Digitizing a 2720-8 keyboard with EK-3 board
Page 1 of 2

Author:  dnigrin [ Sun Jan 25, 2015 9:01 am ]
Post subject:  Digitizing a 2720-8 keyboard with EK-3 board

I've been restoring a friend's 4700 system, and am now trying to get the keyboard he has working with it as well. The keyboard seems like a 2720-8, that has been retrofitted with some custom digitizing circuitry for the note CVs, but which I can't figure out. A lot of the information (including links to images) is summarized on this Yahoo Paia group thread: - if you're not a member of that group, here's a pdf of the entire thread:

I'm at the point now where I don't think the 6 digital outputs coming from the custom board will work with the 8780 module I have, as outlined in the thread above. I have gotten ahold of an EK-3 board, so was planning to remove the custom digital board and the 2720-8b analog board, and replacing them with the EK3. I have a few questions though, and in general would love any tips I would need to know to do this replacement:

1) The EK-3 document mentions cutting the bus rods at specific places. But when I look underneath the keybed, I see *two* bus rods. Which one is meant to be cut? Or both? See photos below.

2) The EK-3 document also mentions adding a diode in series with each key switch, and creating a matrix or something. One of the posters in the thread above mentioned a Paia "Diode Matrix PCB" - is this something I need, and if so, is it still get-able from Paia? Along similar lines, it's not clear to me where the individual keyswitches connect to the EK-3.

Here's images of the underneath of the keybed:

Thanks in advance for any and all help!


Author:  toneman [ Sun Jan 25, 2015 3:24 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

Hey Dan!

Have U read "Friendly Stories About Computers/Synthesizers" by Simonton?

The multi-part article explains how the (digital)keyboard scanning is accomplished and how it is implemented.

It's a must read.

Don't know if Scott posted it at this site somewhere(?)

Long time ago,
I uploaded pdf's of the Friendly Stories Lab Notes series to the Yahoo PaiaGroup. But, I think the moderator of the group is gone(?)

Just remembered....
there is a copy of the EK-3 info here:

Keep That PAiA PHlying!!

Author:  dnigrin [ Sun Jan 25, 2015 7:58 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

Thanks for your reply, but yes, I have read all those things (and I mentioned in my post that I have the EK-3 document already). The problem is that my electronics knowledge is very basic, and even though I can roughly understand the documents you mention, I still need a bit more "cookbook" style help.


Author:  toneman [ Sun Jan 25, 2015 10:35 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

I just quickly scanned that PaiaGroup pdf, but there were a lot of repetitions that I got brain flooze.
Re-re-looking at the jpgs of the keyboard,
I don't see the diode matrix pcb(?) Also, the keyboard buss looks like it has NOT been split into octaves since there is no clear tubing supporting the buss bar together after it's cut. The wiring from the split keyboard buss should goto the "input" of the diode matrix. The "output" of the diode matrix goes to the keyboard scanner (EK-3). Then the scanner outputs goto the D/A. Without the split buss AND diode matrix pcb, the scanner can't work properly.

The digital keyboard case should have 3 pcbs: scanner, matrix, and 5V powersupply.

Since both buss bars look intact, Maybe you could consider restoring the keyboard to it's original analog configuration or, maybe make a linear analog keyboard for V/oct stuff?

Other than that, I don't know what else to suggest.


edit: jpg4252 seems to show a bunch of diodes at the edge; possibily the "missing diode matrix"? there are no '154's(1-of-16) on the EK-3, so this is a total DIY invention. Still doesn't seem like either of the busses were cut at octave intervals for the binary decoding.

Author:  dnigrin [ Mon Jan 26, 2015 7:59 am ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

Thanks again toneman.

So one thing you've mentioned helps me a lot - the confirmation that I need the diode matrix pcb, and that the wiring from the keyboard goes to it first, and then on to the EK-3. This was basically my question #2 in my first post here.

I already knew that my bus rods had not been cut, which is why I asked my original question #1. Could you comment on that - which bus rod do I cut?

I have no interest in trying to keep this all analog, I'd still need to source the missing analog board within the keyboard, and I have the 8780 D/A converter, and I've read that there's far greater stability in using the digital scanning approach.

Author:  toneman [ Mon Jan 26, 2015 11:17 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

U didn’t realize there was a “diode board”(?) It’s easy to see why… on….
For the following comments, please refer to the EK-3 article and the “Friendly Stories…” article. Pages and figures referred to are the actual page numbers at the bottom, not the pdf page number.
OK…..? ok...
It’s easy to see how U might not see the “diode board”. If U Refer to the EK-3 article, there is NO diode matrix on the pcb. Yet, Figure10 shows “keyboard” and “encoder” blocks, but no “diode matrix” block.
I can’t remember if my 8782 has the scanner pcb with or without the matrix(?) It’s been too long.
If U refer to the 8782 assembly drawings, figure8 shows the larger scanner pcb WITH the diodes.
Look on this page for an example of the scanner pcb WITH the diodes.
I think this is a DIY clone pcb not an original(?)
In both articles, the cutting of the keyboard voltage buss is explained. The cuts are at every 8 keys, Not every octave as I stated previously. See figures 7 & 8 in the Friendly Stories article, and the same Figures in the EK-3 article.
I had to re-read all of the digital keyboard articles since its been almost 30yrs since I assembled my P4700J.
Most of the Pratt&Reed keyboards have 2 busses:
A gate buss and a voltage buss. The voltage buss closes first, then the gate buss. I believe U only need To cut the voltage buss to make the keyboard matrix. Clean the buss first with some tarnix or something To remove the oxidation. Avoid steel wool as this will leave metal strands that will be hard to clean out.
I hope this info helps in your quest to restore your PAiA gear. Good Luck!! 8-)

Author:  dnigrin [ Fri Jan 30, 2015 8:59 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

Very helpful info, thanks again!!

Author:  toneman [ Fri Jan 30, 2015 11:29 pm ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

found a another pic of an 8782A keyboard pcb.

it has the INfamous PAiA blue board. ... 1600/4.jpg

Author:  dnigrin [ Sat Jan 31, 2015 7:44 am ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

Helpful again, thank you. So this image shows a board that is essentially the combination of the EK-3 board, and the "Diode Matrix" board, correct?

Author:  toneman [ Sat Jan 31, 2015 10:51 am ]
Post subject:  Re: Digitizing a 2720-8 keyboard with EK-3 board

this is the pcb referred to in figure 8 in the 8782A digital keyboard assembly supplement manual. Figure 7 of the same manual shows the up-side-down keyboard and the 2 buss bars and the pieces of support tubing needed after the rear buss bar is cut. Figure 6 shows the mounting position of the transformer and pc board and the 8700/front panel. U cut the buss bar every 8 keys to make a binary grid or matrix with the diodes so that each key has a unique binary value. When a key is pressed, the EK-3 stops scanning and the binary value of that key is available at the EK-3 output. This binary output goes to the D/A's 25pin cable. The D/A has LEDs on the front panel that actually shows the binary value!! this is so cute as well as so helpfull !! 8-)

Page 1 of 2 All times are UTC - 6 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
Not able to open ./cache/data_global.php