modification Casio Mt65... adding MIDI control?

[the_duckchild]

Hi, 

I have had a little search around and couldn't see any reference to this....

I have an MT-65 and found the service manual here;

https://www.synthxl.com/casio-mt-65/

It appears to state in pretty clear terms how the keyboard matrix is wired. I *think* this would allow it to be pretty easily retrofitted for MIDI using a teensy microcontroller or similar.
it also appears that the sound engine is actually controlled by the switches sending voltages to the keyboard matrix... this seems to open up the possibility of doing the kind of quick waveform switching that works very well on 8-bit chips of a similat vintage.

Just wondering if anyone has seen evidence of anyone achieving similar with old Casiotone keyboards? Or has anything else to share on this....This is probably getting towards the edge of my knowledge, but keen to have a crack at it and see what happens.....

[IanB]

It looks doable. The keyboard matrix is, er, a matrix, so you can't just put voltages onto it because it strobes one set of connections and reads with the other set, in sync. You could use either a MCU with open collector outputs (PICs do this, the Teensy doesn't seem to but I may be wrong) or read the strobes and set outputs accordingly which would be fewer connections so probably best.

 

Would be an interesting project

[the_duckchild]

Ah, thanks -maybe I need to do some more reading, I assumed it worked pretty much like an LED matrix but I guess it's actually that in reverse, pthinking about it... plus the synching bit?
the concept is right though I guess - I can get a microcontroller to perform whatever connection a keypress presents and  at least get note on/off?

 

[pianokeyjoe]

Have you guys heard of the Highly Liquid UMR2 universal Midi Retrofit version 2.0? The company and ALL information on this board is gone but some folks on github had the schematics and source code and released it publicly. Not to mention there are some sellers on EBAY and JAPAN that sell these boards both fully assembled and ready to use, and just the raw board for others to fit themselves. I bought one direct from Japan back in 2017, but when I could not find ANY info on how to hook it up and NO ONE seems to have that info even now in 2021, I just gave up and sold it forward. THAT board would be the answer you are looking for as it was originally made for the SK1/SK5 and then Highly Liquid made other version for YAMAHA keyboards and some obscure Casio keyboards who's matrices did not conform to the more popular SK1 scheme. Then he released the UMR2 which is the UNIVERSAL MIDI RETROFIT 2.0 and that board has hook ups for BOTH YAMAHA and CASIO and ACTIVE HIGH and ACTIVE LOW matrices. Being that if YOU hook up the data and control line wrong or if you get the ACTIVE HIGH/LOW wrong, you WILL fry the board, I did not want to chance my $66 purchase and just sold it forward and gave up until someday some one has the right hook up details to get this working. I HAD a Casio Tone 405 that was slated to use that board, but again, since I did not have the info, I gave up and sold BOTH the card and the board. I DO still have a MT65, and when I took it apart I saw to my dismay, that the keyboard contact boards are split and hooked up to 2 different chips, so.. not so simple to retrofit... Lack of info is the main issue here. I think this is the last bastion of Casio knowledge and goodness left on the net, so if no one here can help us, then it just may be a fruitless endeavor.

 

I have contacted some users in the past that did tell me how to midify an SK1 outside the Highly Liquid route but it was indeed more complicated and requires programming in assembly or something like that. Arduino programming and so on. Which I am not able to do as I am not a coder.. I am a musician and Carpenter.

 

Alot of the past info I HAD from Highly Liquid when they still existed, is gone sadly, due to hard drive crashes and old CDROMs delaminating over the years.

 

Maybe some one here can help us.. please...help us 😛

 

[IanB]

@pianokeyjoeSorry, don't know anything specific about that board

 

@the_duckchild Well yes it is kind of like an LED matrix in reverse. It cycles through a set of "strobe" (select) lines and each is connected to some number of keys, the CPU then reads the inputs to see if they are high. So you just need to use an input port on the MCU and an output connected to the CPU inputs. If it senses strobe 1 is high, it sets the outputs as if those keys are pressed or not pressed. To be really safe you could use a bank of analogue switches like the CD4066 with them connected across each keyboard switch (or other switch) since they don't introduce voltages into the circuit, only downside with that is that a 4066 only has 4 switches onboard, so that's number of switches divided by 4 chips which soon turns into quite a lot- 3 per octave. Also referring to Joe's fear of frying, you can prevent that generally by putting resistors e.g. 4k7 or so in series with voltage outputs from the MCU. You'd need to do some experimentation with just getting the MCU to "play" a single key to start with.

 

The analogue switch design is simple in operation and as I said the safest option. If you go with MCU outputs, an open collector system is safer but *may* not work depending on the CPU's input design.

 

Other than that it's just some MIDI decoding on the MCU using the onboard UART which is easy, I've got one sitting next to me doing that right now on a breadboard.

[pianokeyjoe]

@IanB Yeah, it is a obscure midi solution for classic Casio and Yamaha non velocity sensitive keyboards. They stopped selling and making them as a company back in 2015, and then they did the biggest disservice to all their customers and potential customers: They closed and wiped all forums and content! Seems that the MODEL specific hook instructions where only in their forum, and there is NOTHING online anymore, on the board's hook up instructions for specific Casio and Yamaha keyboard models. It sucks. I did find their github and circuit hub sites that are from 2015. he boards are for sale on Ebay but none of the sellers have info or documentation. So yes, it seems it will have to be a trial and error type of situation. Thanks for the 4.7K ohm resistor tip. I will try this again when I buy another 2 boards(they are $44 in the USA via EBAY now). I really DO want to midi the MT65 up as well as my slew of MT205 keyboards that I can and do sonically circuit bend, so those are worth it to midi up. I have the Highly Liquid UMR2 user manual but it is very basic. I can do a one key circuit at a time deal to see in which direction the keyboard works(active high, or active low).

 

The MT65 having it's on keyboard contact boards SPLIT between 2 chips, is still a concern. The MT205 is connected to ONE chip.

[IanB]

I'm just looking at the PCB and firmware and it looks pretty straightforward. There's a "learning mode" for teaching it which key is which and other than that it's waggling ports as I implied above. You could make it even simpler if you already know which keyboard it's going into. It would be easy to functionally replicate without creating any potential copyright issues.

 

And it uses 2k2 rather than 4k7 resistors, which is also fine. These things are pretty "whatever you've got"

[pianokeyjoe]

@IanBI have been searching for this stupid thing till my eyes grew tired. I do have the general instructions for it and ironically I found Model specific instructions for at least my SK1 and Yamaha PSS470, which is very important, since I own 4 or 5 PSS470 keyboards.. But the MT45/65 and MT205, no clue. The issue here is the Select lines installation, and the data lines installations. And lastly(and this part is very important), the Keyboard sensing polarity(is the matrix active high sensing or active low sensing). The jumper for the polarity of the keyboard is via a jumper of 3 spots of which you have - c + and you have to set that jumper BEFORE you power the unit on. So the basic question is: What keyboard polarity is the MT65 set to? The MT45? the MT205? Once that has been determined then it is matching the select lines and data lines on each keyboard with the numbered solder spots on the UMR2. I think today I have gotten much further than I had when I actually had the board. I want to buy the boards for the keyboards I have but I do need to have a proper hookup diagram for those models. I did not want to join yet another forum, but seems folks in muffwiggler actually have the answer! I will definitely share my findings in THIS forum as soon as I can get the goods! Cause we be Casio lovers here and we need the info. LOL! Wow! To illustrate my point, I shall attach 2 pdfs that show the hook up of the Yamaha PSS470(for illustration of the data lines and select lines and keyboard active sensing polarity purposes only), and the CASIO SK1. The SK1 was the original keyboard this product was designed for. The 3rd file is the keyboard matrix explained which has info on the active low and active high polarities and what to look for on the DIODES direction on the circuit boards. Now THIS FILE, has some KEY info. The Casio MT210 which has a similar engine to the MT65 but the keyboard contact board and ribbon cable are NOT the same. Non the less, like I said.. I am close..

Casio sk1_umr2.pdf Yamaha PSS-470 MIDI Retrofit Guide - Highly Liquid Forum.pdf kbd_matrix_explained.pdf

[IanB]

The MT65 is positive polarity, you can tell from the direction of the diodes. On the circuit diagram and hopefully the PCB, the input (select) lines are labelled KIx and the sense lines are KCx. Also the strobing select lines could easily be checked with an oscilloscope or logic probe. Or you could trace it with a multimeter/continuity tester; the sense lines are connected to one side of the keyboard switch, the select lines are connected to the tail ends of the diodes.

[pianokeyjoe]

11 hours ago, IanB said:

The MT65 is positive polarity, you can tell from the direction of the diodes. On the circuit diagram and hopefully the PCB, the input (select) lines are labelled KIx and the sense lines are KCx. Also the strobing select lines could easily be checked with an oscilloscope or logic probe. Or you could trace it with a multimeter/continuity tester; the sense lines are connected to one side of the keyboard switch, the select lines are connected to the tail ends of the diodes.

YOU sir, are a scholar and a gentleman! THANKYOU!!! Wow.. Well now I have some fun to do! LOL!

[the_duckchild]

great stuff everyone thanks! now the other side of a protracted house move and thinking about this again 



have put one of the UMR-2 copies into a Yamaha CS01 already so might give that a bash armed with this info! thanks.

 

[IanB]

@the_duckchildGreat, let us know how it goes!

[pianokeyjoe]

@the_duckchildyes please! Do post results and what connections you made, where both on the keyboard AND the UMR2 board? Pretty please. Thankyou!

[littleghost]

this might be the most informative UMR2 thread currently on the internet lmao. really just illustrates how totally Highly Liquid managed to wipe documentation on the thing

 

ive been desperately wanting to install a UMR2 in my Casio VL Tone but I can't find anything in terms of tutorials. it really kind of stings because Highly Liquid actually made a VL-1 specific MIDI kit at one point, with custom firmware and everything. god what id do to get my hands on one of those 😭

 

I finally managed to find a photo tutorial on archive.org's backup of their forums (which is a literal nightmare to navigate, btw. like close to impossible) and wouldn't you know, all of the images are missing. I think theyre archived somewhere, but theyre completely unsorted, meaning id have to blindly search through thousands of random images to find them, and then hope id even RECOGNIZE them if I did.

 

oof.

[littleghost]

oops, double posted on accident.

Edited January 1, 2022 by littleghost
oops

[pianokeyjoe]

36 minutes ago, littleghost said:

oops, double posted on accident.

I hear ya littleghost! I have a room FULL of Classic Casiotone keyboards that I have been collecting from fleamarkets and so on for years, and I was putting my trust in Highly Liquid's Midi retrofit kit when it was still sold as individual model specific kits, that someday I would afford each kit for each keyboard in the future. I did not see coming the eventual demise of that company and ALL the data online for their products! So yeah.. painful. Whats worse is the monetary investments into these keyboards only to have them collect dust now until I can find better documentation for the last product rev of the midi retrofit. I did buy one board all the way from Japan, but I sold it forward when I realized I could not hook it up with NO instructions whatsoever. So it is still on the back burner. I found the info I found, and posted what I could here for others and for myself, God willing THIS site does not go down, for future reference. Funny how our entire lives can be wiped out in a second at the stroke of a key.. Think about it.. Its almost like a twilight zone episode..

[IanB]

@littleghost If you post pictures of both sides of the VL1 PCB we can probably work it out. It's monophonic so it probably doesn't have diode steering to handle key rollover.

[IanB]

Funnily enough I'm actually right now working on converting a children's toy keyboard into a MIDI controller

[CYBERYOGI =CO=Windler]

Most interesting is that the MT-65 sound IC "D931C" can be software controlled. Robin Whittle identified in 1980th the (very counterintuitive) serial data format, so a microcontroller can be hooked to its 4-bit bus to define own preset sounds. Here someone made an editor for it:

https://wolfeffect.wordpress.com/casio-931-chip-editor

[pianokeyjoe]

14 hours ago, CYBERYOGI =CO=Windler said:

Most interesting is that the MT-65 sound IC "D931C" can be software controlled. Robin Whittle identified in 1980th the (very counterintuitive) serial data format, so a microcontroller can be hooked to its 4-bit bus to define own preset sounds. Here someone made an editor for it:

https://wolfeffect.wordpress.com/casio-931-chip-editor

Sadly the hardware editor and related software is NOWHERE to be found. No contact info, no name of the maker and no way to get this. The pictures are tiny small and no way to get schematics so far as I am seeing. But at least the Highly Liquid UMR2 midi retrofit board can make this keyboard a midi controller(in controller mode)or a midi controlled synth(in input mode). Would have been nice to have full control of the synth chip inside with a dedicated hardware/software solution but even in 2022 that is a pipe dream? Or is it?

[CYBERYOGI =CO=Windler]

You may need to port Robin Whittle's code to a RasPi or even Arduino to control the synth parameters through midi. It is a slow serial protocal likely not hard to interface with a cheap MCU.

[IanB]

If the code or just a description of the protocol are available I can write code to run on a PIC microcontroller, are these available?

[CYBERYOGI =CO=Windler]

AFAIK generic midi controller code for arduino exists. You would need to program the MT-65 keyboard matrix into it for accessing control panel keys. Keep all locking preset sound/rhythm keys unswitched (press one of them half down) to make midi selection work.