CPU upgrade x10

[blighty]

4 hours ago, Gangst3r said:

Hi,

I have my 1.4Ghz Xbox from N64 Freak and is very Similiar to this project. It works flawless with new built-in XboxHd+ mod but only one thing cannot give me complete happiness - switch to stock 733Mhz CPU on which I can play in normal speed some games Like e.g. Rallysport Challenge 2 or GTA San Andreas

I've tried all the patches available on the web but no luck - any ideas how to get it back to normal?

Play those games on a stock console instead.

[Bowlsnapper]

zz, where are you man? Any updates on this endeavor?

[zzattack]

Well, I finished my batch and moved on. Maybe stellar will revigorate some interest. The only one I've got left is spoken for but I might do another few when I can get my hands on some stellars.

[zzattack]

I recently got asked for some tips on this upgrade and figured others could benefit from the info as well.

So for posterity, here's what I recall as being the most relevant info, condensed.

 

There's 2 important topics to get right which require some strategy: alignment needs to be dead-on, and the interposer needs some help to be mounted flat. I'll explain why.

For alignment, generally you can be just under half a pitch off with your BGA alignment and it'll work out fine when the chip settles.  So, the BGA pitch on this motherboard of 1.25mm suggests alignment is very forgiving. However, vias centered within groups of pads are not nicely tented, and instead take up most of the leeway you'd normally have with alignment. I'd say you don't have ~0.6mm of tolerance, but rather about 0.15mm.

 

To assure I get good alignment, the first thing I do is eliminate the possibility of bridges forming between the BGA pads and nearby vias. You can simply tent these vias manually using UV curable solder mask. Takes about 10-15 minutes, and with this, a lot of alignment margin is recovered.

At the same time, I place some 0603 resistors near corners of the BGA array and interposer outline. See the red marked circles below. The BGA pads on the underside of the interposer do not cover the entire interposer area, so the interposer tends to tilt/slant during reflow because it's gravity center is off. When the interposer is tilted, BGA balls can get too compressed and are unnecessarily stressed. Now the 0603 resistors just happen to have a height very similar to properly compressed 0.76mm solder balls, so this very simple approach ensures the interposer will get mounted flat.

 

Next, my definitive way to get perfect alignments, after a number of less precise prior attempts: I drilled through the corner BGA pads on the underside of a sacrificial interposer. See circles in picture below. The resulting holes allows to see the xbox BGA pads underneath, and thereby align the sacrificial interposer perfectly. Then, I tape it down temporarily.

 

The next step is to fix this interposer in place. The idea is that by placing e.g. 0805 capacitors along the edges of this interposer, we can lift it up and replace it with a fresh interposer in exactly the same position. Furthermore, because the interposer cannot move laterally, the interposer won't "dance", and therefore balls cannot join with the vias centered within groups of pads either. With the alignment also being spot on, there shouldn't be a "snap" when the BGA aligns itself due to surface tension either. I believe this thoroughly eliminates everything that could potentially yield faulty mounts.

The very last optimization I added was to place a single case screw diagonally over the interposer as depicted above. This counteracts the weight imbalance due to the BGA pads being off-center underneath the interposer. The positioning I found is determined experimentally and results in the interposer settling down perfectly straight, i.e. all 4 corners go simultaneously. This isn't super important since we already guarantee flat mounts due to the 0603 resistors acting as standoffs from a prior step.

Running short on time for this writeup so I'll conclude with the only other vital step: use a bismuth compound solder paste for the CPU itself. Bismuth solder has much lower melting point than normal leaded, so this allows soldering of the CPU without inadvertently reflowing the interposer (but with a lot of added weight on top, potentially squashing the balls!).

[ibbanez]

well if you do a few more, put me down as interested. If you don't, no worries, but I did want to throw my hat into the ring, so to speak. 

[Bowlsnapper]

This is my favorite thread.

[Gangst3r]

Issues with some part of games which running too fast on my 1.4Ghz upgraded CPU Xbox is now Resolved! I bought Project Stellar and use Enhanced Game Compatibility without annoying patch!

I've already tested all incompatible games and all of them works fine!

[Bowlsnapper]

12 hours ago, Gangst3r said:

Issues with some part of games which running too fast on my 1.4Ghz upgraded CPU Xbox is now Resolved! I bought Project Stellar and use Enhanced Game Compatibility without annoying patch!

I've already tested all incompatible games and all of them works fine!

I'm editing my post due to new information. I'm glad this works for you.

How does it match up with the other features, though? I know that MakeMHZ is excluding a lot of parts of the scene in order to carve their own niche out, so I wanted to know how you feel about that and how it's working for you. Does it bother you at all or are you able to use stellar's feature set with zero annoyance?

[coldasijs]

@Bowlsnapper if you hadn’t flooded the stellar topic with your bitching about asshat Dustin, you would have seen this feature was released 6 weeks ago in the 1.10 release  

[Bowlsnapper]

9 hours ago, coldasijs said:

@Bowlsnapper if you hadn’t flooded the stellar topic with your bitching about asshat Dustin, you would have seen this feature was released 6 weeks ago in the 1.10 release  

If Dustin hadn't been acting like an asshat, I probably wouldn't have flooded it with that stuff. And I'm pretty sure you can tell by reading the thread that I'm not the only one that feels the way I do. Sorry if I didn't bone up on his friggin' features list: I'm honestly pissed at the guy because of his actions and I don't want to support his product because I feel morally opposed to it... Evidently, quite strongly. I'm just trying to listen to my conscience. If it were really that inaccurate, I doubt people would have liked my posts or responded positively to them and I probably would have faced much wider criticism for my opinions... Not that that's why I'm here in the first place; I myself found it interesting that ANYBODY agreed with me.

What would I do if I had a stellar? Would I buy it for a certain Xbox, (certainly not any of the ones that I regularly use) and then use it strictly for THAT purpose? How am I supposed to use that feature and then deal with all the other issues that he's caused because of his unwillingness to support anything? Or are you just saying I did so much bitching that I ignored that part of his change log and made a public ass of myself as a result... And that's pretty much your only criticism? For some reason, you felt the need to post about it sarcastically and publicly the way that you did, instead of just simply informing me of the feature so that I could be factually correct. That would have been decent of you.

But I'm sure Dustin is a WONDERFUL person just based on his technical achievements alone... And nothing else. And because of that he should be supported... Look I've already said the guy has done some majorly progressive things. I'll add that he should be proud of some of the things he's accomplished. He's just making it extremely difficult to impossible to appreciate because of all the other actions that he's committed... And it INDESCRIBABLY sucks and is hurtful... Yes I take it THAT personally. I've been in the scene since it started, and to see what's happening with MakeMHZ truly stresses me out and makes me sad.

Thanks for the update. By the way, this is one of my favorite threads. I do not want to pollute it with our bitching back and forth about this topic. (If that's what would even happen in the first place; I honestly don't know how you feel about any of this or my opinions. In my experience, you've always been a cool guy and somebody's been pleasant to exchange with. You've been here for a while and as someone that I actually respect, I would hate to start a dispute with you... Over anything. I think we're divided enough in this world, in many ways) I've done enough flaming in threads in my day. If you'd like to argue about any of the points I've just presented and why, please PM me or, actually, go ahead and start a new thread. In fact, I will tonight: "Your opinions of MakeMHZ". I think that would be a more appropriate place. People would be able to express their opinions and we would be able to offer criticism or support to each other based on what is said. I think that would be a lot healthier than expressing these opinions inside a bubble like PMs where no one is able to share how they view them.

And yeah this post is long. So the fuck what?  I fill up threads. It's what I do.

[coldasijs]

@Bowlsnapper sorry just yanking your chain 

 

[Bowlsnapper]

1 hour ago, coldasijs said:

@Bowlsnapper sorry just yanking your chain 

 

I'm too easy. Lol!

[Bowlsnapper]

Hey @zzattack, does the processor have to be 1.45V core voltage?

I have this processor: CPU SL5QJ, 1000/256/133/1.75V, and they're only 20 bucks right now, 15 in bulk, but if I can't set the core voltage correctly (maybe with the proper resistor) then it's not usable. Are you able to offer any insight? I imagine it would just be undervolted, but I don't know if that's too much. The processors tested on Lukas's github suggest that it should be fine, but I wanted to know if I should do anything differently than say the SL6BY?

@SS_Dave@KaosEngineer

[trencherfield]

23 hours ago, Bowlsnapper said:

Hey @zzattack, does the processor have to be 1.45V core voltage?

I have this processor: CPU SL5QJ, 1000/256/133/1.75V, and they're only 20 bucks right now, 15 in bulk, but if I can't set the core voltage correctly (maybe with the proper resistor) then it's not usable. Are you able to offer any insight? I imagine it would just be undervolted, but I don't know if that's too much. The processors tested on Lukas's github suggest that it should be fine, but I wanted to know if I should do anything differently than say the SL6BY?

@SS_Dave@KaosEngineer

You going to need to match up the CPU core voltage. You can measure it after changing the resistors in the guide, if needed. There has been some conjecture over the values needed regarding this. Depends on the processor of course too. Yours at 1.75v means it's probably already about that standard? Whatever the 733mhz standard cpu runs at. I've not checked the value yet.

https://github.com/N64-Freak/Xbox-CPU-interposer/issues/1

Think 'Computer Booter' on youtube mentioned about this too somewhere.

Probably best to measure the standard voltage and then depending on which board revision work out the resistor values you need, fit and meter read them until you get to 1.75v for your CPU.

 

[Bowlsnapper]

27 minutes ago, trencherfield said:

You going to need to match up the CPU core voltage. You can measure it after changing the resistors in the guide, if needed. There has been some conjecture over the values needed regarding this. Depends on the processor of course too. Yours at 1.75v means it's probably already about that standard? Whatever the 733mhz standard cpu runs at. I've not checked the value yet.

https://github.com/N64-Freak/Xbox-CPU-interposer/issues/1

Think 'Computer Booter' on youtube mentioned about this too somewhere.

Probably best to measure the standard voltage and then depending on which board revision work out the resistor values you need, fit and meter read them until you get to 1.75v for your CPU.

 

Aw trench, I have no idea how to calculate that because my electronics knowledge is unfortunately very limited. I think the standard voltage is 1.45, if I'm recalling correctly. That's what the good old SLB6Y is.

[trencherfield]

For a 1.2 to 1.4 board as Maximus64 said...

"I'm in process of upgrade my CPU and after replacing R2E8 to 1.5K, the output voltage still measure around ~1.55V instead of 1.45V that the CPU want. I think the resistor value feedback network mention the the manual is wrong.

SC2688 datasheet mention the reference voltage is 1.25V
So base on my calculation with 1.5K resistor:
(1.25v)(1.5K + 379)/(1.5K) = 1.566V

This is too much for the CPU since spec is 1.45V. If we want 1.45V output I think the resistor should be ~2.4K (2368 ohm to be exact)"

There's the equation in bold to work out to the SC2688 (look at the attached pdf as it provides the VCC_CORE).

Maybe email Lukas and ask? I have not looked into it any further myself yet. Depends on which board you are upgrading of course too, so need to know that.

SC2688.PDF

[Bowlsnapper]

8 minutes ago, trencherfield said:

For a 1.2 to 1.4 board as Maximus64 said...

"I'm in process of upgrade my CPU and after replacing R2E8 to 1.5K, the output voltage still measure around ~1.55V instead of 1.45V that the CPU want. I think the resistor value feedback network mention the the manual is wrong.

SC2688 datasheet mention the reference voltage is 1.25V
So base on my calculation with 1.5K resistor:
(1.25v)(1.5K + 379)/(1.5K) = 1.566V

This is too much for the CPU since spec is 1.45V. If we want 1.45V output I think the resistor should be ~2.4K (2368 ohm to be exact)"

There's the equation in bold to work out to the SC2688 (look at the attached pdf as it provides the VCC_CORE).

Maybe email Lukas and ask? I have not looked into it any further myself yet. Depends on which board you are upgrading of course too, so need to know that.

SC2688.PDF 240.59 kB · 0 downloads

I want to ask him, but the email provided on his github is giving me a subsystem return error, so I guess it's not active. I've messaged him on here and I hope he responds.

Thank you for the info.

[trencherfield]

I saw Maximus64 he has a video of his CPU upgrade on youtube. At one point (after his test motherboard) you will see him check his good board for the VCC_CORE with a multimeter and he got it spot on. This is plugged in with no cpu on the board.

The standard voltage must be higher as the resistor value is increased to drop the VCC_CORE down to 1.45v in this instance. For your 1.75v core cpu, my guess is the resistor value will either be as it is now or lower to increase the VCC_CORE up to 1.75v.

The equation is there, I haven't worked the equation for 1.75v though, plus I haven't checked what the standard VCC_CORE is either, so that's as much as I can offer unless someone else cares to chime in with the values for you.

Standard voltage is prolly on the net somewhere or you can measure it with a meter, again the points to measure are prolly on the net somewhere too.

[trencherfield]

https://youtu.be/Fq7TdN0crSY?feature=shared&t=2881

That's the video.

Also can see where he tested, makes sense I suppose. Presume he used R2.4k at R2E8 to get 1.45v

[trencherfield]

Test point.

[Bowlsnapper]

14 minutes ago, trencherfield said:

Test point.

Wow I'm in over my head... lol

Okay, thank you sir. I appreciate your time in answering my questions and researching this for me! All I have for the time being is hot air, so I think I would need to pick up and ACHI so that I can do this the right way. Manual heating scares me, unless there's a way to use my hot air without worry of components flying (kapton tape, probably) and heating properly and symmetrically/equally. I do have a good hot air machine: an Atten that Sweetdarkdestiny recommended.

Also, if I choose to go the BGA route and decide to go the BGA slot processor route, is there a better way to source them than evilbay? They are so god damn expensive! but at least then I wouldn't have to use the interposer if I'm learning how to BGA solder and it would cut out a LOT of potential for error...

[trencherfield]

8 hours ago, Bowlsnapper said:

Wow I'm in over my head... lol

Okay, thank you sir. I appreciate your time in answering my questions and researching this for me! All I have for the time being is hot air, so I think I would need to pick up and ACHI so that I can do this the right way. Manual heating scares me, unless there's a way to use my hot air without worry of components flying (kapton tape, probably) and heating properly and symmetrically/equally. I do have a good hot air machine: an Atten that Sweetdarkdestiny recommended.

Also, if I choose to go the BGA route and decide to go the BGA slot processor route, is there a better way to source them than evilbay? They are so god damn expensive! but at least then I wouldn't have to use the interposer if I'm learning how to BGA solder and it would cut out a LOT of potential for error...

Will definitely need a preheater. A hot air station will have trouble heating the board uniformly without thermal stress. The board will also sap away the heat without a preheater under it, making it harder to do the CPU.

I have the gear & more and been making the xeniums/ogx360 for years and other chips, plus other smd stuff besides, but even I have not done the CPU upgrade. Yet. It's a mammoth task, there's no blagging around it, it's going to take time (a lot) & patience (a lot). Watch the other videos, computer booter etc. Sometimes he is not successful, sometimes he is. Nature of the task.

I've picked up a few CPU's off ebay recently to test with in due course. It's about the best place to pick them up. You might also consider to have a P3/Tualatin PC test rig with a socket 370 board to check the CPU's first... no point soldering a duff CPU on despite what the listing states hey.

Everything you need for this job, including the parts, gets very expensive very quickly

[GoTeamScotch]

4 hours ago, trencherfield said:

Will definitely need a preheater. A hot air station will have trouble heating the board uniformly without thermal stress. The board will also sap away the heat without a preheater under it, making it harder to do the CPU.

I have the gear & more and been making the xeniums/ogx360 for years and other chips, plus other smd stuff besides, but even I have not done the CPU upgrade. Yet. It's a mammoth task, there's no blagging around it, it's going to take time (a lot) & patience (a lot). Watch the other videos, computer booter etc. Sometimes he is not successful, sometimes he is. Nature of the task.

I've picked up a few CPU's off ebay recently to test with in due course. It's about the best place to pick them up. You might also consider to have a P3/Tualatin PC test rig with a socket 370 board to check the CPU's first... no point soldering a duff CPU on despite what the listing states hey.

Everything you need for this job, including the parts, gets very expensive very quickly

I used to do reflows/reballs for a living during the rrod era and even I don't wanna do this myself.

Definitely not a project for the faint of heart.

[Bowlsnapper]

5 hours ago, trencherfield said:

Will definitely need a preheater. A hot air station will have trouble heating the board uniformly without thermal stress. The board will also sap away the heat without a preheater under it, making it harder to do the CPU.

I have the gear & more and been making the xeniums/ogx360 for years and other chips, plus other smd stuff besides, but even I have not done the CPU upgrade. Yet. It's a mammoth task, there's no blagging around it, it's going to take time (a lot) & patience (a lot). Watch the other videos, computer booter etc. Sometimes he is not successful, sometimes he is. Nature of the task.

I've picked up a few CPU's off ebay recently to test with in due course. It's about the best place to pick them up. You might also consider to have a P3/Tualatin PC test rig with a socket 370 board to check the CPU's first... no point soldering a duff CPU on despite what the listing states hey.

Everything you need for this job, including the parts, gets very expensive very quickly

I assumed that uniform heating would be a barrier, with the limited experience that I have with it. It seemed like preheating was necessary... that makes me worry about other components on the board. Not so much the SMD stuff, but like electrolytic capacitors and things like that. This is why I wanna stick to slot processors at first, so I only have to do it once instead of twice at different temperatures and don't have to source the smd components, which I have not soldered yet. I do have the oscillators though and the hard to find stuff for the performance reduction circuit. I'm paranoid and sourced 10 of each because I'm like that. lol. I need to get an old Dell inspiron or something to test these in, absolutely. I have 3 now! 1ghz, 1.2 and 1.344, I think. They were only 20 bucks so I said, why not? Have to figure out the resistances though, since I believe the voltage of the mobo output is 1.45V.

 

25 minutes ago, GoTeamScotch said:

I used to do reflows/reballs for a living during the rrod era and even I don't wanna do this myself.

Definitely not a project for the faint of heart.

Would YOU feel more comfortable just doing the BGA processors to start?

[trencherfield]

Just had three 1GHZ 133mhz slot 1 Piii's delivered from usa. Look mint, have to see. Need to disable speedstep too of course. Will try the cheaper units first before these when I feel like it.

Don't think it's 1.45v standard Snappy.... else there would be no need to change the VCC_CORE resistor for the 1.4Ghz. I would guess around 1.7v same as these slot 1's for Coppermine.

I'll measure it now for you.