GGbuild freezing problem on jasper

[Dumpkin]

Hi guys,

I have got ggbuild running on jasper 512 on 13604 in most of the times console frezzes and sometimes mke a squeel sound at xbox 360 animation logo after. But sometimes though it boots properly.

Any one else getting the same prob ... any way to fix it.??

[Pacote-san]

I have the same problem o my 512mb -> 16mb jasper

I could upload some videos if anyone want to see exactly whats going on....

I i boot it cold, it always boot properly

[Dumpkin]

Pacote-san,

I have got 3 boxes with same issues all are jasper,

I think we should try rgloader on them.

and yes they boot perfectly when cold, The reason of freezing according to me is the glitch is still being botted ie cpld still sending the reset pulse.

[cory1492]

On my falcon, 1/5 times (using .ecc image) xell is incapable of bringing up video though I can see the xell chatter on UART. I think the glitch puts things in a somewhat random/undefined state... such is it's nature. Basically whats happening while the bootanim is running, is kernel is bringing up hardware (training memory, starting up the other 5 cores, initializing USB, that sort of thing) and interrupts - the trick is figuring out which hardware has the problem, detecting that problem early and letting the machine reset to re-glitch instead of going into kernel/xell.

Just to throw something else into the mix, due to component tolerances not every glitch board of the same make (or every console mobo of the same make) is equal. What works perfectly on one board may not work at all on another...

And no, the CPLD doesn't deliver the 'pulse' more than once per reset, having a LED hooked up to the DBG pin makes that pretty clear. It won't try to glitch it a second time unless the console resets itself (the CPU JTAG reset pin gets pulled low by the console.)

[Pacote-san]

On my falcon, 1/5 times (using .ecc image) xell is incapable of bringing up video though I can see the xell chatter on UART. I think the glitch puts things in a somewhat random/undefined state... such is it's nature. Basically whats happening while the bootanim is running, is kernel is bringing up hardware (training memory, starting up the other 5 cores, initializing USB, that sort of thing) and interrupts - the trick is figuring out which hardware has the problem, detecting that problem early and letting the machine reset to re-glitch instead of going into kernel/xell.

Just to throw something else into the mix, due to component tolerances not every glitch board of the same make (or every console mobo of the same make) is equal. What works perfectly on one board may not work at all on another...

And no, the CPLD doesn't deliver the 'pulse' more than once per reset, having a LED hooked up to the DBG pin makes that pretty clear. It won't try to glitch it a second time unless the console resets itself (the CPU JTAG reset pin gets pulled low by the console.)

That explains a lot... way before gbbuild was released i got this problem on this jasper...

Sometimes Xell would boot but only half the XELL ASCII would be draw on the screen and the freeze

And with gbbuild there was one time that the console booted fine, freestyle loaded and when i tried to choose a game it froze

[ZerOneX]

those reports around the globe are very very weird!!!

I did a few jasper (12 in total) and 3 of them refuses to glitch at all

Today I did my first Falcon and guess what?!! It was the best one... far far way from the others.

It boots in 7 seconds at maximum, I did not see a single freeze till now..... who knows why this is happening... but one thing is sure... the real cause is still unknown!!!! because it not chip, solder, wire (size, thick, thin, etc..) problem!!! it is something else.

[Pacote-san]

those reports around the globe are very very weird!!!

I did a few jasper (12 in total) and 3 of them refuses to glitch at all

Today I did my first Falcon and guess what?!! It was the best one... far far way from the others.

It boots in 7 seconds at maximum, I did not see a single freeze till now..... who knows why this is happening... but one thing is sure... the real cause is still unknown!!!! because it not chip, solder, wire (size, thick, thin, etc..) problem!!! it is something else.

Yep and just to add, i encountered 3 jaspers (1x CB6751 and 2x CB675) that refused to glitch aswell... =/

All the time and work invested for nothing... really makes me just want to start only receiving slims for the glitch....

[cory1492]

Something you might try on a fat that is showing these problems more often than not is +/- a tick or two on the WIDTH_RESET_START variable in the VHDL. It's a little frustrating to experiment with the VHDL by hand like that, but maybe it will pay off for a particular console...

[rf1911]

I can report the issue as well on jasper 256. Im a bit busy atm, if anyone can modify the thing and share the jed that would be great!

[Pacote-san]

Aside from that, just got a slim today that just wont glitch... tried everything you can possible imagine...

took the coolrunner, installed on another slim and glitched just fine...

This one i coudlnt even get xell running... what a pity

[Dumpkin]

same here got another jasper 512 which will not even glitch.... but i can confirm that it does glitch since while flashing to original it boot very late with diffrent timings and sometime do go rrod when too late on original nand.... not able to find the prob

[Pacote-san]

And funny thing... i got a slim that WILL ONLY GLITCH HOT LOL

if i let it cool down and try to glitch, not a chance... let it glitching while taking a 10 min shower, no dice

Took the power cable off, waited 10 seconds, put it back, turned it on, glitch in less then 20 seconds... lol

The glitching on different on different consoles is too unstable =/


But i got a falcon today that i did 10 boots in a row, all of them less then 5 seconds... amazing! Zeronex got a falcon as fast as this aswell

[xboxbreaker]

And funny thing... i got a slim that WILL ONLY GLITCH HOT LOL

Glitching as a method is not 100% reliable, especially when the code testing that was successful on one console is being expected to work across the millions of consoles out there.

In time I think there will either be more successful glitch timings found that will work across the board (I'm sure people are working on it), CPLD code that will try some random timings within a range, or code that will do some profiling of the glitch on first attempt, rather than repeatedly punt the same timings over and over again.

Some of you should try experimenting with adjusting the times as cory suggested, then compare notes, might yield a useful pattern.

It is early days friends! :-)

EDIT: Regarding the original issue of freezing, is that not perhaps because the reset pulse is still being sent? Maybe the CPLD is not picking up the post codes reliably on certain consoles. Just a guess.

[cory1492]

Unless there is a bug I'm not seeing, it takes more than the POST lead to cause a second reset pulse (specifically, reset being drawn low by the console instead of the CPLD.) Definitely not impossible though.

[Blackaddr]

It takes 4-5 seconds for ONE attempt.  If it boots in about 5 seconds, it got it on the first try.  Everyone seems to have no trouble glitching a Falcon, but I haven't seen any report profiling their console them self for a Zephyr, Jasper or Slim.

I profiled my Falcon and found the console glitched successfully 100% of the time across FOUR sequential offset values, while keeping a constant temperature of about 30 C for consistency (big ass fan on it).  once you were +/- 3 from the central window, it dropped to about a 1 in 5 chance, one more tick away and it pretty much didn't glitch.  That is a VERY small window and that's on an 'easy-glitching-falcon'.

All you guys need to do is take the VHDL file, change the value as Cory suggested, rebuild with ISE and reprogram with impact.  It's worth taking the time learn this.  You will only have to try +/- a few ticks to find the optimal value for YOUR console, so it's not a lot of effort.  You might not find a better value than the one you are already starting with, but you won't know unless you try.

[Pacote-san]

The video of the falcon i did today.... really surprising

http://www.youtube.com/watch?v=cnMd9xvSrsk

[ddsdavey]

Yep,did a falcon yesterday,boots like an angel.Jasper bb16 today just crashes at splash/dash,took about ten attempts at xell to scroll to cpu key etc. Jasper 256 other day ticks over in blackness!
Its so easy to sound like im complaining when i really am not,its a superb find none the less!

[playonlcd]

All you guys need to do is take the VHDL file, change the value as Cory suggested, rebuild with ISE and reprogram with impact.  It's worth taking the time learn this.  You will only have to try +/- a few ticks to find the optimal value for YOUR console, so it's not a lot of effort.  You might not find a better value than the one you are already starting with, but you won't know unless you try.

As nandpro v3 is ofering support for POST codes it would be nice to see around a tutorial to calculate the right timings for each console after POST codes. This way you are boting in no time.

More of this it seems that it only needs LPC 4148 and the CPLD board, witch any of us are using it.

[joelbb]

I also have a jasper 256 that freezes in ggbuild sometimes in freestyledash and maybe couple minutes into games also then sometimes 20mins into a game. It has no problems glitching and booting to freestyle dash though glitches within 30 secs most times. This has me tearing my hair out.

I also tried using a config.bin and smc.bin from a 16mb jasper and the console booted but the video was stuttering and until it froze up. which sort of makes me think it has some thing to do with the config and smc files.

I did have a 16mb jasper that wouldn't glitch at all until i used a donor smc.bin file and it glitches straight away and doesnt freeze at all.

What i would like to try is a donor 256 jasper that is working perfectly and encrypt with my cpu for this console to see if it solves the freezing problem or if it is hardware related.

So if anyone would like to upload a 256 jasper.bin file that works would be awesome and much appreciated unencrypted or with a cpu key. Cheers
 
update:  just worked out why ggbuild freezing i had a poor connection on stby clock resoldered and hasnt freezed up at all now

Damn freezing again butnot as bad though