Print

[PlaStiCam]

Lockedfroze after around 7 minutes of gameplay

EDIT: mod is now dead when rebooting after lock up (black screen, no sound etc.)

[Xconsole]

Ouch ... that sucks ...

There is a thread about the new BIOS in the BIOS-area of the forum.

I tried it on my old 1.6 (Xenium Ice), and it didn't appear to have any problems, but then again, it's an old 1.6 that always worked well with the "old" M8_16.

[nachomans]

Could someone post a picture of the hinix ram chips??
There is slightly problem with xbox with samsung too...maybe there are new samsung chips..too...

So it could be called 1.6HR (hinix ram) instead of 1.6B..

[PlaStiCam]

Here is the offending ram:

[pimpkid]

Looks like the ram theory might not hold up.  *edit*..i just now read the post right above this regarding samsung chips ;p .  In any event, here are my findings.

see snip from my post over at yoshihiro's thread

QUOTE

Just to add my findings: I have a 1.6 dated 9.21.2004 Phillips dvd drive seagate hd (skinny white one) yellow dvd ribbon ** **samsung memory chips** modded w/ a cheapmod/m8 (m8.67 ). threw a seagate 200gb hd in, formatted for f+g, threw some apps on there, all normal noticed when copying games either via qwix OR xbox>xbox, transfer would die after about 1gb of data had gone. popped 200gb drive in another xbox, was able to xfer some games onto it. popped 200gb back in 1.6(b?) fired up a game, let it run for about 10 mins, then freeze. now i reflashed bios to beta4, reformatted hd (this time it puts extra space on /f/) things seem to be working, i'll post again if that isn't true. but: so far as i can tell, the samsung/hynix theory is thrown out the window, i had samsung mem chips and got the same symptoms as those who had hynix.

word of note: the beta4 bios that Yoshihiro made seems to be working for me.  Xbox has been powered on and running for almost 2 hrs and no problems whatsoever.

[Xconsole]

Well, as I wrote in my reply to that, Samsung probably makes more then one kind of memory.

They no doubt have different qualities and speeds on them, and MS just decided they didn't need the really good ones anymore

To be certain, we should be comparing the exact writing on the "bugging" Samsung rams to the "old and stable" Samsung rams.

My old (may or so) 1.6, that always worked just fine with the old M8_16 have Samsung chips that says:

SAMSUNG 416
K4D263238F-QC50
WTC086PB KOREA

Some of it is probably things like manufacturing year, week, date, batch and so, but still.

Anyway.
What has been proved is that the M8_16 was missing RAM-timings included in the 1.6-retail-kernel, and also, the Beta4 of M8 with the full table inserted appears to be working.

Therefore, we can with a lot of certainity blame the RAM-chips for these problems, and everything about faulty "1.6B" och "1.7" machines.

Well, actually, we can't really blame the RAM either, as the 1.6-machine is designed to use different kinds of RAM, and the new kernel was written to handle them.

Problem is that that feature was not included in the hacked BIOSes.

Anyway, all is good now. We can be quite confident that all 1.6:es will work with a correct BIOS, Yoshis fixed M8, or the soon to be released X2 and X3.

It's still impressive to see how quickly this rather complex and, to begin with, strange, matter got cleared up.

[Vrtclhykr]

Just want to ad my FYI:
Crystal LE (Canada) 2004-09-07
Philips DVD........now switched with Sammy
Brown DVD Power
Samsung Ram

Not sure if this is a related issue.
All games tested seem to run with no lockup.......BUT....a certain game . (PAL English version)
With my network plugged into the xbox this games freezes when you press Start.
Without network plugged in there is no problem.

[cap269]

Here's one I came across today to add to this growing list:

XBOX 1.6
S/N: xxxxx36 43205
Mfg Date 8-4-2004
"Samsung 425" memory chips (K4D263238F-QC50)
Yellow P/E cable
Yellow DVD cable
Seagate HD
Toshiba/Samsung DVD
Xenium ICE

Xenium OS works fine. Original bios (5838) works fine both from the onboard chip and when flashed to the modchip. When EvoX M8_16 is launched, the system hangs right before the flubber (screen is black). When Yoshihiro's bios is flashed, it boots perfectly. The M8_16 bios used is the same one I have used with every single 1.6 mod I've performed (at least 50), so I know the M8 bios is good.

[feflicker]

What I don't understand is why Flashbios doesn't display video... (It is Cromwell based...).

I don't see how the RAM could cause that, but apparently Cromwell had the same problem as the EvoX M8_16?

[A@ron]

the xbox shares the ram between the video-GPU system and normal memory... so therefore if the ram tables aren't right you won't get proper/any video.

[feflicker]

QUOTE

the xbox shares the ram between the video-GPU system and normal memory... so therefore if the ram tables aren't right you won't get proper/any video.

So cromwell had the same problem then, the RAM tables? Interesting...

But wait, there is nothing to see here, just another ploy by X-Ecuter to cover up their crappy chips    

[SantMat]

 

[nachomans]

2004-08-13
Serial   06
philips
mem samsung
Made in china

Chips boot to black screen...

[John Hoek]

==============
The only thing related to RAM (timings/speeds etc.) that COULD explain those vague problems is that the ram on some Xbox V1.6 is UNDERVOLTAGED!

It is known (ATI for instance) that some manufacturers even forces to undervoltage RAM to prevent overclocking, so that just only plain official speeds can be archived.
In the case of Xbox, this is not an issue. But it IS A POSSIBILITY THAT SOME XBOXES V1.6 HAVE VOLTAGE ISSUES on the RAM.  It chould be 3,15Volts at very minimum, better 3,3 -3,4Volts. (normal voltage on SDR Rams)

Anybody who can measure this problem and to ged rid of this theory!  

[John Hoek]

How on earth can a program find out which type/brand of RAM you have in you're xbox? only the RAM configuration could be checked; 8x32bits, 4x64bits etc. nothing more.... Thei're is NO SPD eeprom inside the xbox with timing valua's.
And second, the RAM configuration could yeh be different between vendors, IF the motherboard accepts this. And with my knowledge; ALL xbox RAMs just have the same config.

The only difference can be the speeds.  The Hynix RAMs in the new xboxes have all 5ns speeds.  This means that even the bad of the bad of the bad RAM's must take at least 166-200mhz clockspeeds at 3-3-3 timings.....

If you know that Xbox uses only 133Mhz clockspeeds and 3-3-3 timingh; then just EVERY 5ns or 6ns model RAM must be working well. Even 6,5ns is rated for 133Mhz.... So basicly... the ram is way below thei're rated specs. Hense, even if MS wanted; just NO ram-manufacturer can sell lower speedgrades anymore. The yields are just to good nowadays!


But i've also seen lot's of post regarding booting problems, and powerunits (PSU's).
It IS PROVEN that the new powerunits are more sensitive and not as good as those used in V1.0 till V1.5.  MS has redesigned the motherboard, not only to prevent modders, but also for cost-reduction. Costreduction is also done on the PSU itself.  


Just use a multimetre and measure the real RAM voltages. If the PSU is faulty, or the voltages are just not stable enough; then yes, you get lockups/resets etc. at bootup and working state.

Don't forget that those modchips uses also EXTRA power from those PSU, and this extra powerconsumption just can be enough for troubles!  This explains directly also that when you TURN OFF the modchip; the xbox boots well to unmodded state; but when you TURN ON (and uses extra power from the PSU); the problems starting.

Just forget the fact that xbox uses different RAM timings and speeds. If they did, then every game just woirk different form each other xbox.... On the left one they work quick, on the second one they work slowly and the thirth they work normal. Only to have this idea is stupid!  


Anyone which measure the voltages inside the xbox, expecially 1,5Volts; 3,3Volts and 5Volts.  Are they within specs?!  I'm convinced; that on so-called V1.6b xboxes they aren't! Or they have to much dorps /variations in the voltage lines.(If this confirms to be true, thei're is the possibility to voltage modding of the PSU, just to put voltages a littlebit up to stable/correct valua's.)