Print

[edbtz123]

Ok guys I looked into it into further detail, I was confused that 3.3 Reset Voltage should be coming from pin 6 of the transistor which is an output to it (not an input as I previously thought, i thought that it was connected to pin 5 of the transistor which is an input).  This makes a lot more sense now.

If pin 5 of the transistor (this is hooked up to the lower of the small two VIAs (holes) on the LPC) is 0 volts then pin 6 of the transistor (this is hooked up to pin 5 of the LPC) should be 0 Volts.  If pin 5 of the transistor is 3.3 Volts then pin 6 should be pulled high to 3.3 Volts (in a V1.6 board the resistor that pulls this voltage high has been depopulated, im not sure why).  If these votlages do not match your transistor is probably blown.

Disconnecting pin 6 of the transistor will pull the signal to 3.3 Volts like it should be, this may fix the problem.  This basically bypasses the transistor logic.

[ferrari_rulz_02]

QUOTE(hedbanger @ Nov 25 2005, 06:02 PM)

Someone said that the 3rd symbol on the transistor shows the week its been built.

i dont think so. markings are usyally just to tell what tehy are, not when they are made

[PimpleX]

QUOTE(edbtz123 @ Nov 25 2005, 07:56 PM)

Ok guys I looked into it into further detail, I was confused that 3.3 Reset Voltage should be coming from pin 6 of the transistor which is an output to it (not an input as I previously thought, i thought that it was connected to pin 5 of the transistor which is an input).  This makes a lot more sense now.

If pin 5 of the transistor (this is hooked up to the lower of the small two VIAs (holes) on the LPC) is 0 volts then pin 6 of the transistor (this is hooked up to pin 5 of the LPC) should be 0 Volts.  If pin 5 of the transistor is 3.3 Volts then pin 6 should be pulled high to 3.3 Volts (in a V1.6 board the resistor that pulls this voltage high has been depopulated, im not sure why).  If these votlages do not match your transistor is probably blown.

Disconnecting pin 6 of the transistor will pull the signal to 3.3 Volts like it should be, this may fix the problem.  This basically bypasses the transistor logic.

Well done, I can confirm that you are correct and lifting pin (the leg) 6 on the on the Q7R1 pulls volts on LPC pin 5 back to 3.3V (IMG:style_emoticons/default/smile.gif)

That just leaves the LPC pin 1 LCLK which is still 0.0v instead of 0.2v otherwise all the volts are good.

On a 1.6 FRAO board it now FRAGs after doing the pin 6 leg lift. I have several other mobos that are FRAG boards but with classic Q7R1 symptoms of 0.2v on LPC pin 5 and 0.0v on LPC pin 1 though, I tried one of them as well and it pulls 3.3v on pin 5 but still FRAGs with 0.0v on pin 1.

So now we need a way of correcting the LPC pin 1 volts WITHOUT replacing the Q7R1 and we are home and dry I think .......That being said, I do have a motherboard with all correct LPC volts and that still FRAGs LOL!!

Replacing the Q7R1 usually corrects both pin 1 and pin 5 volts together (I'll try that on the FRAO 1.6 mentioned above and report back if it fixed it as this is the only FRAO board I've got now to verify if the Q7R1 actually fixes the FRAO problem or not...as long as there are no other problems with the board that is of course.)

EDIT: ...I just pulled off the Q7R1 and tried it naked (not me, the motherboard without the Q7R1!) and it's the same ...3.3v on LPC pin 5 and 0.0v on LPC pin 1. You sure we can't just use 1 trannie on the 3 Q1 pads instead?

(IMG:http://www.meltdown.fsnet.co.uk/pics/XBOX/q7r1pins.jpg)

Cheers,

PimpleX

This post has been edited by PimpleX: Nov 26 2005, 01:37 PM

[PimpleX]

Ok no luck with swapping the dual trannie on this 1.6 as I'd missed some other problem and it was just blowing them on boot up I think.

I tried 2 used ones and 1 new one. If it isn't on right the board will FRAG because if pin 1 or 6 isn't making contact (doesn't matter which) then pin 5 gets 3.3v and pin 1 gets 0.0v (exactly the same as if it is removed entirely). If it's blown but all the pins are good then it will FRAO with 0.2v on LPC pin 5 and 0.0v on pin 1.

Hooking 0.2v up to pin 1 does nothing to help BTW if pin 5 has 3.3v as I tried that, even with a mod chip on it as it's some sort of logic signal probably comming FROM the MCPX and not a voltage required to go TO the MCPX. Pin 1 comes straight from the MCPX so that needs some more looking at.

Cheers,

PimpleX

[hedbanger]

i replaced the MA with two bc637 and it works fine now. when i previously installed other ma´s from a 1.0 board, it fraged. i think i burned them by removing them from the old board, although i was extremly careful.

[hippo]

The Lpc lads are tied high and the xbox drives them low to start the boot for one or more of the lclock (lpc pin 1) clock cycles. If the bios fails to load reset goes low, lclock shuts off due to reset going low and it tries the now famous 3 times and the final attempt it frags and reset is high, the lads are high and lclock is shut off at zero volts or .2 v if you take in to account the floating ground. Now most dead Xboxes I get have all the lines tied high but often lad1 wont drive low or is stuck low and never comes high, Ive seen it both ways so it must be a common open collector in the MCPX that blows and you can't repair. Also every time you push the DVD tray or the power button lrst has to drop to zero you can't just tie this high it's a control signal like Lframe, Lclck (33 mhz clock). That also means the lad lines are only driven low 1/33 millionth of a second, you'll need a scope to see it. Lclckand lrst and shared with the tsop loading and are the only signals that affect both. Also since it's active only between reset and oscillating between 0 and 3.3v 33 million times a second Lclck should read 1.6v breifly on a good meter before it starts fragging.

Most the dead ones I can't fix are a problem on lad 1.

This post has been edited by hippo: Nov 26 2005, 07:42 PM

[PimpleX]

Thanks Hippo, the description is helpful, so basically some will fix with a replaced dual trannie (or pair of singles) if that doesn't do it then it's the MCPX that's duff as originally thought and as such unfixable.

Hedbanger those fairchild BC637 may be easier to source so i'll look into getting some of them instead.... You got a photo or diagram of how you did yours?

Cheers,

PimpleX

[hedbanger]

its pretty easy to figure out with the original pinout of the MA trannie shown above. you just have to connect the pins accordingly and it should work. i even moved the whole construction to the topside to avoid a short. the cables are about 5 cm long so no prob there.

[PimpleX]

Hedbanger ...Not sure if you'd be willing to try this, ...but could you disconnect the transistor on the Q2 part of Q7R1 (pins 1, 2 and 6) and see if your mobo still works?

And what your LPC pin 5 and pin 1 volts are without it?

Cheers,

PimpleX

[G0t M4xx 21]

OK, just one other thing, just because Lclck is 0v and Lrst is 0.2v does not mean the transistor pair is bad.

This has been confirmed on a 1.5 that is FRAOing due to what i assume is bad ram. Replacing the tranny did nothing, and grounding d0 without a chip connected resulted in frag (because it tries to load the bios before checking the ram, if the tranny or the mcpx were botched it wouldn't even get that far).

The board FRAO's most of the time, sometimes boots coma or locks up in the flubber, and even less often boots to the dash but freezes in less than 30 seconds. When it FRAO's Lckck is 0v and Lrst is 0.2V, but if it actually boots (coma, or actual flubber) the voltages are normal.

So IMO frag is that it cant load the bios for some reason, and frao is either circuitry related to Lclck and Lrst, or bad ram. Also these odd voltages may actually be the effects of frao not the cause.

[G0t M4xx 21]

Nevermind stupid me.

I finished my 128mb mod yesterday and it booted up fine ,and ran perfectly for a few hours then started doing stupid stuff. I found a loose pin, and it worked again for a hour or two then started locking up and doing stupid stuff.

By this morning, it was stuck in its frao/coma/freezing flubber thing. Since I already checked all my loose pins I assumed the ram was bad.

I jsut resoldered EVERY pin and its back working again. I musta had one that was connected but barely and was an intermittant joint.

Lesson learned: Even if a 128mb box boots it doesn't mean there isn't a loose pin.

[hedbanger]

the box is assembled again and only waiting for a bigger hd, but if noone else can/wants to do it i can do it tommorow or the day after tommorow.

[G0t M4xx 21]

yeah mine's all good now. I musta rebooted it 20-30 times so far, boots up every time. XBMC sees the 128mb, so does linux, dvd2xbox, etc. It's been running for baout 2 hours with no lockups

[ferrari_rulz_02]

QUOTE(G0t M4xx 21 @ Nov 28 2005, 12:18 PM)

yeah mine's all good now. I musta rebooted it 20-30 times so far, boots up every time. XBMC sees the 128mb, so does linux, dvd2xbox, etc. It's been running for baout 2 hours with no lockups

excellent. just out of intrest, where did you get the ram, and how much did it cost?

[G0t M4xx 21]

i pulled em off a dead mobo I had laying around (mcpx fried). Cost: $0

I used a heat gun to remove em

I also used my "ghetto hot air reflow" to replace the tsop on that 1.5 board, I bought the board on ebay as broken it jsut had a dead tsop. I chipped it as soon as i got it but then yesterday i was bored so i pulled the tsop and ram off that dead mobo and put it on there. I got really lucky and the tsop i used as a replacement was already reflashed! (isn't that a nice surprise, i was dreading getting raincoat running again)

You know ive got too many xbox parts if i cant remember if i tsop'ed a motherboard or not

This post has been edited by G0t M4xx 21: Nov 28 2005, 03:22 AM