produced and no one as of yet has found anything... but your gonna ask your mates, so we can all sleep easy at night now :p

Just joking, but this info won't be found easily on the internet as no one knows of anyone who has done it yet.

LOL they r not internet searchers, they love to trace circuits. They are some of my military buddys and thats what they do for the military. I don't expect them to do any better but we all have people who if we think someone could do it we would say them. I trust them just because I know if its beyond me they always seem to love jumping in and doing it.

obelisk
June 28th, 2005, 22:31
bump for dc. keep up the stuff, and let us know your findings.

Morph
June 29th, 2005, 06:04
I have faith in you man. Hell, if I had a logic analyzer, I'd do it myself. But, eh, not so lucky.

Cross
June 29th, 2005, 08:32
We shall see I am taking them my spare DC tomorrow. I am not sure how long it will take but here goes nothing lol.

Morph
June 29th, 2005, 13:34
good luck! keep us posted.

ptr.exe
June 29th, 2005, 14:29
Cross i apologize, but when someone on a forum says "My mate..." you immediately just assume its nonsense. Hopefully they can find something.

Morph
June 30th, 2005, 23:31
Well, I just got a reply from someone at that site who definately seems to know his stuff. Get this, that CPU line? Thats NOT a CPU line. The the motherboard FSB line. So when we overclock the CPU, we are actually overclocking the RAM and everything else. Here is the reply.

"This is not entirely correct. The basic clock signal comes from a 13.5MHz
oscillator that is fed to a PLL that indeed generates a 54MHz and a
33.33MHz signal. The 33.33MHz signal is fed to the SH4 which generates a
200MHz signal that is used internally, but it also generates a 100MHz
clock that is fed to the rest of the system, including the Holly chip
(which includes the graphics core but also lots of bus control logic) and
SDRAM. That 54MHz from the PLL is indeed also fed to Holly, perhaps it is
even used internally by the graphics system, I'm not sure; the main stuff
would be running off the 100MHz clock though. The primary use of this
clock seems to be to as a pixel clock for the video encoder.

You state that you want to overclock the SH4 to 300MHz, by I'm guessing,
feeding it a 50MHz clock instead of 33MHz. You think the problem is the
1.8V. Sorry, I can't help with that at the moment, but I'm thinking there
are lots of other problems too, such as clocking the GPU and main/video
RAM at 150MHz instead of 100MHz..."

If I understand his point correctly, I will probably also have to boost the FSB (ex-CPU) line to keep it synced with the CPU. Which isn't a problem, it simply changes the playing field a bit.

Cross
July 1st, 2005, 07:14
SO we are also OCing the RAM and other parts...hmm I will pass this on.

Morph
July 1st, 2005, 07:40
nonono, ignore that whole bit of information. The CPU speed is still independant of the FSB line. So all I need still is the 1.8v thing. Thats just clarifying that what was previosly considered a CPU-only overclock is in fact a FSB overclock.

ptr.exe
July 1st, 2005, 13:39
what was previosly considered a GPU overclock is in fact a FSB overclock.
If that's true then it would explain alot of strange behavior with some games on a GPU OCed DC.

Morph
July 1st, 2005, 15:14
Oh crap. I totally misinterpreted that. Get ready for long ass post, strted off by my reply to what he said X.X

"> VERY interesting. I had no idea the so-called "GPU" pin was actually
> the FSB pin. I appreciate the information. So what your saying, is
> that to make overclocking the CPU more possible, I would also possibly
> have to raise the FSB so that the CPU doesn't go out of sync?

I think you have misunderstood what I wrote. Forget about that 54MHz "GPU"
pin for a moment. Here's a simplified clock diagram (use fixed width
font to view properly):

.-------. .---------------.
| Xtal | | SH4 |
| 13.5M | .---+--+---x6-CORE | .------------.
`---+--' | | | 200M | | SDRAM |
| | | | | 100M | Main mem. |
.---+------. | | `---x3-------+---+---------+ 100MHz |
| PLL | | | 100M | | | |
| 54M 33M | | `---------------' |100M `------------'
`--+----+--' | .-------------------+------. .------------.
| | | | Holly | | | SDRAM |
| `----' | (System ASIC) .--+---. | | Video mem. |
`-------------+---. | `-+--+ 100MHz |
| | .-------+----. | | |
| /n | PowerVR | | `------------'
| | | 100MHz | |
| | `------------' |
`---+----------------------'
|13.5M/27M
.---+---.
| Video |
| DAC |
`-------'

As you can see, the 33MHz clock is multiplied by 6 inside the SH4 to
generate its core operating frequency of 200MHz. It is also multiplied by
3 to generate a 100MHz clock that in your terminology might be called the
"FSB" clock.

This "FSB" clock is used to drive the SDRAM and most stuff in Holly
(including the GPU). Since the 100MHz is generated by multiplying the
33MHz clock by 3, if you change the input to 50MHz to get a 300MHz CPU
clock, the "FSB" clock will be half of that ie 150MHz instead!

Listen to what Simon said: the GPU limit might well be 133MHz (and since
he designed the damn chip, you might want to consider he knows what he's
talking about!). If the GPU is running at 133MHz, the SH4 will be running
at twice that, namely 266MHz, which is close to the limit you state:
270MHz. So it seems quite probable to me that it's not the SH4 that is
the limiting factor. It very well could be the GPU!

Returning to that 54MHz "GPU" clock now. As you can see it is not really a
"GPU" clock at all; it is more properly termed "pixel clock" since it is
used to drive the video encoder/DAC. The 54MHz pixel clock is not
multiplied by 2 (as I've seen asserted); it is actually divided by a
divisor n that is controlled by bit 23 in hardware register a05f8044:

bit 23
------
0: n = 4 (pixel clock=13.5MHz NTSC/PAL mode)
1: n = 2 (pixel clock=27MHz VGA mode)

So overclocking this clock will do little to increase the GPU speed. It
will only screw up the refresh rates so the signal is no longer NTSC/PAL
compatible :)

> But yeah, still looking for the 1.8v thing-a-ma-doo-dad. :D

I'm **GUESSING** that IC104 **MIGHT** be the one to generate the 1.8V. It's a
five-pin 1.5-7V variable voltage regulator that you'll find it on the
flip side of the PCB, near the SH4 but further towards the edge. I would
not recommend you to attempt anything drastic without being sure you know
what you're doing, and if anything goes wrong, don't blame me!"

So in fact, the GPU/FSB pin only controls the pixel clock, not either the GPU nor the FSB. The *CPU* pin controls EVERYTHING.

Cross
July 1st, 2005, 19:07
lol, more info to the flame. Well lets see where this takes us

ptr.exe
July 1st, 2005, 20:46
again that explains alot, i have to say congrats for finding this out, so many people are completely wrong about what those generated clocks do.

Who is telling you all this btw?

quzar
July 1st, 2005, 21:01
This was just being discussed on the DCdev mailing list, so i assume it was from there.

ClumsyCliff
July 1st, 2005, 21:01
That would be Lars Olsson from the DCDev list on Yahoo.

ptr.exe
July 1st, 2005, 21:05
I see, how do you get on that mailing list?

ClumsyCliff
July 1st, 2005, 21:50
You'll need an account on Yahoo Groups, go to http://groups.yahoo.com, then search for DCDev or just use this link http://groups.yahoo.com/group/dcdev/

Morph
July 1st, 2005, 22:41
yeah, just go there and sign up, then you are put on the mailing list. The "group" is basically a single forum where people post in depth information or questions about the hardware/software of the Dreamcast. And every reply is sent through the mailing list, and is posted, so you know asap.

Pretty cool stuff. I got forwarded there by Dan Potter, and yes it was Lars who replied.

So basically, anyone with a voltometer or the like, would you mind testing IC104 on the bottom of the Dreamcast motherboard? Its only 5 pins, so it shouldn't take long.

ptr.exe
July 2nd, 2005, 11:30
You don't have a multimeter and yet your doing all this? lol.

Yeah the pinout is:


IC 104
________
| |
| |
| |
5 4 3 2 1

1: ground
2: 1.24V
3: 1.99V
4: 3.24V
5: 3.24V

thats what my multimeter said, no 1.8V though, perhaps Sega underclocked the SH-4? so maybe its the 1.24V?

Morph
July 2nd, 2005, 21:03
Nah, the SH4 isnt underclocked that i can tell. The official SH4 product thingamajigger from Hitachi says its 200MHz. Odds are, either the DC version is a low power version (seeing how is hard coded low-endian) or perhaps it gets more juice to keep it more stable. I would doubt the assertion of more juice, since that would have increased the amount of heat, then again, that explains how the DC can hit a nearly 40% overclock without a voltage increase, something impossible on any other processor I have ever seen.

So, odds are, its probably the 1.99v lead, unless there is still another circuit on the DC motherboard that we don't know about. But I doubt that since I am almost certain that the 3.24v leads are for the I/O voltage. Btw, did you find how many amps or amp increments that thos pins run at? I don't wanna blow the circuit overloading on amps.

Nice finds there ptr!

ptr.exe
July 2nd, 2005, 21:43
I'll check amps probably tommorow.

funny thing is when you told me it was a 1.8V fed to the SH-4 i decided to waste some time and check the various ICs on the mobo for something matching/near that, and IC 104 was one of the ones i thought may be it, also a Sega chip with alot of pins next to the SH-4 makes 1.74V on a couple of the pins but that's probably nothing.

Morph
July 2nd, 2005, 23:49
are you referring to IC501? If so, then thats the DC BIOS. And if that is indeed where the DC gets its 1.8v* from, then that still shouldnt be too hard to intercept. And it makes sense too, if it is coming from the BIOS chip, because that means the the voltage is BIOS controlled. But a few wires and a switch can change all of that.

Btw, is this it? (threw this together real quick in Paint)

http://img.photobucket.com/albums/v176/Morph777/DC_mobo.png

If so, which pins? Fi you wanna tell me, you can just edit this pic and mark the ones red or something, so I know. You dont have to tell e the pin # etc. :D

And those revision numbers, I have no idea on their accuracy, since those are only the ones I have at my disposal (from my Rev: 0 board and one of my Rev: 1 boards) and could in fact vary by region etc.

Cross
July 3rd, 2005, 07:52
Nah, the SH4 isnt underclocked that i can tell. The official SH4 product thingamajigger from Hitachi says its 200MHz. Odds are, either the DC version is a low power version (seeing how is hard coded low-endian) or perhaps it gets more juice to keep it more stable. I would doubt the assertion of more juice, since that would have increased the amount of heat, then again, that explains how the DC can hit a nearly 40% overclock without a voltage increase, something impossible on any other processor I have ever seen.

So, odds are, its probably the 1.99v lead, unless there is still another circuit on the DC motherboard that we don't know about. But I doubt that since I am almost certain that the 3.24v leads are for the I/O voltage. Btw, did you find how many amps or amp increments that thos pins run at? I don't wanna blow the circuit overloading on amps.

Nice finds there ptr!

My first thought as well with how well it pushes forward without a voltage increase 1.99 seems to be the most likely.

quzar
July 3rd, 2005, 12:21
morph: He said IC 104 because he is talking about IC 104, it's a small chip (looks almost like a complex voltage regulator) on the back of the board almost directly under CE106.

On VA0 boards it is a Sharp 07VZ1H chip and on VA1 boards it's a Sharp 070XZ1H chip. If anyone cares, it is also a 070XZ1H on VA2.1 boards, but it has been moved to directly under the power supply pins.

I checked that on 2~3 motherboards a piece except the 2.1 (since i only have one).

ptr.exe
July 3rd, 2005, 16:58
that's correct quzar, but in my post i started talking about another chip that generated 1.74V, thats what morph is talking about, also i think it looks like a voltage regulator because it is :p As for VA2 boards that doesnt matter as they dont OC well at all... just thought of something, maybe the reason they wont OC is because there isnt the massive excess of V that all the other revs have considering they have a differant chip in another place.
quzar could you check IC 104 on a rev 2 and post the V's, if the one on the others that has 1.99V is lower on the rev2's then its a good bet that that is the chip 'feeding' the SH-4 the vcore.

morph: No its not the bios, i know where that is, on the diagram you drew, go up to where the PVR2 is and above that theres a small chip that says 'Sega 315-6258', sorry when i said near the SH-4 i meant the PVR2. Probably isnt that. As for the bios, i thought it would be that so checked that first but found nothing near the 1.8.

tommorow i have a day off college, i'll check every pin again and post a pic showing all the points that are near the 1.8, their V and current. If i remember there are only about 5 excluding the IC 104.

However i really doubt its any of them, most likely the IC 104.

Morph
July 3rd, 2005, 19:59
Ok, I know what chip you are talking about, IC401. Thats on the Rev: 1 board and I believe that might be the Dreamcast's ARM7 sound processor. On a Rev: 0 board it is marked 315-6145.

And if you would make that diagram ptr, I'd be much abliged. But you are probably correct about the 1.99v, since that would be most logical.

If we succeed in this, I will make a manual on how to go about all of this, and put it on my site. I'll submit it to Devcast if they'd like, and I can probably rewrite and polish up the old overclocking guide. Since this is a little more than overclocking the DC, I'm gonna affectionately dub what we are doing powerclocking (overclocking + power increase) :p.

ptr.exe
July 3rd, 2005, 20:29
Yeah, 1.99 seems most likely, there doesnt seem much point in the diagram as majority of 'likely candidates' were quite below the 1.8V, such as 1.24V and 1.5V. I'll still check over tommorow and if i find anything new thats near 1.8 i'll post it, also i'll post the current as you asked previously. Just wondering; do you not have a multimeter? cant believe your doing this without one, they cost like £5.

Later on.

Morph
July 3rd, 2005, 20:35
Ha, I couldnt possibly afford one at least until next Friday, as I am flat broke. Paycheck was WAY smaller than I had thought it would be. However I'll probably get one at the same time I get a 50MHz CO, which are actually sold nearby.

Once the OC is working, I'm gonna do some benchmarks vs a standard 200MHz DC and a standard OD'd 240MHz DC. I was thinking that, along with the Delicious Hardware Demo, I could also run Super Mario War, Quake 3 Arena, and Unreal Tournament all with FPS counter on and set on custom bot-matches to tax the hardware.

Sound good?

ptr.exe
July 3rd, 2005, 20:40
Do you live in the UK? as i can not find any C/O over 25Mhz anywhere. Really annoying.

Morph
July 3rd, 2005, 21:04
Nah, I'm based in Michigan. But that sucks. I just happen to live down the road from a privately owned electronics store who has stuff that you couldn't even hope to find at like Radio Shack or the like. A little more pricey, but, eh, never had no trouble with um.

ptr.exe
July 4th, 2005, 12:12
Sorry only just saw the secind bit of the other post, yeah that should be a good test, but dont forget to run DSNES with something demanding, a certain []Enix game should do, and see if it can get above 80%.

I'll check amps in a minute, i just woke up and am dazed and confused...

ptr.exe
July 4th, 2005, 12:48
Okay, pin 3 of IC 104 that puts out 1.99V has a current of 0.23A. However when i first turned the DC on and checked the current fluxuated alot around 0.3 it was going between .2 and .4 until after about 30secs it settled at 0.3A, so i turned the DC off and on to check again and it was at 0.23 i turned off and on and left for a few seconds and it was still at 0.23A. Perhaps that fluxuation was due to the caps filling up, not sure. Might wanna check into that.

As for the new vcore, what are you going to use to generate the new V? you'll probably want something variable to use at first when your deciding how much of an increase is required. But it needs to be very stable, small fluxuations in V could destroy the SH-4.

Morph
July 4th, 2005, 17:11
Correct. the current could kill it, but the SH4 specs sheet labels it as capable up to 2.4v with a 2.5v limit. So I'll probably throw together a small circuit that throws out those specific volts at that specific current. Shouldn't be too hard, I can power it from the Dreamcast PSU like any other mod and have the circuit break down the voltage and the amps.

But because the pin seems to fluctuate a bit, what current would you suggest? Its gonna be hit or miss either way, so there really is no wrong answer :p. Yeah, and as a bonus test, I might run Star Ocean thru DreamSNES with the sound enabled and see how it runs.

ptr.exe
July 4th, 2005, 17:27
Better to be lower than too high, use 0.23A but ensure that 0.23 is the calculated value not the value from a multimeter, as you know a multimeter will always give a value lower than a true value as connecting it reduces resistance in a circuit. So 0.23A is probably slightly lower so thats a good starting point. If it doesnt work simply replace a resistor with a lower value to give a slightly higher current and repeat until you get the thing to boot. Simple :D

Star Ocean would be a good test, as it usually just judders along at 55-60% with sound. Make sure you switch the little speed readout thing on though :)

Morph
July 4th, 2005, 19:46
Does the current eventaully level off? Or does it continue to fluctuate?

ptr.exe
July 7th, 2005, 18:24
Sorry took so long to reply, the current levels off and stays at 0.23A without any fluxuations. It fluxuates for only about 20 seconds when you first turn the DC.

semicolo
July 7th, 2005, 19:02
do you desolder the pins, solder wires in between and tie you ampmeter to the wires to take the measures ???

ptr.exe
July 7th, 2005, 22:08
do you desolder the pins, solder wires in between and tie you ampmeter to the wires to take the measures ???
No. Why would i do that?

Cross
July 8th, 2005, 09:16
Sorry guys I did not disappear lol. Well I spoke to the guys today on the phone. Nothing yet, but they did say that they all enjoy playing Crazy Taxi.....

ptr.exe
July 9th, 2005, 15:51
Morph, any progress? have you tried this yet?

Morph
July 9th, 2005, 18:25
Havent had the opportunity. I must first get a new soldering iron tip, since my old one is worn all to hell. I will as soon as possible.

Btw, on a seperate note, I figured out why Revision 2 DC's won't play CD-R's and probably why they dont overclock.

ptr.exe
July 9th, 2005, 20:48
Care to mention why?

Morph
July 10th, 2005, 15:01
The entire reason why CD-R backups worked on the DC's was because the Sega programmers had put a "back door" sequence in the bios so that they could boot from CD-R's and develop games and such faster without wasting valuable GD-Roms. No one knew about it for a long time, and even Sega forgot. By the time that hackers had learned of how to take advantage of the sequence, the Revision: 1 models had been out for some time.

So odds are, Sega replaced the bios in its Revision: 2 with a more "airtight" version that won't accept backups, like the Dreamcast was originally intended. That could also explain its inability to overclock.

ptr.exe
July 10th, 2005, 15:37
Seems feasible but unlikely as the whole reason, the bios will cause the DC to not boot CD-R's but it doesnt really explain why the A02 DC's cant OC.

I posted before a theory and i think that its likely, and could be proved/disproved easily if i had a rev2 mobo.

The chip that (possibly) generates the SH-4's power - IC 104. It's in a differant place on a A02 and is a differant type of V. regulator. You said before that the DC OC's well because of the high vcore of 1.99V so perhaps on rev2 boards the new IC only generates the required V. for the SH-4 to run at 200Mhz rather than the excess the rev1 and 0 produce.

Its possible but again unlikely, as why would they change it?

Morph
July 13th, 2005, 03:11
Sorry about being all gone and stuff. For some reason, the dcemu forums won't load in IE anymore on my PC. But anyways. what you suggest sounds plausable. Perhaps Sega used cheaper parts/configuartions for the Rev2 making it incompatible with our prior OCing knowledge.

Cross
July 14th, 2005, 02:31
Its nice to be back, dispite the debacle in the PSP EMU forums. I have not gotten any word from those I asked for help, lol good thing the DC's not worth much or I would really be worried. I hope to hear from them before its to late and someone else does it, lol it would seem Morph and PTR (and their fast and excellent work!!) have beaten me to the punch.

Hopefully either way we will have the answer we seek soon.

ptr.exe
July 14th, 2005, 16:58
lol it would seem Morph (and his fast and excellent work!!) have beaten me to the punch.
Morph hasnt actually done anything, lol.

Morph you probably need to buy a multimeter and a soldering iron tip, doubt you'll get very far without them :p

Morph
July 14th, 2005, 19:24
Well, just a soldering tip. You covered the whole voltage part, so now I can procrastinate in getting a multimeter sommore. And as for my "hard work", shoot, I only did a little research and asked questions, ptr and thems guys from dcdev did the work, I just out 2+2 together and got fish.

waitaminute......that isnt....

Cross
July 14th, 2005, 19:47
Corrected, my mistake did not mean to leave u out ptr, lol. Fish sounds right Morph lol.

ptr.exe
July 14th, 2005, 21:25
fish? yukky. :p

No, no, i dont want any credit, if it werent for Morph i never would of thought about this. btw, ive figured a way of finding out if pin 3 on IC 104 is the SH-4 power supply, it involves no fancy equipment, so i'll give it a try tommorow.

Also morph, i have a DC that wont OC to 276 for some reason so i may do some research by building a circuit to produce 2.01 V with 0.23A, i have it designed just need to get some resisitors with a variance of 1% to ensure its accurate.

quzar
July 14th, 2005, 21:43
Seems feasible but unlikely as the whole reason, the bios will cause the DC to not boot CD-R's but it doesnt really explain why the A02 DC's cant OC.

I posted before a theory and i think that its likely, and could be proved/disproved easily if i had a rev2 mobo.

The chip that (possibly) generates the SH-4's power - IC 104. It's in a differant place on a A02 and is a differant type of V. regulator. You said before that the DC OC's well because of the high vcore of 1.99V so perhaps on rev2 boards the new IC only generates the required V. for the SH-4 to run at 200Mhz rather than the excess the rev1 and 0 produce.

Its possible but again unlikely, as why would they change it?

It is in a different place (makes sense, things have to be reorganized to add chips to the board) and is a slightly different markings, but they probably reflect a revision as opposed to it's voltage output. If anything, it provides higher voltage or more current due to the fact that the mobo itself needs more power than the external gd-rom (then again, it could be less =P)

ptr.exe
July 14th, 2005, 21:51
It was only a (wrong) theory :)

Why would it need more? that chip has nothing to do with the GD-ROM does it?

semicolo
July 16th, 2005, 19:58
Sorry it took so long, I lost the post (in fact I searched on dcemulation 8-).
So how do you measure Intensities ? Do you have some sort of inductancy ampmeter ? (sorry I don't know how it's called in english) But those just work for several amps, so how ?
(If you just measured between ground and the spot where you want to read intensity, I'm afraid your measures are wrong).

quzar
July 16th, 2005, 21:13
It was only a (wrong) theory :)

Why would it need more? that chip has nothing to do with the GD-ROM does it?
with the v2 the gdrom controller was biuilt into the dreamcast motherboard itself, so it is possible that there needed to be different power levels going to different things (not that the cpu would need more power, but possibly more or less current going into the VRM which would require the use of a different one possibly...)

flaming_carrot
July 17th, 2005, 15:48
Sorry it took so long, I lost the post (in fact I searched on dcemulation 8-).
So how do you measure Intensities ? Do you have some sort of inductancy ampmeter ? (sorry I don't know how it's called in english) But those just work for several amps, so how ?
(If you just measured between ground and the spot where you want to read intensity, I'm afraid your measures are wrong).

Yes, you measure volts in parallel and current in series. But why are you people bothering with all that stuff anyway? Seems very cumbersome. Instead of going to all the trouble of building your own power supply, wouldn't it be much easier to just replace the controlling resistor with a higher rated one to up the voltage? No muss, no fuss :cool:

quzar
July 17th, 2005, 22:26
nobody implied replacing the powersupply... the current discussion was mostly about modifiying something near the vrm that takes power to the cpu (be it changing it for a different model, adding a pot, or adding a resistor)

flaming_carrot
July 17th, 2005, 23:08
Sorry but the discussion about current was all about building their own circuit to supply the CPU power, ie replacing/bypassing IC104. Which I find a little amusing since the goal seems to be raising the voltage, and that can be done by simply replacing the one resistor that controls the output of IC104. Seems a lot more efficient to me, but what do I know... :)

ptr.exe
July 18th, 2005, 14:29
you measure volts in parallel
Doh! how could i forget such a simple thing!?

To clarify, i read V. by using the ground coming from IC 104, which i assumed would be the ground fed to the SH-4 therefore the multimeter was in parallel. Perhaps im wrong?

Anyway, flaming_carrot, thats a good point but the problem is that if i was bothered to find a resistor that was in series with the output of pin 3 of IC 104 then i would have to replace it with something so close to the original value as the V. is only increasing so slightly (eg. V=IR 1.99/0.23= 8.65, so to increase that: V'= 2.01 so: 2.01/0.23= 8.74 - i know that these values for resistance arent true but for an example) i dont think i could find a resistor so close to a value as the ones used are gonna be nice round no.s such as 100 then im gonna need something like 101.7. The other way would be potential dividers but thats even messier on an already packed board.

Maybe i'm wrong and over complicating it, this is only GCSE physics, and me doing A2 i should know this stuff :(

semicolo
July 18th, 2005, 15:19
You're right for measuring the tension as ground are likely to be the same, but your intensity measure is likely to be bad if you measured in parallel like the tension, you short circuited the power supply (that's probably why you see fluctuation at first).
That's why I asked if you desolderd pins to take the measures.

ptr.exe
July 18th, 2005, 15:32
Ah, i see, really should think more about what i'm doing :p Thanks for the help Semicolo :)

oh, and in english, tension is voltage, and intensity is current.

Just hope Morph doesnt use those values and blow up his DC :D

flaming_carrot
July 18th, 2005, 17:04
Doh! how could i forget such a simple thing!?

To clarify, i read V. by using the ground coming from IC 104, which i assumed would be the ground fed to the SH-4 therefore the multimeter was in parallel. Perhaps im wrong?

Anyway, flaming_carrot, thats a good point but the problem is that if i was bothered to find a resistor that was in series with the output of pin 3 of IC 104 then i would have to replace it with something so close to the original value as the V. is only increasing so slightly (eg. V=IR 1.99/0.23= 8.65, so to increase that: V'= 2.01 so: 2.01/0.23= 8.74 - i know that these values for resistance arent true but for an example) i dont think i could find a resistor so close to a value as the ones used are gonna be nice round no.s such as 100 then im gonna need something like 101.7. The other way would be potential dividers but thats even messier on an already packed board.

Maybe i'm wrong and over complicating it, this is only GCSE physics, and me doing A2 i should know this stuff :(

No, you read the voltage properly, but like semicolo said, you need to connect the meter in series (ie desolder the pin and route the signal through your meter) to get a correct current reading.

And yeah, I think you are overcomplicating things. First off, a resistor after pin 3 (by the way, you have switched the numbering of the pins; pin1 is Vcc, pin 5 is ground) wouldn't work that way, it would limit the current, not raise the voltage. Here's a diagram I threw together quickly:


http://www.rootshell.be/~nocturne/vrm.png

I may have switched R133 and R134, I'm at work right now and this is from memory. Pin 4 is the Vref pin and it should have a voltage of 1.25V over R134. R134 is prolly 1kohm. R133 will control Vout according to:

Vout = Vref*(1 + R133/R134)

So R133 is prolly 600ohm now. If you replace this with a 619ohm resistor (use 0.1% tolerance) you will raise Vout to a little over 2.02V.

Unfortunately resistors only come in 600, 604, and 619 values so if you want values in between, solder two on top of each other in parallel fashion, eg use 1k580 and 1k to get circa 612 ohm, giving Vout ~= 2.015V (umm..uncertainty is really greater than that but whatever...)


Important notice! Caveat emptor! I take no responsibility for any damages that may result from this!

ptr.exe
July 19th, 2005, 16:47
Yeah, i mixed up the no's in the text and so thought it easier to change the diagram. don't know why.

You obviously know so much more than me its ridiculous. That method makes alot more sense than before. yeah i remember how to use resistors in parallel to have a lower value, so i'll have a go at that.

Thanks for all the help :)

Morph
July 20th, 2005, 02:46
Sorry about being gone. Been busy as all get out, and haven't had the opportunity to post. if everything goes well, I should be able to try this by this coming Monday. I see someone else posted with new data. So intercepting the lead via lifting a pin is not necessary? Because the diagram you drew up is all like...

EDIT: nvm, now it works

But yeah, I would like to hear about ur second DC ptr. Wonder if the voltage has to do with the speed or whatnot.

ptr.exe
July 20th, 2005, 12:53
Can't you see the pic?

Yeah, i dont know why we assumed the easiest method would be a completely new circuit for making the 2.01V increase, i may give this a go soon as im on holdiays soon :D

Morph
July 23rd, 2005, 16:01
Been too busy to give this much thought myself. been working on a different undiscussible project on the side, and really haven't had the chance to tinker with the hardware.

flaming_carrot
August 3rd, 2005, 15:17
No one tried this yet? I'm bored, I wanna hear some stories about SH4s bursting into flame, muhahahaha!

Anyway, R133 and R134 are correct in the diagram, so any bold ones can just replace R133 appropriately. It's located on the same side of the PCB as the SH4 and is pretty close to it.

ptr.exe
August 6th, 2005, 11:29
I'll try this as soon as i can find a cheap DC. At the moment i have one stock one that i don't wanna screw up as its mint, my MF DC which i don't wanna screw up as it took too long to build. I bought a DC the other day for £4 and got it working (F1 fix) and then sold it for £15, i'll find another cheap one then try this out.

flaming_carrot
August 9th, 2005, 09:24
Understandable. I was thinking more of the other guy who seemed so eager before :)

ptr.exe
August 9th, 2005, 10:47
Yeah, morph this was pretty much all your idea, since you hijacked the topic from cross, who is probably still patiently waiting for his original questions to be answered 8 pages on :p

Any luck yet? you must have bought a soldering iron by now :)

Morph
August 10th, 2005, 20:11
Well, yes and no. I have a new soldering tip now, yes. Problem being I have to find the time to do it. Hopefully this coming week will allow me a chance to have a go over it.

Morph
November 13th, 2005, 20:32
Hey, back again, and I've gotten some work done. Thru some very precise coaxing, I have gotten my DC to hit 288Mhz. I still have not done a voltage mod. I was going to try what flaming carrot said, but I cannot find R133 to do it, the bottom of the DC motherboard is a gaggle of the little buggers. I've even found C133, but not R133.

Can anyone help?

EDIT: OH! btw, do you think that only an extra .02 volts (2.02V) can hit 300Mhz? This is being asked because the SH-4 can run at 2.5V max, and it is at 2.0V now. I was actually shooting for 2.2V.

EDIT2: OHH! Another idea! Is there any way we could try the "pencil trick"? Heres a link to an example: http://www.pcstats.com/articleview.cfm?articleID=953 Seems to work fine in that situation, whos to say it wont in this one?

flaming_carrot
November 14th, 2005, 12:25
I think you are looking on the wrong side of the board. R133 and R134 are on the same side as the SH4 and PowerVR chips, and pretty close to the SH4.

Here's from memory to give a rough idea:

.---------.
| powervr |
`---------'
.-----.
| SH4 |
`-----'

R133
R134

semicolo
November 14th, 2005, 19:46
This pencil trick won't last, you can make some tests with it, but be sure to use real resistors when you're happy with the settings you found.

Morph
November 14th, 2005, 23:54
Ok, I found it. I should be able to get a 4B pencil tomarrow, and I'll give it a whirl then. And yeah, I am not shooting for anything permanent at the moment. I am simply attempting if it can be done.

ptr.exe
November 22nd, 2005, 17:24
Hey morph, i havent been on this board for ages, good to see you're still making progress.

What did it take to hit 288Mhz?

Morph
November 28th, 2005, 05:57
I bought a bulk load of 40-50Mhz crystal oscillators, and I had like 4 different ones rated for 48Mhz. So I tried them all, and one was kinda wierd. It would boot the opening DC intro audio, but no video. I reset it a good 4-5 times, and it ran fine from there. WIIIEEEERD.

Also, I cannot find a 4B pencil for the life of me.

ptr.exe
November 28th, 2005, 16:55
That is odd, because I have the same thing when trying to OC one of my DC's to 276Mhz, sometimes i can hear audio but no video, sometimes there is video but it locks up after a few seconds.

Would you like to sell some 44Mhz C/O? pm me if you do.

Cross
August 17th, 2006, 20:51
Wow this is AWESOME!!! I wish I could have ever come back and said we found out something great but all the guys did in the shop was take it apart, play with it and then have it out during a walkthrough and it all got thrown away.....

Anyway I would REALLY like if you have any extra crystal oscillators if I could buy one from ya. Def been missing my DC and I just picked up another at a yard sale... black to hehe :p

Morph
August 17th, 2006, 23:46
Well, talk about dusting off an old topic. Ok, let me see what i can refresh:

I bought a ton of C.O.'s in bulk, and was willing to just hand them out. I knew people needed them. Well, low and behold, my temporary PC finally gives out. Then my next temp gives out to.

I've been without a decent PC this whole time (this thing is NOT decent). I sold off a lot of unecessarry stuff to get some new parts, and unfortunately, I got rid of my C.O.'s in the process.

My DevDreamcast turned into a flop. The custom GD cable snapped, and was pretty much a lost cause. I tried to borrow a 4B pencil from a friend of ours, but they didn't have one. I was not able to find one. Then, while working on my main DC, we got hit by a CRAZY power-surge, and my new test DC's power supply lit up like the 4th of July. Fire extinguisher required.

Well, I kind of gave this project the boot after that. I am satisfied with 288, and maybe someone else can do 300. I MIGHT try again in the future (picked up some REALLY low profile 486 heatsinks, that can fit under the heat-plate) but it won't be any time soon. I REALLY have too much I am working on.

Cross
August 18th, 2006, 01:16
Yeah I know I had to go away for awhile, lol new family member!

Its to bad to hear about your DC. I would have been happy to send you the 4B I am sure the shop had a few laying around they would not have noticed for a few weeks :p

But alas I am no longer at that station I am DONE! WWHHOOWWW :p

Anyay I guess I overlooked how old it was I just kinda came back to see what happened. *Shrug* ADHD what can I say :p

but with how much Good Info this thread has I don't think it was completely evil to dust it off :p

Morph
August 22nd, 2006, 15:55
That's fine, ain't no harm to it.

I may go back to it again another day, but I'm within firing distance of accomplishing a couple big things for the DC community as it is. All things must be prioritized.

ßüboni¢ $oñic
August 28th, 2006, 17:36
uh what does this thing do to ur DC?

Morph
August 28th, 2006, 20:58
*Suggests reading the whole topic and Googling*

ptr.exe
August 31st, 2006, 19:42
Hey Morph, sorry to hear about the bad luck.

I too worked on this for a while but soon gave up out of frustration.

I've still got a stack of spare DCs and hopefully I'll try this again sometime, I'll hit 300Mhz one day :)

Morph
September 1st, 2006, 05:35
I read your PM, but my interwebs suck as of now, and am unsure about even the reliability of this post.

The custom GD cable. I took 2 old ATA33 IDE cables, and diced them up so I had 2 cables for even pinage for every interconnect on the mainboard and Rom GD connectors. I basically graverobbed from an old non-working DC, and took the male and female connectors from it's mainboard and GD-Rom respectively.

I tried all the soldering myself, but gave up out of a combination of frustration and a short attention span. I then sent it to a guy I know from another website. Amazing with a soldering iron, etc, I got him to work with it.

The result was a functional (but fragile, as I learned later) GD-Rom cable. Wasn't more than 7 inches long. He used some plastic stuff to kind of mold the ends so the wires remain out of contact with one another.

Worked fine, its a shame I snapped it. I was not about to pay to get it fixed, so, scratch that.

On the other hand, I am experimenting with another method. I am going for no records with this one, but it involves removing the spreader plate from the main metal board on early Rev1 DC's (to compensate for the loss of the Heatpipe Cooler), and replacing it with a couple of ULTRA low profile i486 heatsinks someone had tossed out with the proc's and all.

I'll pull off the bit of metal covering the back vent, and place a fan to blow air in. Perhaps that will work.

ptr.exe
October 17th, 2006, 03:04
Thanks for the info, my intention with the GD cable was to make space for some nice big heatsinks, probably some stock Socket A HSFs I've got lying around.

But I think improved cooling wont help much without the vcore increase.

quzar
October 17th, 2006, 03:09
you may be able to get more speed out of a rev2 dc. all rev2s (that I know of) have an sh 200R model processor (as opposed to 200) which according to rensas was a non-functional update, which usually means decreased energy use and decreased heat output.

Morph
October 26th, 2006, 22:56
you may be able to get more speed out of a rev2 dc. all rev2s (that I know of) have an sh 200R model processor (as opposed to 200) which according to rensas was a non-functional update, which usually means decreased energy use and decreased heat output.

Of course, that is still assuming that:

A) It is voltage related.
B) The Rev2 is undervolting when overclock is tried.
C) The voltage can be changed somehow.


I might be able to get my mits on a digital camera at some point or another, in which case I will take pictures at my progress. Again, not going to 300 (haven't bothered with voltage), but I am running @ 270Mhz rather nicely.

Also, my Rev1 DC sports an "R" CPU. And, as a side note, I am not necessarily certain that the CPU is the bottleneck. We have to take into consideration typical RAM speeds and limitations. If anything, the RAM is probably bottlenecking.

A Rev1 board I have (just laying out), sports 8ns (125Mhz) Low Density memory for the GPU and the Sound systems. Heres where the fun part comes in, the main system RAM is rated for 10ns (100Mhz). So, if any RAM is going to put a strain, it's going to be the system RAM (seeing it'd be a full 50% over-spec).

For reference, the RAM that is on the left side of the GPU/Northbridge and above the PSU connector is the GPU memory (4x 2MB chips), and the RAM that is on the right side of the GPU/Northbridge and next to the Ext Connector are the chips for the Sound system (1x 2MB chip). The system memory would one of the chips below the SH4, and one on the right side, both right next to it (2x 8MB chips).

Sega basically saved money, by getting larger, slower chips for system memory, and smaller faster chips for GPU and sound memory. So, if we are looking for a highly overclockable Dreamcast, suggest that people look for a Dreamcast with system memory above 10ns.


Using a GPU RAM chip for example:
http://img.photobucket.com/albums/v176/Morph777/rambig.jpg


Thats how you identify ns speed. To find total speed, take 1000, the divide it by the ns # (for example, 8), and you will get the maximum rated speed.

Get huntin!

kohan69
October 28th, 2006, 02:32
Good post Morph, I will now which of my 6 dreamcasts I can expect the best clocks from :D

Question- did anyone dare to change the clockspeed down or up WHILE A GAME WAS RUNNING ?! :D

Morph
October 30th, 2006, 04:39
Yes, ptr mentioned that he did, and he can no longer run at 200Mhz. "Hotswitching" is not reccomended, because you are halting and restarting the entire system when you do. The 33Mhz signal is multiplied by 3 to run the entire unit, not just CPU, GPU, and memory.

And remember kids, small "ns" numbers yield higher clockspeeds. Simple equation for SDRAM:

1000 / ns-speed = max rated speed

Example: 1000 / 8ns = 125Mhz

I can't remember where the whole 1000 thing came from, I assume it was from me cutting corners.

kohan69
October 30th, 2006, 08:28
Yes, ptr mentioned that he did, and he can no longer run at 200Mhz.


As in- if he turned off and swirched to either position, it will always tay overclocked!?




And remember kids, small "ns" numbers yield higher clockspeeds. Simple equation for SDRAM:

1000 / ns-speed = max rated speed

Example: 1000 / 8ns = 125Mhz

I can't remember where the whole 1000 thing came from, I assume it was from me cutting corners.

Are you referring to the memory timings?
PS; did anyone solder out and put in some better ram yet?

:o

quzar
October 30th, 2006, 10:16
You'd have to be a robot to be able to do that. The rom chip has fairly large legs, but the ram's are tiny tiny tiny.

kohan69
October 30th, 2006, 22:22
You'd have to be a robot to be able to do that. The rom chip has fairly large legs, but the ram's are tiny tiny tiny.

or have a friend that fixes laptop hardware for a living :p

Morph
October 30th, 2006, 23:02
You still need some mad skillz. I know someone from GameTZ who MIGHT be able to do it. He's one of the best solderers on the US, but even then it'd be hella hard. You'd probably have to get the same RAM from the same manufacturer, and hope it's faster.

Lord knows if the DC will accept "non-standard" RAM. Hell, for some ungodly reason it might NEED 10ns RAM (couldn't understand why, though). So far, I have seen NEC, and Hyundai memory as main memory. Both 10ns. Again, it'd be much easier if we can find one with faster memory.

I actually suggest opening up a Rev2, and checking the main memory. It's the newest, it might also be the fastest.

Morph
October 30th, 2006, 23:05
As in- if he turned off and swirched to either position, it will always tay overclocked!?


as in, it no longer runs at 200Mhz, the setting doesn't even work any more. You do realize backtracking through this topic can answer many of your questions, right?

Morph
November 2nd, 2006, 03:10
It's funny, I decided to look over my old DC motherboards, adn I found that my ODDBALL Rev0 has 9ns NEC memory for both GPU, Sound, and Main. Hm.

kohan69
November 2nd, 2006, 06:20
I read all previous posts :)

btw, talking about old posts:



Once the OC is working, I'm gonna do some benchmarks vs a standard 200MHz DC and a standard OD'd 240MHz DC. I was thinking that, along with the Delicious Hardware Demo, I could also run Super Mario War, Quake 3 Arena, and Unreal Tournament all with FPS counter on and set on custom bot-matches to tax the hardware.


Wouldn't timedemos be more accurate since bots produce random behavior?

Also- did someone try measuring the voltages (IC 104 ?) while the dreamcast was
ON idle (system menu)
heavy load <25fps (shenmue rain, snes emu)

and overclocked ON idle (system menu)

and overclocked heavy load <25fps (shenmue rain, snes emu)

That could lead to some thing ;)


and one another thing:
http://www.dcemu.co.uk/vbulletin/showpost.php?p=36258&postcount=21

What textviewer has 'fixed font view' ?

is this it?

http://img165.imageshack.us/img165/811/untitledji6.png







But why are you people bothering with all that stuff anyway? Seems very cumbersome. Instead of going to all the trouble of building your own power supply, wouldn't it be much easier to just replace the controlling resistor with a higher rated one to up the voltage? No muss, no fuss :cool:

I see his point, but wouldn't that increase the voltages to ALL components? May that have an unforeseen negative effect?
I doubt +0.05v could do any significant damage though.


~Peace Outside,
KoHaN :p

kohan69
November 7th, 2006, 04:11
It's funny, I decided to look over my old DC motherboards, adn I found that my ODDBALL Rev0 has 9ns NEC memory for both GPU, Sound, and Main. Hm.

Does it outperform the 10ns one? (benchies! :D)

Do you think 10ns + 9ns may be mixed, and will they bot run at 10ns?


I got some info on the voltmod - it is possible

the voltage is converted onboard from the 3.3v line
mod in the PSU's 3.3v line and u'll get more volts on the ~1.8v cpu line

Morph
November 9th, 2006, 07:03
I believe that that diagram came from a DCDEV member.

Also, I can try benchmarks, but I have yet to find a game with a frikking FPS counter. It is quite frustrating. Also voltage is not generate from the 3.3v line. DO NOT MESS WITH THAT! That voltage is used for the CPU I/O, and not as vcore.

Thats a sure way to blow a processor, mess with the I/O voltage. And the voltage is converted from (possibly), IC104, which generates the 3.3v, AND a 2v line (or was it 1.8v, can't remember), which might explain why the CPU can handle a 20% overclock as "standard" and not even flinch.

Also, I am not sure if that board even works. Its.....ODD.

Morph
November 28th, 2006, 22:44
Ya'll are gonna LOVE this.

I just purchased a Revision 0 Dreamcast. And it isn't just a Revision 0, it's a pre-launch DC. Manufacturing date is August 1999 (if everyone remembers, the N.A. DC release date was 9-9-99, and this console predates that by a month).

It has a heatpipe cooler. The SH4 is:


9E3, BP200, HD6417091R, Japan


Thats a revised SH4, something I've only seen from Revision 1 and 2's up til now.

But wait, it gets better. The GPU and sound memory are the usual Vanguard 8ns (125Mhz) chips. BUT! Get a load of the SYSTEM chips:


SEC Korea 928
KM432S2030CT-G8


8ns MEMORY!! This means the CPU and GPU memory have max clockrates before you're actually overclocking, of 125Mhz. Thats great news, because 133Mhz is just a stones throw away, and 150Mhz (300Mhz SH4), is only 25Mhz away for the both of them.

I HAVE to get a digital camera!

Morph
December 1st, 2006, 05:21
Wow, it's like a ghost town thread now. Everyone's lost interest... :(

Shinebi
December 1st, 2006, 11:57
I think you guys know far more than the average DCowner.... too technical.

I had mine OC'ed, but the chrystal could handle two frequencies, need to replace it, cuz the OC didn't work, the original signal was still ok.

Morph
February 26th, 2007, 21:08
Bump

I am thinking of using ThermoElectric Peltiers for cooling.

ptr.exe
March 10th, 2007, 02:31
Bump

I am thinking of using ThermoElectric Peltiers for cooling.
Sub zero temps ftw :D

You can get cheap 80w peltiers off eBay. Then just use an ATX PSU as a dedicated power supply for it, should easily power a 80w TEC off the 12V. You could get two, one for GPU aswell.

Make sure you take the proper precautions with a TEC though, they can be dangerous, and need adequate cooling on the hot side, a socket A HSF at the least.

I get the impression that it's not the SH4 thats holding the OC back, but some other component that is clocked from the system bus. And I think 300Mhz is impossible to hit without first figuring out what that is. As you said, the RAM is a possibility. You could transfer that 8ns RAM over to one of your other DCs that OC well. Google for the Xbox RAM upgrade to see how to solder RAM using a basic soldering iron.

Good luck with the TECs :D

btw, if you're wondering why I'm here, every now and again I check in the dcemu hardware section :o

Morph
March 21st, 2007, 02:40
Yeah, I'd rather shoot myself than do that sort of soldering ^_______^

Even if I could, I'd just pull some 5.5ns RAM off of an old Voodoo3, thats 183Mhz rated.

Also, I'll be running them on low low voltage. I just need it to cool better than the current solution.

ptr.exe
April 20th, 2007, 16:59
You'll be cooling the RAM?

Thing is, RAM is the same as northbridges etc. improved cooling does little for better clocks. As long as you're dissipating the generated heat (RAM sinks etc) then you wont see any improvement from going further.

A volt mod for the RAM would be interesting, but we'd need a volt mod for the SH-4 first, lol.

kohan69
April 21st, 2007, 08:57
Hey Morph- how did the 8ns ram do?

For cooling, check out:
http://xtremesystems.org/forums/forumdisplay.php?f=10

All your needs and guides.

However, I don't think that extreme cooling will make significant (10MHz+) difference unless the voltage is increased -any progress? :confused:

PS: the corrupt save problem due to asynchronism was solved by OCing the vmu :thumbup:

Tomlo
April 21st, 2007, 09:28
How do you OC a vmu?

btw nice to see ptr.exe is still alive <3

kohan69
April 21st, 2007, 10:31
How do you OC a vmu?



crystal

http://dcemulation.com/phpBB/viewtopic.php?t=60642&sid=c2184d4b70840b35e41946a34f17ada0

Morph
July 16th, 2007, 20:08
Sorry, I've kinda lost interest as of late. I've been busy with my computer hardware and the stuff related to that.

Shinebi
January 7th, 2009, 13:46
Maybe a stupid question: was there ever a benchmark app for checking if the OC is succesfull? (I thought there was one...)
And didn't DCEMU have a DL section? Been looking, but it's been a while since I visited.

EDIT:
Nevermind, i just got word on the DCEVO forum that there's one on the SAPPLAY disc and got another one elsewhere too...
Clock succeeded!