XP-ALLin - the latest and greatest in the XP-ALL project.

[digi2t]

First off, a mighty THANKS!!, and a well deserved round of applause for Prehistoricman for this latest update to the XP-ALL.

Managing to reverse engineer the Digitech .bin files, he rearranged the code, allowing each mode to now retain whatever user presets one wishes to save in any mode. Each mode will now retain those presets in memory, regardless of what presets are programmed in other modes.

This, is a major advancement in the project. Previously, the user presets section of the XP-ALL+ was rendered useless, due to the fact that only one section of memory was being used by all the modes. Since some modes had more/less presets than other modes, and some modes had preset numbers that did not exist in other modes, saving user presets would sometimes overwrite saved presets in other modes, or sometimes create errors that would require factory resets.

The new code, baptized "XP-ALLin", has been tested on my XP-ALL+ platform (converted XP-200), and is fully functional.

The updated XP-ALL build document can be found here;
https://drive.google.com/open?id=1aIbRsYi9-_odxECMjSIw4PFZCpxpx_GD

The new XP-ALLin .bin file (for XP-ALL+ 27C1001 application) can be downloaded here;
http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=53391&g2_GALLERYSID=b9edd9290db937f0dfd0be36edb4e9a4

Prehistoricman's video on how the Digitech code was reverse engineered, and modified to all for user presets to be saved, can be viewed here;

When we initially released the XP-ALL, there were two functional drawbacks; the loss of stereo output in 100 mode, and the inability to securely store user presets in all modes. While the relay workaround solved the stereo issue, the XP-ALLin now addresses the user presets problem of the XP-ALL.

Coming soon - replacement of the preset mode switching with a rotary switch, to a footswitch style mode switching.

[Prehistoricman]

Regarding stereo output for the XP-100 mode, this is the bottom line:

Don't hold your breath. I'll have a real crack at it in a couple of weeks when I have my EEPROMs and a programmer.
The DSP controls stereo output and there is very little documentation on the internet about the DSP. That's the big issue with the XP-100 stereo output.

However, I'm pleased with the current progress. Big props to digi2t and Govmnt_Lacky for doing all the work towards the XP-ALL. They're still the true heroes here 

[MüThing]

Some really excellent work from all you folks, this is an absolutely amazing development for the XP! Am I right in saying if you already have the XP-All mod installed the new rom should work without an issue?

[digi2t]

Some really excellent work from all you folks, this is an absolutely amazing development for the XP! Am I right in saying if you already have the XP-All mod installed the new rom should work without an issue?

Yes. I currently have the new XP-ALLin chip running in my XP-ALL+ unit. It was a straight swap out. I do recommend doing a factory reset once the new chip is in however.

I've done a complete shake down on it;
- tested all modes for user preset memory retention.
- memory retention when other mode user presets changed.
- no problem in saving presets in non-linear fashion (they don't have to be in an ascending /descending order).
- complete functionality of all factory patches.

Nothing out of the ordinary has popped out to date. No C5 or C7 errors at all. Same 'ole XP, just better.

[vigilante397]

I don't even do digital stuff, but I want you all to know I think this is remarkably cool

[Govmnt_Lacky]

I believe this will only work with the "+" models. It will not work with the stacked EEPROM conversions.

Correct me if I am incorrect.

If you have a stacked EEPROM model then you will need to convert to th single 27C1001 version.

[digi2t]

I believe this will only work with the "+" models. It will not work with the stacked EEPROM conversions.

Correct me if I am incorrect.

If you have a stacked EEPROM model then you will need to convert to th single 27C1001 version.

I think it would work with the stacked chips. Since each .bin file is modified individually, once the stack is assembled, it should work. After all, the XP-ALL+ is simply the same four .bin files, but all together on one chip.

The pain with the XP-ALL is that you need to either disassemble and reprogram your stack, or make a new stack altogether.

[MüThing]

Yes. I currently have the new XP-ALLin chip running in my XP-ALL+ unit. It was a straight swap out. I do recommend doing a factory reset once the new chip is in however.

I've done a complete shake down on it;
- tested all modes for user preset memory retention.
- memory retention when other mode user presets changed.
- no problem in saving presets in non-linear fashion (they don't have to be in an ascending /descending order).
- complete functionality of all factory patches.

Nothing out of the ordinary has popped out to date. No C5 or C7 errors at all. Same 'ole XP, just better.

Lovely, lovely, lovely!  I'll be digging out the MiniPro and putting my XP back to work on my board.

[Govmnt_Lacky]

The pain with the XP-ALL is that you need to either disassemble and reprogram your stack, or make a new stack altogether.

This was more of the reasoning I wanted to convey. 

If you are going to disassemble a 4 chip stack to do this and then re-assemble it.... you might as well convert to the 27C1001 anyways 

[Prehistoricman]

I found out the importance of decoupling yesterday. I received my extra RAM chips, solder suckered all the holes in the board and added sockets for them. Upon adding the extra RAMs, my Whammy didn't quite work as I remembered. The frequency of the pitch shift would oscillate quite badly, and I couldn't push the pedal far enough to get the entire shift. (On this note, why is the expression pedal LED powered by the digital 5V?) The signal LED was also behaving slightly weirdly.

I tried to do some decoupling. I had a couple of ceramic 2.2uF caps but it's still 5x less capacitance than the tantalums that should have been there. I added a cheap Chinese 470uF electrolytic and it didn't help. Then I replaced it with a 3.3mF electrolytic and it worked great! I suspect the lower ESR of the huge cap is what did the trick. The designers wouldn't have asked for tantalums if they didn't need to. However, I still plan to test out some higher capacity MLCCs as these would be smaller than the tantalum.

[Govmnt_Lacky]

Yep! If you install the 3 additional RAM chips, you definitely want to install C61, C64, and C65 or else there will be some "issues."

[digi2t]

Unrelated, but important nonetheless, when doing mod work on these units. A factory reset and calibration should be done before using the unit.

Factory Reset/Pedal Calibrate -
- To reset the User programs to their factory settings and calibrate the expression pedal, with power off, press and hold the <BYPASS/(Hold)TUNER> foot switch, and apply power to the unit.
- Wait until the letters rS appear in the display and release the foot switch.
- The display will now show: Pd and  CA , indicating that you are in pedal calibrate mode 
- Once  Pb appears, rock the pedal back and press the <BYPASS> foot switch.
- When PF appears, rock the pedal forward and press the <BYPASS> foot switch.
- SA will next appear in the display to confirm the pedal is now calibrated.

I myself, as well as one or two others that have done these mods, have encountered the units doing "weird things" after a mod. A simple factory reset has cleared things up in some cases.

[Prehistoricman]

What kind of weird? I'm interested.

[digi2t]

What kind of weird? I'm interested.

Patches not loading (dead air), patch working but not modulating with the treadle, were two that come to mind.

[Prehistoricman]

Update on the decoupling issue:

Wasn't actually decoupling. I scoped the power supply after adding a considerable amount of SMD ceramic caps (9) and found that there was a 0.5V dip at 50Hz. This led me to believe that the input cap wasn't holding enough juice for the next cycle of the mains supply. I doubled the capacitance of the input cap and it works fine.

[Govmnt_Lacky]

Update on the decoupling issue:

Wasn't actually decoupling. I scoped the power supply after adding a considerable amount of SMD ceramic caps (9) and found that there was a 0.5V dip at 50Hz. This led me to believe that the input cap wasn't holding enough juice for the next cycle of the mains supply. I doubled the capacitance of the input cap and it works fine.

Which input caps were you having trouble with? C?

[Prehistoricman]

C39

I suspect the issue might have something to do with the fact that I'm using my own power adapter (a 9V transformer made in '92 in England!). It outputs exactly 9VAC and I suspect that a stock power supply outputs a bit more than that.


On a different note, my EEPROM programmer is coming along. I spent 5 hours in my university's lab today checking the timing and making code

[mitchschaft]

Excellent, glad I checked this it when I did abd found the update.

[Govmnt_Lacky]

Although I do love all of the advancements that have occurred with the XP platform, it is painful to see where the prices of used or second-hand XPs have gone. Someone used to be able to get a used XP100 for anywhere from $40-70. Those days are long gone now. Going rate for them is about $125! XP200s and 400s are worse.

Hopefully, the market will settle and the prices will go back down 

[digi2t]

Although I do love all of the advancements that have occurred with the XP platform, it is painful to see where the prices of used or second-hand XPs have gone. Someone used to be able to get a used XP100 for anywhere from $40-70. Those days are long gone now. Going rate for them is about $125! XP200s and 400s are worse.

Hopefully, the market will settle and the prices will go back down 

All your fault pal.