Axon - Repairing an Axon AX 100 guitar-to-MIDI converter

Started by Elantric, August 26, 2015, 08:58:41 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Elantric

http://edinburghhacklab.com/2015/03/axon-ax-100-repair-pt1/

Repairing an Axon AX 100 guitar-to-MIDI converter (part 1)

Gareth writes: I've wanted a Guitar-to-MIDI converter (a.k.a. a guitar synth) for yonks. In particular, I've been looking for an Axon AX 100 unit; they were renowned for the speed and accuracy of the tracking that the proprietary neural network software could achieve. Unfortunately the AX 100 has been out of production for years, and used units are commanding prices in the high hundreds of pounds. However, I recently found a reasonably priced spares-or-repair unit on eBay, so I picked it up.

The Axon AX-100 was originally designed in Germany by Blue Chip Audio in the 1990s, then the company was acquired by Terratec, who slowly strangled the product line.

Here's a video of Burr Johnson demonstrating the Mark II unit. If you can cope with the cheesy synth patches he's using, it's a good explanation of what a guitar synthesiser can do.


I picked up a spares-or-repair unit on eBay recently and on Saturday I got down to the business of repair. This is one of the original Mark I units from Blue Chip Audio, built in 1998; it can be seen below. The seller had said that it wasn't starting up and that the NiCad had leaked; he had removed that NiCad from the board before shipping.


Internally, none of the parts looked to be non-standard apart from the programming of the EPROM and the Lattice PLDs that are in it, so I was optimistic that any failures could be repaired without having to locate another unit to cannibalise. There was a small amount of corrosion around the NiCad site, but nothing too sinister. The picture below shows the board layout with my interpretation of the devices and functions on it.


Powering up the unit showed two LEDs and the LCD backlight on the front panel coming on, but no further life from the unit, as shown in the picture below.

AXON AX-100 failing to boot
The NiCad battery that was removed sits next to a SRAM device that holds the settings and customisation for the box; everything I'd read on the NiCad failure (and it seems to be a common failure mode) suggested that the only symptom would be no retention between power cycles, rather than an outright failure. Something else was wrong.

Disconnecting the Yamaha synthesiser board made no difference to the behaviour; this was disappointing, as I am not planning to use the Yamaha synth card if I can repair the unit.

Al suggested disconnecting the display and trying again, and that stroke of genius was a big step forward. About a second after the same two LEDs lit as before, they went out and one LED lit elsewhere on the panel, and I could move the lit LED around the front panel by pressing the corresponding buttons.

AXON AX-100 booting (no display)
A bit of investigation of the type number of the LCD showed that it was a 16×2 with a standard interface, but the pinout was somewhat unusual. Most 16×2 LCD units these days have solder holes along one edge for hardwiring a harness to, but this had a two-row IDC header connector on it. A bit of digging yielded an equivalent Midas part to the original.

AXON AX-100 - back of display
So now I wait for the replacement LCD to arrive, and then part 2 of the post will follow once I've done some testing.

Continue to part 2.

http://edinburghhacklab.com/2015/04/repairing-an-axon-ax-100-pt2/
So the replacement LCD (a Midas MC21605B6WK-FPTLW) arrived last week, and I headed to the lab to try it out.


Replacement display
The new LCD did not come with the header pins already attached so I scavenged some from the copious lab stock of Useful Things, then plugged in to the board and powered up.

The good news: the board booted (it hadn't booted at all with the old LCD in place).

The bad news: there was nothing on the display.

I fired up the 'scope and started walking down the pins one by one, checking against the datasheet.

http://www.farnell.com/datasheets/1663609.pdf

+5V and GND were OK, but something was odd on the "contrast adjust" pin 3. This should be a constant voltage between 0V and 5V, set by a potentiometer on the front panel of the unit, but what I was seeing on the 'scope was clearly a digital pulse train. A closer look at the connections showed that the ribbon cable jumping from the mezzanine board to the LCD was in fact fitted the wrong way round; this was the state it had arrived in (and I have the photo to prove it).
https://www.flickr.com/photos/gareth_edwards/16897323682/

Reversing the connector and re-powering the unit gave this wonderful sight:

No. 5 Is Alive!
While I was soldering in the replacement NiCads, Martin pointed out that the sticker on the EPROM was lifting a bit and it might be a good idea to grab a copy of the firmware while the lid was off. This was a very good idea; EPROMs typically retain contents for 10 to 20 years, and this unit is now 17 years old. We pulled the ROM (a 27256, with a massive 32KB of storage) and grabbed a dump in the programmer.

After a bit of work on the mounting holes with the rotary hardware editor

http://www.dremeleurope.com/gb/en/

and the application of some of the Great Joiner Of Things,

http://en.wikipedia.org/wiki/Hot-melt_adhesive

everything went back together and a last test of the boot is shown here:




Next steps: although the processor now boots, I need to get hold of a "GK-ready" or "synth-ready" guitar to try out the audio pathway and see how the conversion is working. I also might need to adjust the resistor that controls the trickle charge of the NiCad since the parts I've replaced the existing unit with are at least 10x the capacity. Lastly, I'm considering using radare2
http://www.radare.org/r/

to do some reverse engineering of the firmware to understand how the unit works.




Quote
QuoteRadare is a portable reversing framework that can...

Disassemble (and assemble for) many different architectures
Debug with local native and remote debuggers (gdb, rap, webui, r2pipe, winedbg, windbg)
Run on Linux, *BSD, Windows, OSX, Android, iOS, Solaris and Haiku
Perform forensics on filesystems and data carving
Be scripted in Python, Javascript, Go and more
Support collaborative analysis using the embedded webserver
Visualize data structures of several file types
Patch programs to uncover new features or fix vulnerabilities
Use powerful analysis capabilities to speed up reversing
Aid in software exploitation

GarveyDesign

Hi, thanks for your highly informative article.

I have an AX100 mkII, and am concerned about the battery, and am considering getting some adjustment done to it in the coming months.
Obviously, losing settings could be a problem, but nothing so bad as a leakage would.

Could you advise as to how much you would charge to replace a battery?

Also you mentioned you had put one in with 10x capacity and had to make some adjustments... would this be viable and safe, and to what advantage, and what would be the cost not including shipping?

Also, what would be the viability of putting a device that could accommodate 2 AA nicads instead, even locating them remotely on the board, and would there be any possibility of enclosing the batteries to prevent any possibility of leakage?

Sincerely,  Joe Garvey, Cork City, Ireland

mignolyx

Dear All,
don't you know if there's a way to modify a Blue Chips Axon AX 100 to MKII (i need it for single string transpose capability and for the use with the Terratec Editor)
Can you help me?

Thank you so much

Mignolyx

Pin

Quote from: mignolyx on January 17, 2016, 02:42:09 PM
Dear All,
don't you know if there's a way to modify a Blue Chips Axon AX 100 to MKII (i need it for single string transpose capability and for the use with the Terratec Editor)
Can you help me?

Thank you so much

Mignolyx

Yes, there is but I am not sure if still available. See:

https://www.vguitarforums.com/smf/index.php?topic=13362.0




Frandid

Hi

I think I have been asked by Elantric to write in this section.

I have started the process of updating my 3rd model Blue Chip Axon AX-100 SB into a Terratec MK2 version. I first replaced eprom 2.09 by MK2 eprom 4.28 which is the last one. It is OK. Everything seems to work. But my own sounds cannot be loaded in this new version. Not a serious issue since the sounds and presets can now be edited easily and quickly on my PC as my Blue Chip AX-100 SB now communicates with my PC.
Only one major issue : I wanted to update the firmware from 4.28 to v. 7.28, so I followed the manual's instructions :
"Press the "PRESET" + "UTILITY" + "EDIT" buttons while switching on the AX 100 MKII. The following appears in the display: "TERRATEC AXON FIRMWARE UPDATE"..."
But unfortunately when I press these 3 buttons and power on the unit this update display NEVER appears...
Is anybody here knowing why ???

Regards.

PS: I also replaced the sound card by one which is more compatible with MK2 version (better sounds and faster response but I don't think this has anything to do with the issue

Elantric


Frandid


admin

Member "Arkieboy" knows the most about Axon here:
https://www.vguitarforums.com/smf/index.php?action=profile;u=4775

Here is a link to AXON Internal memory backup Battery replacement ( likely issue in 2021 )
http://web.archive.org/web/20151002144859/http://me.ineptum.home.comcast.net/~me.ineptum/ax-100/battery/



Replacing the battery in an AX-100
The Axon AX-100 guitar-to-midi converter uses a NiCd battery to power its CMOS backup memory. Eventually this battery will stop working and you will lose whatever setup you had stored and you will have a hard time replacing the battery. Here's how I replaced mine.
The original battery
The original battery in my AX-100 was an Emmerich Z2A which is a 2.4V 65mAh NiCd battery. A Google search for "65mAh nicd" turned up some stuff in Poland and Germany and some other places, but nothing nearby. Try it yourself. Emmerich actually brags about producing non-standard batteries if you can believe it.
Update: It's now 2008. I clicked the "try it yourself" link above and behold, 65mAh NiCd batteries are now available for about $1+shipping on my continent, although they're 1.2V so it would still take two of them.
The original battery was soldered onto the main circuit board. This is bad design for a couple reasons: it is not convenient to replace, and if it leaks it may take out something on the board with it.
The new battery
Since I couldn't find anything on the same continent in a 65mAh NiCd, I went with two 1.2V 110mAh NiCd cells from DigiKey. I've bought a lot of stuff from DigiKey and like them a lot.
http://www.digikey.com/product-detail/en/P-11AAH%2FA7/P109-ND/6652?WT.srch=1&WT.medium=cpc&WT.mc_id=IQ65015717-VQ2-g-VQ6-40609416435-VQ15-1t1-VQ16-c

The two cells together make a 2.4V 110mAh battery.
The battery holder
The new cells are 1/3AA size. I couldn't find any battery holders for these, so I got an enclosed battery holder for 2 AA batteries at my local Radio Shack.
http://www.radioshack.com/product/index.jsp?productId=2062252&filterName=Type&filterValue=AA

I hacksawed out 2/3 of a AA battery worth of length, which was conveniently equal to the section between the two springs, and super glued the holder and cover back together after doing some sanding on the cut edges to make sure everything was straight and even and would fit back together nicely.
Unlike the AA batteries the holder was designed for, the new batteries do not have protruding buttons on the positive end to make contact with the holder's metal tabs, so I compensated by putting solder blobs on the holder's metal tabs. Now I have a very nice enclosed battery holder that shuts securely with a screw, and the batteries will be easily replaceable, and if anything leaks it should be contained.
The finished battery holder is really quite cute. Here it is with the cover removed:


   Photograph of shortened battery holder
Increasing the charging current
Since my new 110mAh battery was greater capacity than the 65mAh original, I thought it would be appropriate to increase the charging current in approximate proportion, since that seems to be what NiCd batteries want. The battery charges through 2K2 resistor R21. I paralleled it with a 2K resistor. This more than doubles the current, to maybe 2mA, which is still pretty weak since a real trickle charge for a 110mAh NiCd would be about 5.5mA.
One end of the new 2K resistor is soldered to the left end of D21, the other to the positive battery lead:
   Closeup photo of how new resistor was added


Installation

I spliced some wires onto the battery holder's wires and covered the splices with heat shrink tubing for insulation and strength, and bound the wires together with heat shrink tubing for neatness. I snipped out the original battery, left the leads for soldering to, and soldered away.
Here it is with the daughtercard removed:
   Overview photo of complete installation

Because I soldered onto the original battery leads instead of using a removable connector, I used velcro to attach the battery holder inside the case so it can be easily removed with the main circuit board should the need arise, e.g., to replace power supply capacitors.
Conclusion
Yes, it was time consuming, but I'm happy with the results.
Disclaimer
If you attempt this or any other modification to your equipment and mess it up, that is entirely your responsibility.


gearhound22

Just to be correct this was only for the bluechip ones and not the mk II and the ax50?

I couldn't see any batteries in either

Shingles

Quote from: gearhound22 on September 03, 2021, 09:01:31 PM
Just to be correct this was only for the bluechip ones and not the mk II and the ax50?

I couldn't see any batteries in either

Only the Blue Chip Axons have internal batteries.
The AX100 MkII and the AX50 do not.
Nik
--------------------------------
Tonelab, VG99, Axon AX100, EDP, Repeater
Godin, PRS, Crafter and Roland guitars
Center Point Stereo Spacestation V3