To cut Lframe, or not to cut Lframe.

[Bowlsnapper]

That is the question. I've heard different things about grounding D0 to a pad without cutting Lframe. I've heard that It will eventually cause fryage, and I've heard that it's an old wive's tale. What is your opinion?

 

I HATE cutting traces...

[SS_Dave]

56 minutes ago, Bowlsnapper said:

That is the question. I've heard different things about grounding D0 to a pad without cutting Lframe. I've heard that It will eventually cause fryage, and I've heard that it's an old wive's tale. What is your opinion?

 

I HATE cutting traces...

I tested the current of the D0 on 1.0-1.4 and it's in the vicinity of single digit micro amps (from memory 7 or 8 micro amps) and a 1.6 is in the  high 2  digit micro amps ( around 35 micro amps from memory) I know it doesn't sound a lot but why stress the components if it's not needed

Chips like the X3, Open Xenium, Xenium only ground the D0 just long enough to make the Xbox load from the LPC and the rest of the time it's floating.

Chips like the Aladdin ground the D0 the entire time the chip is active, And some of the Aladdin chips can't pull the D0 on a 1.6 to ground at all and you need to link the D0 and L1 on the chip in order for the chip to work on a version 1.6

 

Me I always cut the LFRAME trace with all chips as I see no reason to ever need the original bios again.

 

Cheers

SS Dave

Soft modding is like masturbating, It gets the job done but it's nothing like the real thing.

[lopenator]

and if you ever wanted to boot retail again ...for whatever reason you have, you can simply repair the trace. it's not permanent.

[Bowlsnapper]

Thank you for the info Dave. I have the D0 grounded to a ground pad. I was told that d0 is janky on the chip. I wouldn't know since I had BT hooked up when I tried it and it shouldn't have been. Should I just try D0 on the chip and leave it if it works? I am ANAL about booting my stock TSOP... I think it's a leftover obsession/habit from my Xecuter 3 days. Lmao. I like to be able to do it. There's NO good reason for this.

What the hell is Lframe, anyway? And how does the Aladdin know which front button is being pressed and thus which bank to boot if it's not tied in to the switch panel? My assumption was that it just saw power and tried to boot, and that would be all it was capable of.

@SS_Dave @lopenator

[SS_Dave]

On a stock Aladdin chip if the power switches on with the BT held low it boots with the chip active

The X3 chip has connections to the power and eject buttons to determine what happens

According to wiki The Lframe  This active-low signal indicates the beginning of an LPC bus transaction. Driven by the host only.

I suspect that when you power on with the eject button there is some data on the LPC port that the chip sees and knows that's a eject button.

18 hours ago, Bowlsnapper said:

Should I just try D0 on the chip and leave it if it works?

You can try it 

 

Cheers

SS Dave

Soft modding is like masturbating, It gets the job done but it's nothing like the real thing.

 

[Bowlsnapper]

6 hours ago, SS_Dave said:

On a stock Aladdin chip if the power switches on with the BT held low it boots with the chip active

The X3 chip has connections to the power and eject buttons to determine what happens

According to wiki The Lframe  This active-low signal indicates the beginning of an LPC bus transaction. Driven by the host only.

I suspect that when you power on with the eject button there is some data on the LPC port that the chip sees and knows that's a eject button.

You can try it 

 

Cheers

SS Dave

Soft modding is like masturbating, It gets the job done but it's nothing like the real thing.

 

I don't even honestly know what BT is... Held low? What about the XBlast? I'm assuming the CPLD has something to do with this? Why was it fragging when connected?

Yes, the X3 has dedicated connections to the front panel. That's why I was so confused about power and eject with this xblast Aladdin. lol

I need to read up on Lframe. I'm confused as to why both Lframe and D0 exist. It seems like they perform the same function.

I do know that Cerbios boots to compatibility mode when eject is pressed so it has to read it somehow. I think you're right.

I will try re-wiring D0 when I do the controller port LEDs.

[SS_Dave]

2 hours ago, Bowlsnapper said:

I don't even honestly know what BT is... Held low? What about the XBlast? I'm assuming the CPLD has something to do with this? Why was it fragging when connected?

I tested my Aladdin Xblast chip and powering on with the power button with the BT held low (ground) to boots to the nominated bios file and with the BT not connected it loads Xblast menu

The term BT is probably short for button

 

Cheers

SS Dave

Soft modding is like masturbating, It gets the job done but it's nothing like the real thing.

[Prehistoricman]

On 6/15/2023 at 9:46 AM, SS_Dave said:

I tested the current of the D0 on 1.0-1.4 and it's in the vicinity of single digit micro amps (from memory 7 or 8 micro amps) and a 1.6 is in the  high 2  digit micro amps ( around 35 micro amps from memory) I know it doesn't sound a lot but why stress the components if it's not needed

Chips like the X3, Open Xenium, Xenium only ground the D0 just long enough to make the Xbox load from the LPC and the rest of the time it's floating.

This was not the experience that I had with my non-CE X3. I measured it and it was driving LFRAME low constantly, taking 2 digit milliamps. 35uA is nothing - even the most sensitive circuits will not be damaged by such a low current. I think I saw around 30mA which is short circuit output current for that kind of IO driver. It is only 0.1 watts, but it will accelerate the lifetime of that driver in the MCPX. That led me to cut LFRAME on my 1.6 and put a resistor on it.

@Bowlsnapper LFRAME is a signal on the LPC bus, the bus which the modchips use to send the BIOS data over. Holding it low on a 1.6 stops the Xyclops chip from using the bus, so the modchip can take over instead.

The REAL D0 signal on a 1.6 board is unused. D0 means data bit 0, and it's part of a parallel bus up to D7. This bus is used for flash on 1.0 - 1.4 but that flash doesn't exist on 1.6. Likewise, LFRAME is unused on 1.0 - 1.4.

The confusion starts when modchip instructions started calling LFRAME D0 because the user needs to wire up the modchip's D0 pin to LFRAME on 1.6. So they make a picture showing the LFRAME trace and mark it as D0.

[Bowlsnapper]

4 minutes ago, Prehistoricman said:

This was not the experience that I had with my non-CE X3. I measured it and it was driving LFRAME low constantly, taking 2 digit milliamps. 35uA is nothing - even the most sensitive circuits will not be damaged by such a low current. I think I saw around 30mA which is short circuit output current for that kind of IO driver. It is only 0.1 watts, but it will accelerate the lifetime of that driver in the MCPX. That led me to cut LFRAME on my 1.6 and put a resistor on it.

@Bowlsnapper LFRAME is a signal on the LPC bus, the bus which the modchips use to send the BIOS data over. Holding it low on a 1.6 stops the Xyclops chip from using the bus, so the modchip can take over instead.

The REAL D0 signal on a 1.6 board is unused. D0 means data bit 0, and it's part of a parallel bus up to D7. This bus is used for flash on 1.0 - 1.4 but that flash doesn't exist on 1.6. Likewise, LFRAME is unused on 1.0 - 1.4.

The confusion starts when modchip instructions started calling LFRAME D0 because the user needs to wire up the modchip's D0 pin to LFRAME on 1.6. So they make a picture showing the LFRAME trace and mark it as D0.

Thank you for basically erasing my confusion about this. Right on. The only other thing I'm wondering about this is: should I go ahead and use the chip D0 or leave Lframe grounded to the pad I have it to? What is typically the current that a 1.6 drives on that circuit? Micro, or milli?

Also, I've heard the bios is stored on the video encoder chip on a 1.6. Is that true? 

[Prehistoricman]

1 hour ago, Bowlsnapper said:

Thank you for basically erasing my confusion about this. Right on. The only other thing I'm wondering about this is: should I go ahead and use the chip D0 or leave Lframe grounded to the pad I have it to? What is typically the current that a 1.6 drives on that circuit? Micro, or milli?

Also, I've heard the bios is stored on the video encoder chip on a 1.6. Is that true? 

Do you mean how much current does it take to drive LFRAME after it's cut? Grounding isn't required after it's cut.

Or how much current it takes when grounded and uncut? It's about the 30mA mark from memory.

No, the Xyclops is the SMC or system management controller and it contains the 1.6 BIOS. The video encoder isn't hooked up to the LPC at all so it can't possibly be used as BIOS storage.

[Bowlsnapper]

12 minutes ago, Prehistoricman said:

Do you mean how much current does it take to drive LFRAME after it's cut? Grounding isn't required after it's cut.

Or how much current it takes when grounded and uncut? It's about the 30mA mark from memory.

No, the Xyclops is the SMC or system management controller and it contains the 1.6 BIOS. The video encoder isn't hooked up to the LPC at all so it can't possibly be used as BIOS storage.

30 MILLIAMPS!? Christ on a CAMEL. I need to try the chip's D0 point. Or just cut the trace. That's a lot of power. No wonder it fries... Whatever IT is... MCPX? Wow I need to study. Any references you could shoot me to read up on would be appreciated!

DId MS hide the bios in the xyclops to keep people from getting at it? If so what did they hope to accomplish by doing this if there was access to the lpc and Lframe? Were they hoping people wouldn't figure out how to rebuild? Lol.

how did you learn so much about the Xbox anyway? Almost seems like you designed it... Well you AND Kaos

[Prehistoricman]

Oh, it's actually worse. Around 71mA on my box

So that draws 0.071 * 3.3 = 0.23W which is enough to burn small components.

The LFRAME grounding trick was another Microsoft oversight I suppose. The BIOS is in the Xyclops because it's convenient and cheap. The 1.6 is basically a cost-down Xbox and this saves them from including a flash chip like the previous models.

 

[SakisTheDrifter]

Cut the Lframe and add a switch between the 2 cuted terminals. That way you can disable\enable it as you wish

[Bowlsnapper]

8 hours ago, Prehistoricman said:

Oh, it's actually worse. Around 71mA on my box

So that draws 0.071 * 3.3 = 0.23W which is enough to burn small components.

The LFRAME grounding trick was another Microsoft oversight I suppose. The BIOS is in the Xyclops because it's convenient and cheap. The 1.6 is basically a cost-down Xbox and this saves them from including a flash chip like the previous models.

 

Oh my FUCK... Okay. Thank you. I'll get it dealt with... If D0 on the chip is fuxored, what's the best way to cut a trace? I've done it with a blade, but I never been sure that it's been cut since the slice is so narrow... 

I didn't know they were trying to save with that hardware revision... They fucked off the LPC though, so I KNOW that was intentional. Lol

3 hours ago, SakisTheDrifter said:

Cut the Lframe and add a switch between the 2 cuted terminals. That way you can disable\enable it as you wish

Possibly, yes.

[Prehistoricman]

5 hours ago, SakisTheDrifter said:

Cut the Lframe and add a switch between the 2 cuted terminals. That way you can disable\enable it as you wish

I cut it in a specific place close to the Xyclops and put a 470 ohm resistor between the two cut ends. This lets the Xbox work as normal with and without the modchip and reduces the current greatly.

1 hour ago, Bowlsnapper said:

Oh my FUCK... Okay. Thank you. I'll get it dealt with... If D0 on the chip is fuxored, what's the best way to cut a trace? I've done it with a blade, but I never been sure that it's been cut since the slice is so narrow... 

I didn't know they were trying to save with that hardware revision... They fucked off the LPC though, so I KNOW that was intentional. Lol

I always use a blade. You can do two cuts at slightly different angles to get a bigger segment out.

Yeah the 1.6 uses fewer caps, fewer chips, a more modern version of the GPU, and a less capable PSU (IIRC) all to reduce the cost.

[Bowlsnapper]

15 minutes ago, Prehistoricman said:

I cut it in a specific place close to the Xyclops and put a 470 ohm resistor between the two cut ends. This lets the Xbox work as normal with and without the modchip and reduces the current greatly.

I always use a blade. You can do two cuts at slightly different angles to get a bigger segment out.

Yeah the 1.6 uses fewer caps, fewer chips, a more modern version of the GPU, and a less capable PSU (IIRC) all to reduce the cost.

That's good advice! I actually have a 470 ohm resistor so I'll go ahead and use it for that... I'd rather have the chip be wireless if I can get away with it instead of hooking up D0.

Also good advice on the segmentation. I just hope there's enough room for me to carve something out like that and I don't f*** something up.

[Prehistoricman]

49 minutes ago, Bowlsnapper said:

That's good advice! I actually have a 470 ohm resistor so I'll go ahead and use it for that... I'd rather have the chip be wireless if I can get away with it instead of hooking up D0.

Ah, but if you want the modchip to always be selected without a D0/LFRAME wire, then you should only cut the trace. Adding a resistor like I did restores the original function where a D0/LFRAME wire is required.

[Bowlsnapper]

36 minutes ago, Prehistoricman said:

Ah, but if you want the modchip to always be selected without a D0/LFRAME wire, then you should only cut the trace. Adding a resistor like I did restores the original function where a D0/LFRAME wire is required.

So LFrame/D0 can be completely disconnected from ground or chip if the Lframe trace is cut. D0 needs to be connected either through ground or on the chip if Lframe is bridged with a resistor, but it at least won't fry if it's hooked up that way. Do I have that right?

[Prehistoricman]

Correct

[Bowlsnapper]

On 6/17/2023 at 6:48 PM, Prehistoricman said:

Correct

Thank you. You guys were right: 450 was way too fuckin hot. I have it at 250 and it still works fine... and it IS Celsius. And it's not blacking up at this temperature. It's silver but solder won't stick to it. I'm making it work though. Trying to repair this trace rot on my other halo console. I think I see which ones are broken. I could use some alternate points if you know what they are. Soldering to traces is not hard but I don't think I can scrape while isolating because these are close together... I need pads I have a thread I just started if you're able to help...

Solder is sticking to the tip now. MUCH easier to use.

[SylverReZ]

There's a spare LFRAME point just right behind on the side of the motherboard as seen in this figure.