XBRv3a on Falcon, lots of strange errors, diodes, soldering or something else?

[Snigel]

Update:
Most errors gone since I resoldered DB1F1, but it seems some errors are related to starting the xbox when a wireless controller is connected for charging. Please see last post for detailed information.

------------------------------------------------------------------------------------------------------------

Jtagged my Falcon with XBRv3a. It works fine for a while then it gave up and it resulted in lots of different errors;

Booting only possible with wired controller, eject, power button, wireless doesn't work.
Boot will lead to either e79, instant shutdown, initial setup without working controller or with working controller for a while.
The machine also boots by itself and then shuts down again sometimes.

The first time this happened I had played around with the xbox for an hour or two, everything was fine. Then I shut it down and left it plugged to the wall (orange light). When I came back a couple of hours later it was booting up and shutting down by itself. Doing all the above errors.

I tried replacing the diodes, 1N4148 with replacements, the machine booted up and worked fine for about an hour before it suddenly shut down while I was playing. I was getting some of the above errors, but not as severe.

First I thought this was a diode issue, so I replaced my 1N4148 diodes with BAT46. It worked for a while, then shut itself down while I was playing, since then it has started up by itself, gone into endless rebooting loops and such.

I have been looking for shorts all over the board. I have tried soldering one of the points to the power+led board or directly to the motherboard.
Tried running it without the box.
I have tried with the optional 330ohm resistor instead of just cable.
I have tried reflashing XBR without my config block.
Nothing seems to change issues.

I've edited this post so you shouldn't have to read through lots of now irrelevant information, perhaps the following posts looks a bit strange because of that.

[Arakon]

Try flashing back your original FW. This sounds more like a short somewhere.. check that no wires are crushed under the MB, near the screw holes.

[Snigel]

Try flashing back your original FW. This sounds more like a short somewhere.. check that no wires are crushed under the MB, near the screw holes.

I tried reflashing my original firmware. It boots up, but the power button only seem to work seldomly if it works at all. I cannot boot it with the eject button.

I have also rechecked all cables, and I desoldered them for running the original firmware (I think that's needed?). Tried to look over the whole board both on the front side and back side, couldn't find any anomalies.

Still no luck. Have no idea what have happened since it wasn't moved during the night which it stopped working.

[EyeFlight]

I had near identical issues to what you've described (Falcon running XBR 3a). I had multiple good orig.bin dumps, tried reflashing both XBR and orig.bin, replaced the wireless ROL board, tested/checked all solder points but nothing fixed it. Console would randomly boot, wireless controller would randomly lose sync after 10min, crazy ROL error lights when console appeared to be running fine, occasional E79 and NXE setup screen when it did boot correctly etc etc etc.

The solution and cause appeared to be a faulty/damaged diode. I resoldered the DB1F1 point and fitted a new diode then retouched all other solder points. The console now works as it should  . So yeah, try replacing your diodes and redoing your solder points.

[Snigel]

I had near identical issues to what you've described (Falcon running XBR 3a). I had multiple good orig.bin dumps, tried reflashing both XBR and orig.bin, replaced the wireless ROL board, tested/checked all solder points but nothing fixed it. Console would randomly boot, wireless controller would randomly lose sync after 10min, crazy ROL error lights when console appeared to be running fine, occasional E79 and NXE setup screen when it did boot correctly etc etc etc.

The solution and cause appeared to be a faulty/damaged diode. I resoldered the DB1F1 point and fitted a new diode then retouched all other solder points. The console now works as it should  . So yeah, try replacing your diodes and redoing your solder points.

Thank you very much for that information. Will try that.
Still the xbox behaved weird even when the diodes was removed.

By the way, could it matter where the diodes are placed along the cable? My diodes are in the opposite end of the cable compared to some guides. Of course they are turned in the right direction and so on.

[Snigel]

Changed the diodes today. Now it works again.
Thank you very much.

[Snigel]

Machine seems dead again. Played GTA:GT for one hour, then suddenly it shut itself off. Trying to boot it straight after just led to that it started up and then shut itself off again, and kept on doing so.
I janked the power plug and waited for a minute, now it boots into the xbox interface, but then shuts down.

I guess the diodes went bad again. I'm using 1N4148. Should I perhaps use some other diodes? Or put a resistor in series with them?
Seems like something is wrong when they die like this.

If it can differ "my" 1N4148 has the following specs 0,15A 75V 4ns. Is really such a fast diode needed? Thinking about putting something that can tolerate a stronger current there, but they are all slower than 4ns.

[littlestevie360]

Machine seems dead again. Played GTA:GT for one hour, then suddenly it shut itself off. Trying to boot it straight after just led to that it started up and then shut itself off again, and kept on doing so.
I janked the power plug and waited for a minute, now it boots into the xbox interface, but then shuts down.

I guess the diodes went bad again. I'm using 1N4148. Should I perhaps use some other diodes? Or put a resistor in series with them?
Seems like something is wrong when they die like this.

If it can differ "my" 1N4148 has the following specs 0,15A 75V 4ns. Is really such a fast diode needed? Thinking about putting something that can tolerate a stronger current there, but they are all slower than 4ns.

AFAIK yes fast is a requirement, personally i use BAT41/46 and they are fine

[Snigel]

littlestevie360: Fast yes, but how fast? Is 4ns a requirement or could 25ns work?

I'm not using a 330ohm resistor between J2D2.4 and J2D2.7 as some suggested over at xs, could that change anything for the diodes?
The first time the machine broke down I tried to use a 100ohm there with no success. Removing it and changing the diodes instantly made the machine work again though.

[littlestevie360]

100ohm was a very bad idea, 330ohm was stated for a very specific reason it was current limiting, maybe you have done damage to the GPU interface already, fingers crossed you didnt. as i said use BAT41/46 they seem to get the best results

as for the not using 330ohm, no that wont change anything, its either OLD(not as safe) 330ohm resistors, or use the diodes

[Snigel]

100ohm was a very bad idea, 330ohm was stated for a very specific reason it was current limiting, maybe you have done damage to the GPU interface already, fingers crossed you didnt. as i said use BAT41/46 they seem to get the best results

as for the not using 330ohm, no that wont change anything, its either OLD(not as safe) 330ohm resistors, or use the diodes

I think you are confusing the old Xenon (non diodes) with what I was doing. Over at XS some are using a 330ohm instead of just wire between the two bridged points. I tried with a 100ohm instead of the wire and got no difference at all though, thus I removed it.

For clarification, this is about using a resistor on the blue wire:
http://pictures.xbox-scene.com/xbox360/free60/SPI_&_JTAG_diagram_%28zephyr-falcon-opus-jasper%29.png

[Blackaddr]

J
I'm thinking that perhaps it's passing a tad more than 0.75A through the suggested soldering points on the Falcon boards which slowly kills the diodes. Or could it be a reverse current which is too high? It could also be too high voltage, I haven't measured but perhaps somebody knows more than me about this.

Where did you get this 750 mA amp number from?  The circuit used for connecting the SMC to the GPU is electrically safe and does NOT draw 750 mA.  The 1.5K pullup on the GPU pin is on the board, it's not part of the mod.

SMC Out (3.3V) ---|<----- R=1.5K ----- 1.8V rail
                                  |
                                  |
                               GPU In

When the SMC output is HI (3.3V) the diode is reverse biased and cut off, no current flows through it.

When the SMC output is LO (0V), the diode is forward biased.  The diode is a low-voltage drop diode but even if it has almost zero voltage drop, the maximum current would be I = V / R = 1.8V / 1500 = 1.2 mA MAX

You cannot burn up the diode if installed correctly.  If your xbox is having electrical problems, somebody did shoddy workmanship.  Either the guys who manufactured the board, or the guy who modded the board.

[Snigel]

Where did you get this 750 mA amp number from?

That's what 1N4148 diodes are rated for at the specs I'm looking at. Since the diodes were malfunctioning after a while I thought they were being overloaded in some way.
I don't have anything to properly test these switching diodes (except for a simple multimeter) so I can't say for sure that they are broken. Just that the machine started working when I exchanged them with new ones, and resoldering the old ones didn't change the issues at all.

 The circuit used for connecting the SMC to the GPU is electrically safe and does NOT draw 750 mA.  The 1.5K pullup on the GPU pin is on the board, it's not part of the mod.

SMC Out (3.3V) ---|<----- R=1.5K ----- 1.8V rail
                                  |
                                  |
                               GPU In

When the SMC output is HI (3.3V) the diode is reverse biased and cut off, no current flows through it.

When the SMC output is LO (0V), the diode is forward biased.  The diode is a low-voltage drop diode but even if it has almost zero voltage drop, the maximum current would be I = V / R = 1.8V / 1500 = 1.2 mA MAX

It's really nice to finally see some numbers. Thank you for that.
Since you have some electrical knowledge of the mod, you don't happen to know why several people at XS report that they need to use a 330ohm resistor instead of just wire between jd2.4 and jd2.7 on their falcon boards? I'm trying to find out if this could be somehow related to my problems.

You cannot burn up the diode if installed correctly.  If your xbox is having electrical problems, somebody did shoddy workmanship.  Either the guys who manufactured the board, or the guy who modded the board.

The guy who helps me is a real pro. We've looked at the solderings several times. Also tried resoldering the diodes before exchanging them. I really doubt that something is wrong with his work since we've gone through it so many times now.
If the board is having electrical problems it has not shown up in any other way. Been running the console for two years before the jtag-mod without any problems.

I guess it's not impossible that the batch of diodes I bought all are bad, but I somehow doubt it.
Suppose that my xbox is doing something strange to the diodes, what would be the best way to measure it?

Lastly, could it hurt anything to run resistors in series with the diodes? I still want to try that out until I can find a better solution.

[Snigel]

Just wanted to post a little update.

After the box stopped working the second time I left it unplugged over the weekend. After getting some new components I just wanted to check if the errors still was there. To my surprise it was working. I've left it in standby over night twice now. Left it running in the menus for hours, also tried playing for several hours.

I really don't understand what has happened. Last time I got errors, I tried unplugging it for at least 30 minutes before trying anything again.

[Snigel]

Now it has gone a couple of more days, the xbox has booted itself up twice. Otherwise everything still works fine. When it has booted itself up i've had my wireless controller connected with the charging cord and a USB harddrive (which is USB powered). Could this cause these issues?

Update: It's not the usb drive or the controller, today it booted up by itself without any of those. Something strange is still going on alright.

Update2: Tried reflashing XBR withouth config block, still same errors.

[maximilian0017]

Now it has gone a couple of more days, the xbox has booted itself up twice. Otherwise everything still works fine. When it has booted itself up i've had my wireless controller connected with the charging cord and a USB harddrive (which is USB powered). Could this cause these issues?

Longshot:
You did remove the LPT plug from the computer i hope, and how long is that cable?, did you check it for shorts?

[Snigel]

Now it has gone a couple of more days, the xbox has booted itself up twice. Otherwise everything still works fine. When it has booted itself up i've had my wireless controller connected with the charging cord and a USB harddrive (which is USB powered). Could this cause these issues?

Longshot:
You did remove the LPT plug from the computer i hope, and how long is that cable?, did you check it for shorts?

I removed the jtag cables when the flash was done, so nothing of that is connected at all.

UPDATE: Changed diodes from 1N4148 to BAT46 today. Noticed that I needed to use a resistor between J2D2.4 and J2D2.7 otherwise the machine wouldn't boot up. I will leave it in stand by over night and see if it boots by itself.

[ZerOneX]

OUTSTANDIG working man... please keep the post updated... it will help a lot the others users.

[Snigel]

Tried to play some GTA4. Resulted in that the machine shut down and went into a loop of starting up and shutting down. What's important here is that the power off isn't like you unplug the machine, it's exactly like when you press the power button. A slow power off while dimming the screen.

What could cause this? Must be the same thing that makes the machine start by itself. Like something randomly presses the power button.

[dak]

Seriously i dont think it are your diodes. Maybe they get damaged by the rebooting, tho.
Anyway, did you try to run the console while it is not in the metal case ?
This sounds more like a "half" short somewhere.
If you google for xbox keeps rebooting you will see this or a similar issues on stock consoles without jtag.
Did you watch the Power Supply's LED while it reboots ? Maybe its a current problem from the PSU.
Maybe measure the 5V 12V voltages if they are constant.