Highly Liquid Forum facebook twitter google plus rss feed

Go Back   Highly Liquid Forum > General Discussion > MIDI, Electronics, & Everything Else

Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 06-16-2012, 02:49 PM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default help with implanting a rotary encoder into a Korg Poly 800?

Hey John and all,


I was hoping maybe I could pick your brains on how to go about and add a rotary encoder to my Korg Poly 800.

Below I have a couple snapshots of the Korg's schematics. What I want to do is place a rotary encoder in parallel with the UP/DOWN switches I have boxed in green.

That way I can either use the original up down switches, or turn the encoder knob to make the value of a parameter go up or down.

It looks like the switches are in a matrix. The IC it goes to on another board (via a group of wires called "SW COLUMN")) is a tc40h240p.. an octal buffer I believe. Again, the incoming connection from the switches are boxed in green. It would be 3-4a & 3-5a.


So, would I be able to get a rotary encoder, decode the pulses with a simple flipflop IC and direct connect the flipflop outputs to those switches?


.
.
.
.
Attached Images
  

Last edited by Synthetech; 06-17-2012 at 01:34 AM.
Reply With Quote
  #2  
Old 06-18-2012, 02:40 PM
John's Avatar
John John is offline
Moderator
 
Join Date: Jan 2009
Posts: 3,007
Default

Hi Seeker,

That's tricky. Since the circuit is designed for regular switches, it might be tough to insert an encoder. I guess you could try wiring the switches that form the encoder in parallel with the up/down switches, but I'm not sure that would work.
Reply With Quote
  #3  
Old 06-18-2012, 11:00 PM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Default

I think you will have to add some nand gates to the process. Normally a cpu determines the direction of travel by whether the B output is leading the A output or vice versa. so B and not A would be one direction for counting and A and not B would be in the other direction -- you could use the logic as an output to a 4066 device whose switches parallel the keyboard switches
Reply With Quote
  #4  
Old 06-19-2012, 01:28 AM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default

If I used a 7474 Dual D-type flip flop like shown below, could I hook the COM on the encoder to what appears to be ground at SW LINE in the first pic above?

Then using the flipflop (and some debouncing circuitry), couldnt I hook the flipflops outputs to the cathode side of the diodes at SW 12 and SW 13 (respectively for up/down.. CW would be up, CCW would be down).


I only think I'd need to power the FlipFlop IC itself.. use the same ground as the COM on the encoder is using.. borrow my synth's PS +5V to provide the IC power...

I'd think it'd work, but I thought to ask your opinions on it. It appears to me that the UP/Down switches shown in the Korg's schematics are actually switching a ground signal (low signal) to that Octal Buffer.
I think the Octal Buffer is probably polled by the Korgs MPU, then reset after it's polled it.

I'm not sure, I'm very green in the digital electronics field. But I'm trying to self study TTL/CMOS and/or Nor/Nand Xor/Xand logic gates and inverters.


Here's a pic of a 7474 hooked up to a simple A/B/COM encoder


Reply With Quote
  #5  
Old 06-21-2012, 05:47 PM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Default

On those pieces of schematic, I am not seeing where 3-9B and 3-10B are connected to grd.
Reply With Quote
  #6  
Old 06-22-2012, 03:22 AM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default

Quote:
Originally Posted by Jim McDougall View Post
On those pieces of schematic, I am not seeing where 3-9B and 3-10B are connected to grd.



3-9B and 3-10B are difficult to find, because there's a typo in the schematic. Look at SW LINE 3-9B and you'll see a second 3-9B... then look further down to 3-10B and right next to it is 3-10A... the 3-9B above should be 3-9A (right terminal)


It appears to be hooked up to a TC40H138 3 to 8 Line Demultiplexer

PDF datasheet here:
http://www.datasheetarchive.com/TC40...atasheet.html#


3-9A and B appear to be hooked up to Y7 output of the 40H138 and 3-10a and B look like they are hooked up to Y6


So my guess is that when the MPU provides a "011" (#6) at the ABC select terminals of the 40H138, it will be providing a low signal to 3-10A & B (Y6). When it provides a 111 at ABC, it will provide a low signal to 3-9a & B (Y7).

When a switch is pressed, it passes that low signal(when it's provided by the 40H138) over to the 40H240 octal buffer.. which will store a bit to hand over to the MPU when it polls that buffer, which it will reset that buffer when it polls it.
Then the cycle starts over again, right?



So would there still be a way to sneak an encoder in there? That 3-8 line demux is what's got me scratching my head. It might not like it if I place an encoder in there and try to use the 3-10A or 3-9A terminals for COM.



Here's the full schematic to the Korg.


http://www.synfo.nl/servicemanuals/K...ICE_MANUAL.pdf


It's the MkII version of a Korg Poly 800, but the main board we are looking at is identical.
Reply With Quote
  #7  
Old 06-23-2012, 03:52 AM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Default

I think there is a error in the schematic showing the switches. On the right side there is a long buss that has both 3-9 A & B and 3-10 A&B connected from the same buss line.

However if you look at the left side you see the lines that are the straight through connects to that buss starting at 3-9B (D9 - D15) and then you have another set of lines from the chip that drop down and connect to that same buss below 3-10B (D1-D8)

Logically the scanning process would not be able to differentiate between a switch press on a straight through line D15 and the switch press on the drop down line from the same terminal D8

So I believe that there should not be a connection bewteen the buss at point Sw8 8 and the point Sw9.

This would only seem to work if this link was broken like the link between Sw15 and Sw16 and 3-1 through 3-8 were data lines and 3-9, 3-10 and 3-11 were select lines.

the only way to verify would be to test/inspect the board itself.

If it is a matrix, then I would take a slightly different approach. I would parallel two sections of a 4066 type chip with the switch contacts and then use external logic with the encoder to drive the control signals to the 4006 control terminals. I think your flip flop logic would work as you are reaaly creating a dual output nand gating. This is something you can breadboard to experiment externally to the synth and when working it would be easy to wire in.

If it was not a matrix, I would still take the above approach. Lower risk !

Last edited by Jim McDougall; 06-23-2012 at 03:53 AM. Reason: typos
Reply With Quote
  #8  
Old 06-23-2012, 02:07 PM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default

Funny you mention the error 3-9 and 3-10 scan groups. I wondered about that myself after my last post.

Here's a snapshot of a MkI Poly 800 schematic.. it has a break between the two as we suspected. Sorry I didnt post that in the first place, I coulda swore the MkII schematic was going to be exactly the same (and the MkII pdf schem. is MUCH clearer to read, thats why I used it in the 1st place). Evidently its the MkII schem. that has the errors.. and more than just the two we saw, I suspect.






Fascinating for me to comprehend, at last, what the Korg is doing inside to scan all the buttons and keys. Evidently they designed it to be efficient with minimal bus lines.. it appears that it is using one 8 bit bus for 8 different groups of switches. So this design appears to be scanning 64 different switches in total. I believe the Korg has approx. 68 switches total (control switches and the entire keyboard keys). So this baby is doing the work of just about all of them!


That means switches 12 and 13 are inactive for the duration of 7 other group scans and only active when Y6 on the 3-8 demux is activated.

So does this mean the encoder's COM connection should be hooked up to the 3-10a/b terminal? Or would it be that the 4066 (quad bilateral switch) have it's Vss hooked up to 3-10a/b and have the encoder's COM grounded.. along with the flip/flop's Vss.
Or.. should it be the flipflop that needs to output the low signal? It appears that the signal I need to be passing on is a 0.5V "low" signal to the oct. buffer.. evidently the buffer must have inverters at it's input to change the low to a high.

Obviously, I dont want the encoder's decoding circuit to pass on a signal when it's not that group's (Y6's) turn to send a signal to the octal buffer, right?

My concern is that the 4066 or the flipflop IC might not like being switched on/off so fast during the MPU's scanning routines.
I'm also assuming I need to keep all power/ground connections within the main boards PS.. meaning if I breadboard this up, I need to use the Korgs PS, not a separate supply like the one in my powered breadboard.




Sorry I sound so green on this.. it's all new to me.

Thing is... if we manage to make this work, it would probably be the very first time (known on the internet anyway..) that a Korg Poly 800 would have a rotary dial for editing parameter values. Most people absolutely HATE those silly UP/DOWN switches.
Alot of people would be thrilled to finally be able to twirl a knob on their old Korg instead of pushing those buttons.

-Blaine

Last edited by Synthetech; 06-23-2012 at 02:28 PM.
Reply With Quote
  #9  
Old 06-24-2012, 02:03 AM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Default

Here is roughly how the 4066 would connect

Use the 74HCT4066 as this is designed for single end 5V use -- the CD4066 uses a bipolar supply.

The switches are isolated on the chip so you can test on the breadboard. Wire a dropping resistors and leds to watch the switchs working.

The E inputs are the control inputs. 4066 were often used for audio switching so there is no need for the switch side to have any commonality with regard to voltage to power the chip itself.

You could implement the chips on a piece of perfboard and then just wire into the switch positions and the Korg 5V supply

I am more of an old analog guy than a digital expert (emphasis on 'old') However I can describe the logic you need to implement and you can go from there.

Referring to your diagram with the pulse trains from an encoder, let's look at the logic required. Assume we are starting between the pulses so both A & B are low.

As we turn the dial clockwise (up) then we get the leading edge of the A pulse and use that to set the E1 input high - closing the up switch on the synth. As we continue then the B pulse will start but because A is already high we need to inhibit the B pulse from raising the E2 input high.

Contining along, the A pulse will trail but the B pulse will still be high - at this point we need to set the E1 input to low but there is no change to the E2 input - still low.

Eventually both pulses go to low and we reset the logic for the next cycle

Turning the other way is the same logic -- sub B for A and A for B in the above logic. Walk this logic through your flip-flop diagram and see it will hold up for both cases. I think it will based on your drawing of the logic. Worst case, you might have to add some logic gates.
Attached Images
 
Reply With Quote
  #10  
Old 06-24-2012, 02:06 AM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Smile

This circuit is small enough you could implement it on a piece of breadboard that it held to the encoder by the solder connection on the bottom/side of the encoder. Done with SMD chips, the circuit would be smaller than the bottom of the encoder
Reply With Quote
  #11  
Old 06-24-2012, 12:54 PM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default

Ohhhhhh! IC, a bilateral switch.
That's what I was thinking of doing the first time I tried to figure this out... but I got sidetracked on understanding how the Korg scanned the switches in the first place.
Guess I've come full circle with this project.


Would a 74hct74 dual d-type flipflop be a direct compatible IC to the 74HCT4066?


Thanks so much for taking the time to help me sort this out. I'll probably get this up and working with little effort.. just need to buy more stuff to do it.




Quote:
I am more of an old analog guy than a digital expert

you, me and Joe Walsh...

http://www.youtube.com/watch?v=RLaQUU_VLMk

Last edited by Synthetech; 06-24-2012 at 01:05 PM.
Reply With Quote
  #12  
Old 06-24-2012, 03:10 PM
Jim McDougall Jim McDougall is offline
Moderator
 
Join Date: Aug 2009
Posts: 395
Default

Yes they are the same family -- if you run both using a common V+ and ground then they will work together very nicely. The HCT version of the 4066 is more than fast enough to handle a pulse train from the encoded and had a very low reisistance across the switch contacts.

I think the biggest problem is going to be how fast the Korg CPU cycles through its scans to test the switch's status. One of the ways to manage this is to use an encoder that is 32 PPR or less (32 pulses per revolution) This means that incrementing a parameter from 0 - 99 would take 3 - 5 revolutions but that is a lot faster than pushing the button and you minimize the risk of the Korg scanning circuit missing switch closures and is easier to dial in on a specfic paramater value -- less sensitivity to jittery fingers !
Reply With Quote
  #13  
Old 07-29-2012, 12:44 AM
MikeK MikeK is offline
Junior Member
 
Join Date: Jul 2012
Posts: 2
Default Korg Poly 800 mod

If you get this mod working please put up all the details as I would like to give it a go with my Poly 800. I may just give it a go anyways if I could get a few pointers as I am pretty new to all this.

Mike
Reply With Quote
  #14  
Old 09-09-2012, 04:10 AM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Exclamation

Well I finally got around to getting some 7474 D flipflops and a 4066 quad latch.


I had some issues in the beginning.. but I finally got around to making it work.

look for a post of vids of it in action soon....

Last edited by Synthetech; 09-14-2012 at 08:02 PM.
Reply With Quote
  #15  
Old 09-14-2012, 08:25 PM
Synthetech's Avatar
Synthetech Synthetech is offline
Moderator
 
Join Date: May 2012
Location: USA
Posts: 711
Default

FINALLY!!!!!


A HUGE thanks goes to Jim for his suggestion to mate a 4066 bilateral switch to the encoder.

As hoped, this setup works quite well!

See the vids.



http://www.youtube.com/user/Synthetech


it's still in development stage. I need to fine tune the debouncing circuit so it will minimize jitter but keep up with the turning action better.. it's a combonation that will require me to order different capacitors/resistors than the ones I had in stock at the time of this post...
Hopefully I can reach a decent compromise.

But anyway, I'm really glad this works... now to finish it up, put it on a protoboard and physically mount it on the keyboard.
Reply With Quote
  #16  
Old 09-18-2012, 05:03 PM
MikeK MikeK is offline
Junior Member
 
Join Date: Jul 2012
Posts: 2
Default Nice!

I am trying to replace an LCD on my Kurzweil at the moment but I would like to try this mod when you get it all ironed out.
Reply With Quote
  #17  
Old 02-11-2016, 06:12 AM
PUllrich PUllrich is offline
Junior Member
 
Join Date: Feb 2016
Posts: 1
Exclamation my little datadial that does the job

Hi!

Reading your discussion I can tell you that my datadial
http://synpro.heimat.eu/datadial.htm

should be exactly what you need. Lots of people have installed them, some of the installation pictures can be found on my webpage.

And yes, my design also uses a 4066 to be switched in parallel to the plus/minus buttons in your synthesizer.

My datadial woorks in the range from 3.3V to 5V - so it should fit in most synthesizers.

Ciao
Peter

http://www.ullrich.at.tt (My homepage)
http://synpro.heimat.eu/datadial.htm (Synth addon)
http://rc100-usb.at.tt (Roland Sampler tool)
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT. The time now is 03:57 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2017, vBulletin Solutions, Inc.