Gp2x Voltage Increase?

[quadomatic]

I was wondering about the reason why the GP2X crashes after pushing to a certain overclock (for me its after 290 i think). Is it because the cpu isn't getting enough voltage for that clock? If so, would it be possible to voltmod the GP2X so that it gets enough power to be stable at higher clocks?

Just wondering, once again, don't hurt me.

Oh yeah, and I was reading another thread mentioning thermal pads being on MK1s and not on MK2s, but it wasn't known for sure (or maybe it was). Anyways, the MK2 does NOT have thermal pads...would putting a thermal pad increase oc'ing potential?

 

[christo930]

I was wondering about the reason why the GP2X crashes after pushing to a certain overclock (for me its after 290 i think). Is it because the cpu isn't getting enough voltage for that clock? If so, would it be possible to voltmod the GP2X so that it gets enough power to be stable at higher clocks?

Just wondering, once again, don't hurt me.

Oh yeah, and I was reading another thread mentioning thermal pads being on MK1s and not on MK2s, but it wasn't known for sure (or maybe it was). Anyways, the MK2 does NOT have thermal pads...would putting a thermal pad increase oc'ing potential?

I'm stuck at 240mhz(MK2, the value pack one with firmware 3), is there some sort of mod I can do to go faster (sort of like the old pencil trick for the GP32)?

Chris

 

[quadomatic]

I found this post in an older thread:

QUOTE
Because I'd like to be able to close my GP2X up when I am done. wink.gif As for ramping up vcore to the ARM920T, I've looked into it and the ARM920T has its own DISTINCT voltage line. It is pin VDDA920 on the MMSP2 and can be found in the datasheet. However, due to the lovely FPBGA package that prevents any of the pins from coming through the PCB, tracing down the line to that pin will be a lot of trial and error, or a process that would necessitate removing the MMSP2 from the machine, after which point I'd not really expect it to go back on very cooperatively..

At any rate, the ARM920T seems to take 1.8vdc by default, and its maximum rating is 1.95v. At 1.9v or 1.95v, I'd expect much more stability past 300 MHz .. Maybe my '2x would hit 320 with stability. Right now it only stands it for several seconds, I cannot even get through one run of 2xMark.


I looked at the MMSP-2 chip on my GP2X (its currently taken apart right now, I'm in the middle of a d-pad mod). I understand what he meant about the MMSP-2 chip. The pins sit underneath the chip. So, soldering to it would require soldering to the traces, which could be difficult. Still, it would be cool.

 

[___]

I am pretty sure that this has been proven to be impossible. I am not certain as to why but I recall a couple of threads discussing the possibility of raising the core voltage and some of the more knowledgeable people concluding that it wouldn't make a difference or was somewhat not possible.

Regarding the thermal pads: The ARM processor in the gp2x doesn't create any kind of heat that might lead to it becoming unstable. Cooling the processor is thus not necessary. Don't even waste your time with a mod likes this, it won't help you with overclocking.

 

[quadomatic]

Regarding the thermal pads: The ARM processor in the gp2x doesn't create any kind of heat that might lead to it becoming unstable. Cooling the processor is thus not necessary. Don't even waste your time with a mod likes this, it won't help you with overclocking.

That's what I thought (I always assumed the crashes were associated with instability that was not caused by temperatures, which is why I asked about a voltage mod instead of a cooling mod), but someone had posted in that same thread that they added a thermal pad from a dreamcast to the MMSP-2 chip, and they got a 15 or 20mhz increase in overclock.

 

[dpoarch]

If I remember right the problem with increasing the voltage was related to the power circuit. Apparently the way it is set up there is no easy way to up the voltage going into the gp2x with a simple trick or hack. I believe that on the f100s the ltsk chip takes the voltage input and puts out a constant 3 volts. Either I am wrong and the LTSK puts out 1.8 volts or there is another circuit involved steps the voltage down to 1.8 volts. Actually come to think of it there has to be. When the LTSK breaks the dc input plug still works. There must be some voltage divider somewhere that steps the voltage down and then some protection circuit to make sure it is a constant 1.8 volts. I havn't looked at the data sheet but sometimes micro controllers will run on a range of voltages and it will automatically set the core voltage at whatever the designer specifies using a sort of internal supply. If this is the case then it seems almost impossible to set the voltage with a simple mod and the voltage is either hard designed somewhere or controlled by some some pins or internal memory. Im sure some one knows more about it.

 

[christo930]

Regarding the thermal pads: The ARM processor in the gp2x doesn't create any kind of heat that might lead to it becoming unstable. Cooling the processor is thus not necessary. Don't even waste your time with a mod likes this, it won't help you with overclocking.

That's what I thought (I always assumed the crashes were associated with instability that was not caused by temperatures, which is why I asked about a voltage mod instead of a cooling mod), but someone had posted in that same thread that they added a thermal pad from a dreamcast to the MMSP-2 chip, and they got a 15 or 20mhz increase in overclock.


How are they able to keep the power consumption and heat down so low? Does it have to do with the lack of an fpu? I noticed that it wasn't until the introduction of the 486dx (which is the first x86 chip to include an on chip fpu) that pc's started needing cooling. I even have a 386 40mhz mhz mother board where the chip is plastic instead of the normal ceramic. Low transistor count? I did a quick google (I mean quick) and I didn't find any data sheet giving the transistor count (The arm 6 had 360k, so I assume the arm 9 has over a million). Power consumption is rated at 675mw (which is ultimately why it's so cool), but how did they accomplish that?

Chris

 

[ledow]

How are they able to keep the power consumption and heat down so low? Does it have to do with the lack of an fpu? I noticed that it wasn't until the introduction of the 486dx (which is the first x86 chip to include an on chip fpu) that pc's started needing cooling. I even have a 386 40mhz mhz mother board where the chip is plastic instead of the normal ceramic. Low transistor count? I did a quick google (I mean quick) and I didn't find any data sheet giving the transistor count (The arm 6 had 360k, so I assume the arm 9 has over a million). Power consumption is rated at 675mw (which is ultimately why it's so cool), but how did they accomplish that?

ARM precedes a lot of x86 chips entirely. They were originally Acorn RISC Machines, they designed the chips for a particular project, and then went through several name changes. The ARM chips were designed from the start to be small, cheap, low-power, Reduced Instruction Set (RISC), etc. Thus it's purely a design issue - ARM designed their processors to do as little as possible as quickly as possible and shut down unused areas of the chip - Intel/AMD etc. just went for sheer brute force to get to the top of the "MHz" and eventually "GHz" stakes, sod the power consumption. Of course, that meant poor design decisions that ultimately mean they are now fighting against their former selves to reduce power enough while remaining compatible with their top-end designs. FPU or not affects very little - it's just a case of not having lots of power running through lots of "wires" inside the chip whenever possible. RISC helps a little because your internal "wiring" is greatly simplified.

This was also part of the idea behind Transmeta's formation back in the day - a (potentially) x86 compatible processor that was completely designed to be power-efficient. ARM processors now run most mobile phones, there are thousands of them in things you see every day and they started off in this embedded field years ago. With x86, Intel were only really focused on getting faster ones into power-hungry desktops for business users. Intel can do embedded but they chose not to. Now that they've saturated all their new ideas and started to hit physical limits (e.g. heat), they are having to focus on reduced power-consumption, laptop & embedded use, dual-core etc. Meanwhile, they are fighting against £2 chips that are just as powerful if someone is willing to recompile their software to target them.