Super Aladdin Live

Not trying to start any arguments over my topic. I appreciate everyone's input. I am having an issue tho. The console I have installed the chip in FRAGs when I tried to boot it after the install. I tried re-doing all the wires on the bottom of the board at the LPC but that didn't work. I also tried using an official Aladdin chip I had from ~2005 and that one didn't FRAG. So, I figured it was a bad chip. I contacted the seller and got the free replacement in today. installed the replacement and....FRAG! IDK what to try next. I can upload images of my work if need be.

 

Here is my work. Let me know if I missed something.

http://imgur.com/a/10SfI

 

In case you didn't know, the other "dumb" CPLD mods I released for Aladdin XT PLUS2 also properly handle LFRAME on 1.6/1.6b. Just wire your Motherboard's LFRAME point to the L1 on the Aladdin. There's a file to use with the 256KB flash chip that comes with stock Aladdin XT PLUS2.

Even if the way most modchips handle LFRAME on 1.6/1.6b does not seem to damage the motherboard, it really does overtime. Putting LFRAME to ground cause an excess current flow from the MCPX. Those kind of logic ICs are not designed to supply much current. This puts extra strain on the IO drivers, generate a good amount of heat locally and could lead to trace damage inside the MCPX. This extra current flow happens the entire time your Xbox is powered ON.

It's like not having enough oil in your car engine. It will seem to work fine for quite some time but will increase wear and make it break prematurely.

Granted there are no reported case of such issue but clearly identifying this particuliar cause of failure is really hard. You must look specifically for it and cut traces on the motherboard just to test. Plus, if the damage is only limited to burning that specific IO driver, you will not witness this issue until you disable the modchip and try to boot the stock ROM. When would you want to do this nowadays? It's a possible outcome but it could also do collateral damage and break other parts of the sillicon, leading to unexplicable FRAGs or other error. It certainly has happened before but most likely, people just witnessed their Xbox die unexplicably and trashed it, not knowing why.

There is no certainty if your 1.6/1.6b will die of this specific issue but not all MCPX chips are created equal. Like CPUs, some are of higher quality, others have it worst. It's the sillicon lottery. Best case is to treat it has prevention.

You could also just cut the LFRAME trace that goes from the MCPX to the Xyclops chip and not wire LFRAME on your modchip. Downside of this is you "permanently" disable the ability to boot stock ROM. But then again, why would you want to boot the stock ROM nowadays if you have a modchip in there?

 

Can you upload a picture of your LFRAME soldering point on the motherboard?

If you completely remove the modchip but leave the LPC rebuild as it is, does your Xbox boots?

 

I'm a bit confused on the LFRAME point. but yes, it does boot normally without the chip. I didn't cut any traces by the xyclops chip. My guess is the chip is blank, and when the console tries to boot off of it then it FRAGs...but I'm just guessing and you probably know way more than I do lol.

 

I didn't see you did this, only the xblast one. Thanks!

 

Ok, the fact your Xbox still boots without a modchip proves the LPC rebuild you did does not short anything. Sometimes it's the most obvious things that we miss!

I'm sorry, LFRAME is called D0 anyway on most modchip install guides... It's just that on 1.6/1.6b you're wiring the modchip's D0 pad to a signal called LFRAME on the Xbox motherboard. Check this pic to identify the d0 point on the motherboard (bottom side). You must have D0 soldered to your modchip for it to work. You may have soldered it on the top side of the motherboard as well, it would be near the Xyclops chip (bottom right corner when reading the markings on the Xyclops the right way).

One other thing, do you know what BIOS comes preflashed with Aladdin Super Live usually? I fear the chip might be flashed with a BIOS that does not support 1.6/1.6b. In case you didn't know, there are BIOS for 1.0-1.5 XBoxes combined and other BIOSes for 1.6/1.6b. Super Aladdin Live is an older generation chip and I wouldn't be surprised they'd be flashed with a X2 4xxxx BIOS which are not compatible with 1.6/1.6b...

If you have a 1.0-1.5 Xbox, try to install the Super Aladdin Live in it, see if it boots. You 'll be able to identify the BIOS on it then.

 

AladdinLive.com tutorial for the Super Aladdin LIve modchip from 2004 (posted to xbox-hq.com):

The rest of the tutorial guide reads:

They said there's a way to flash for 1.6 but I read no details as they stated would be given later in the tutorial. I see using the external programmer but no details on it's use. Do you have the programmer for the Super Aladdin Live modchip? Or, hotswap the modchip to a non-v1.6 Xbox and reflash it with a v1.6 compatible BIOS.

 

I followed this tutorial except for cutting the trace at the end. link:
@bennydiamond , I'm not sure if I actually soldered to the D0 or not. I know it sounds dumb but I figured that's what the rebuild on the bottom was doing lol. Here are the pics. http://imgur.com/a/PeCL6

 

Yeah, those chips most likely come preflashed with Evox M8+ for 1.6/1.6b which do boot on 1.0-1.5 Xbox but shouldn't be used extensively with this BIOS. The RAM timings in this BIOS suits 1.6/1.6b but are not recommended for earlier revisions.

Anyway, you should always reflash the BIOS on them with an appropriate BIOS for your Xbox revision once you've established it boots fine, whatever your console's revision.


@ToXZiN 1 Your issue is most likely the LFRAME signal on the motherboard not properly interfaced. If you say your Xbox boots without the modchip seated, it means the Xyclops still tries to feed the stock ROM on boot.

In the video you linked, the guy cuts the LFRAME trace that goes to the Xyclops chip. It starts at 18:30. That's why he doesn't bother soldering the D0 point on the Aladdin XT. The Xyclops, without the LFRAME signal coming in, just sits idle and never tries to send the stock ROM, giving away the LPC bus to the Aladdin chip.

You have 3 choices here.

1- Solder a wire from the D0 (on Aladdin) to the LFRAME (on Xbox motherboard) which will ground it permanently as I see from the pictures, your Aladdin is set to always ON (wire soldered on the BT pad).

2- Solder a wire from the LFRAME(on Xbox motherboard) to a ground point on the motherboard directly. I acts identical to solution 1 but you don't have a wire going from your motherboard to the Aladdin, leaving you the ability to freely remove the Aladdin without having to desolder the wire from it first.

3- Cut the LFRAME trace on the Xbox motherboard just as instructed in the video you linked. It's a "permanent" solution to disable stock ROM loading and you don't tie the LFRAME to ground, alleviating the inherent problem of grounding the LFRAME signal.

After that, your Xbox should boot. Once you've established the modchip works, reflash your Aladdin with Evox M8+ for 1.6. Use EVTool to modify the settings like fan speed, boot paths and other stuff to your liking.

 

@bennydiamond You are a badass. Got it working by soldering the D0 to the proper point. Chip booted Evox M8.
Thanks again to everyone who commented suggestions!

 

How much current is drawn when done this way? What is the maximum amount of current the mcpx is capable of supplying?

 

Someone experimented on this a while ago. His calculations of current draw are theoritical only. In reality, you can see on the oscilloscope capture that the LFRAME signal drops around 1.2V when normally it should be 3.3V. This signify a higher current draw than what the MCPX chipset can supply.

Ohm's law states that the voltage of a circuit is directly proportional to the total current multiplied by the equivalent resistance of the circuit (V = R*I). If you reduce the voltage to 0V or close to it by forcing it to GND (what the modchip does while the MCPX tries to stay at 3.3V), you effectively brings the resistance close to 0 ohms (not 0 but way more close to it than the couple MegaOhms it's supposed to be at). Then, you must balance the equation to try to stay at 3.3V. The only way to do this is to increase the current supplied. If the the current is not sufficient to balance it (because the transistors inside the MCPX physically cannot let an infinite amount of current flow), then the only mathematical outcome is to reduce the voltage, hence the 1.2V seen on the oscilloscope. At that level, the equation is balanced. However, this is has the effect of saturating the transistors to their physical limit, which results in excess heat and reduced life expectancy.

Since nVidia is a bitch and do not make their documentation available, we can only speculate on the max current rating of the IO drivers on its LPC interface. We can assume it's close to some of their competitor's corresponding ICs. Here's a datasheet of a AMD's south bridge chipset for comparison. Table 50 states IO lines shouldn't exceed 12mA in normal condition. Absolute maximal operating specs won't be much higher than that.

 

Glad you found that link, I was looking for it but couldn't find it to use on way back machine.

 

If we're THAT far out of spec, it shouldn't be lasting years like it does. So I may go ahead and connect my clamp meter to make sure the current draw isn't more than an amp or two. (It's clearly not or it wouldn't be working.) Then I'll hook my dmm up in series to measure the current draw.
What I get from your post is that every sense you have seems to be telling you that it puts the xbox in imminent danger in spite of all real world results being to the contrary. I get that. I can follow that. Then we blindly speculate on current draw and maximum limits to base our supposition on. I think we need to get some more concrete numbers, but the fact remains. Xbox can be had for 20.00. Modded in the way I provided has lasted years without issue. So 20.00 every few years isn't the dire issue we're saying it is. That much is clear.... but I'll try to get some measurements to see if there isn't some slight modification that can be made to put the alarmists to rest. If I know alarmists though, putting one thing to rest isn't enough. There will be something else. We might be embarking down primrose's path.

 

Or you could just do it correctly in your guide, rather than damaging hardware that isn't made anymore.

Your digital multimeter isn't going to very good at measuring 12ma (unless it's a good one) . Probably better to use scope and measure voltage drop over a known resistance, assuming you have one. No idea where you are pulling the 1 or 2 amps figure from though.

But the measurements have been made. The math is absolute, just correct the guide instead of continuing the ignorance that you are correct.

The most likely failure from this chip is that io gets burnt out.... Which means the xbox will only boot from a modchip. But it could cause other failures, but I'm not into damaging things that don't get made anymore..... Especially when there's a proper way to do it and it's just cutting a trace! You have already been told this, nothing you measure and "check to see if there's a slight modification to put the alarmists at rest" will be anything that's not already been said in this thread to you.

Benny made his own chip, he does know what he's talking about. He's also an electrical engineer. I'm currently working on my ee degree too.

 

The guide seems to BE correct since there are no failures and that "absolute" math hasn't been provided. You should check your attitude at the door because you've already proven you don't know as much as you think you do.

Cutting a trace is more likely to destroy the xbox than what you're concerned about. Lol! The fact that you think you have the logical high ground is hysterical. Especially when you look where that trace is located and the size of it. We'd have every new hobbyist with a 1.6 with a destroyed xbox. Not just one that might only last half a decade. You're not into destroying things that aren't made anymore! Keep cutting traces.

Don't lump yourself in with Benny. He's being logical and sticking to the facts. Your ego is on the line here. You need to be right and you're insistent on attacking me. At this point, I've done all I can to avoid it... and you're not deterred. So, provide this math you say is absolute or stop talking.

 

Says the guy trying to push his forum here and has a vested interest in looking like he knows what he's talking about.

The fact is, you don't know enough to understand it's been laid out in front of you. So how can you expect to judge what I know?

Do remember, I said use a chip that does it properly in the first place. (which bennydiamond makes or has provided the files to flash you own aladdins for free!)

The link Bennydiamonds provided has the current measurements in it. Go read that. Bennydiamonds post explaining the voltage drop is the maths, go read that again too.

Your entire argument is "works for me", nothing based on measurements or basic electronics knowledge. I'm quite done arguing with you, as it's obviously a waste of time, it's like talking to religious folk, you will believe what ever you believe because you believe it. If you come back with some actual science and can prove you are correct, I'm happy to be corrected.

Until then, have fun!

 

The forum I shared isn't about me personally, so my credibility has absolutely nothing to do with it... so you fail there.
The fact is, I DO know enough to know it has not been laid out in front of me. I see that Benny said that NVidia did not release the specs of the MCPX so he doesn't know what the tolerance is. I see that we didn't know what the resistance was, so we couldn't do the math.
So, it wasn't laid out in front of me. It was a big guess. The amount of current drawn was never said because it cannot be calculated because the only known is the voltage.
My entire argument is that it works for everyone who has tried it with no issues and there is no evidence to the contrary.
Nothing you and frankly that Benny has said has been anything other than a guess. So until you have some of what you claim to have, you lose your own argument.

 

Current measurements here, as linked previously.

https://web.archive.org/web/20050306201343/http://xbox.bula.nu:80/resistor_testresults.asp