Author |
Message |
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
Posted: Sun Jan 19, 2014 12:58 pm Post subject:
Top-down octave divider keyboard Subject description: would like to add octave shift |
|
|
Hello,
Just a quick general question really: Is it likely to be a reasonably painless task to add an octave shift to an old string machine/synth that uses an organ-style top-down octave divider?
A bit more detail: It's a 4 octave Elka Rhapsody 490 string machine. I bought a PDF of the schematic a while back, but I thought I'd get a general idea before posting a load of info.
Thanks for any help. |
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
Posted: Sun Jan 19, 2014 7:25 pm Post subject:
Re: Top-down octave divider keyboard Subject description: would like to add octave shift |
|
|
tonewill wrote: | Hello,
Just a quick general question really: Is it likely to be a reasonably painless task to add an octave shift to an old string machine/synth that uses an organ-style top-down octave divider?
A bit more detail: It's a 4 octave Elka Rhapsody 490 string machine. I bought a PDF of the schematic a while back, but I thought I'd get a general idea before posting a load of info.
Thanks for any help. |
A few educated guesses having never seen the specific circuit in question, in increasing order of difficulty:
Option 1 - double/halve the top octave oscillator frequency
Lilkely to work but the TOGs and dividers may not appreciate being clocked wildly out of spec.
Option 2 - Add extra /2 stage before each divider note input
More likely to work than option 1 as the TOGs would still be clocked normally. But would only allow for lower octaves, not higher
Option 3 - Rewire all the divider outputs and make them switchable
Definitely would work but would be a wiring and switching nightmare. I shudder at the thought. See 4016/4066 datasheets
Option 4 - Highly Liquid UMR2 (or similar) midification
There's an outside change you might be able to MIDIfy the string machine and simply use a master keyboard with octave shift. But many (most?) string machines don't use a keyboard matrix so you'll need to find a suitable MIDI-to-keybed technology solution.
Would you mind sending me the PDF so I can give you a more informed answer?
Cheers,
A. |
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
Posted: Sun Jan 19, 2014 9:54 pm Post subject:
Re: Top-down octave divider keyboard Subject description: would like to add octave shift |
|
|
aladan wrote: | tonewill wrote: | Hello,
Just a quick general question really: Is it likely to be a reasonably painless task to add an octave shift to an old string machine/synth that uses an organ-style top-down octave divider?
A bit more detail: It's a 4 octave Elka Rhapsody 490 string machine. I bought a PDF of the schematic a while back, but I thought I'd get a general idea before posting a load of info.
Thanks for any help. |
A few educated guesses having never seen the specific circuit in question, in increasing order of difficulty:
Option 1 - double/halve the top octave oscillator frequency
Lilkely to work but the TOGs and dividers may not appreciate being clocked wildly out of spec.
Option 2 - Add extra /2 stage before each divider note input
More likely to work than option 1 as the TOGs would still be clocked normally. But would only allow for lower octaves, not higher
Option 3 - Rewire all the divider outputs and make them switchable
Definitely would work but would be a wiring and switching nightmare. I shudder at the thought. See 4016/4066 datasheets
Option 4 - Highly Liquid UMR2 (or similar) midification
There's an outside change you might be able to MIDIfy the string machine and simply use a master keyboard with octave shift. But many (most?) string machines don't use a keyboard matrix so you'll need to find a suitable MIDI-to-keybed technology solution.
Would you mind sending me the PDF so I can give you a more informed answer?
Cheers,
A. |
Ok, found one c/- Mr. Google.
Option 1 still a possibility; what is IC1 on Sheet No. 2/8? I can't read the part, looks like 7473? No idea what the maximum clock rate for IC2 & IC3 (AMI S2555 and S2556) would be.
Option 2 would definitely work. You could use 3 x SAJ110 and insert them in circuit between board M000419 and the 6 x M000420. You could even look at using 12 x PLLs to provide octave-up functionality. Max input frequency on SAJ110 is 50 kHz so you should be OK for one or two octaves up.
Option 3 would require you to rewire all the Kn and Yn lines on the M000420s (6 x 9 = 54 of them) into a switching mechanism. You wouldn't get K and Y lines for an octave above though, so this is probably not useful.
Option 4 you couldn't use a matrix; each key is individually switched so you'd need 54 individually switched lines - but the service manual I have doesn't show a wiring diagram so you might be unlucky and it could be a multiple of that (108 or 162 switches etc) depending on how the waveform shaping is done (i.e. square wave output, individual per-note filters or waveform approximation using summed squares?)
The service manual I found is horrible quality; if you have a good one I'd love a copy of it.
Cheers,
A. Last edited by aladan on Sun Jan 19, 2014 10:03 pm; edited 1 time in total |
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
|
Back to top
|
|
|
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
Posted: Mon Jan 20, 2014 4:32 am Post subject:
|
|
|
aladan, I appreciate all your help, thank you!
I think I have the same schematic as you, I just searched and found the one I have here. I couldn't find one a few years ago and bought this from someone online.
I've had a look at the 419 board and IC1 is a 7473 as you suspected. Option 2 seems to be the way to go but I have to confess now, I'm not as advanced in electronics as I'd like to be and would need help with the mods required. This is to let you bail-out now if you like!
I'm definitely able to do the work, I'm okay with creating circuits on stripboard and the like, just not designing them to begin with. Perhaps I'll try just the octave-down to start with by including the extra divide-by-two into each M000420 board. What would you suggest my next move should be?
Thanks again for you help, and I completely understand if you unable to go deeper into this. |
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
Posted: Mon Jan 20, 2014 4:14 pm Post subject:
|
|
|
tonewill wrote: | aladan, I appreciate all your help, thank you!
I think I have the same schematic as you, I just searched and found the one I have here. I couldn't find one a few years ago and bought this from someone online.
I've had a look at the 419 board and IC1 is a 7473 as you suspected. Option 2 seems to be the way to go but I have to confess now, I'm not as advanced in electronics as I'd like to be and would need help with the mods required. This is to let you bail-out now if you like!
I'm definitely able to do the work, I'm okay with creating circuits on stripboard and the like, just not designing them to begin with. Perhaps I'll try just the octave-down to start with by including the extra divide-by-two into each M000420 board. What would you suggest my next move should be?
Thanks again for you help, and I completely understand if you unable to go deeper into this. |
Ok, here goes. Some of this may be obvious but it doesn't hurt to say it for others who may be reading so I hope you don't mind.
You will need reasonable soldering/electronics skills. If you don't have these skills then it would be best not to attempt this as your Elka could be damaged by unskilled efforts. If this is the case, I recommend spending a few months practicing on DIY synth PCBs like http://cgs.synth.net or MFOS or similar first.
Before you start, and during the process, get a digital camera and take copious photos with clear and complete detail so you can refer back at each step and know where wires and pins were connected before you started. I use my phone so I always have them handy.
Page 4 of the PDF (Sheet No. 3/ shows the SAJ110 octave dividers IC4 on the six M000420 boards (each handling two notes). Are those ICs socketed? If yes then very carefully bend pins 2 and 4 out and solder fly wires on to them. On the bottom of the board, solder fly wires onto pins 2 and 4 of the socket. This is where you'll be inserting the additional divide by 2 (or 4) stage. If not socketed then you'll need to find some other way to carefully insert your flywires between the collectors of T5/T6 and IC4.
Find yourself some SAJ110s and breadboard/prototype up a circuit similar to the the IC4 section of M000420 with +10V/GND power from the M000420 board and the seven resistors to GND on the outputs. Use pins 2, 3, 4, 5, 6 as inputs (coming from the IC4 socket flywires on boards M000420) and pins 14, 12, 10, 9, 8 as the corresponding outputs (going to the pins 2 and 4 of the SAJ110s on the M000420 boards). With 5 divide by 2 stages on each SAJ110 and twelve notes you'll need three SAJ110. You'll have three unused divider inputs left over on SAJ110s which should not be left unconnected so just connect them to one of the outputs so they do something normal.
Damn, I just noticed the two direct lines from the collectors of T5 and T6 (K and Y/16(misprint? check the boards)) going to C47 (near T15) and C43 (near T13). You'll need to connect them to pins 2 and 4 of IC4 (respectively) as well, as they'll need to come from the additional divide by 2 stage you're adding too.
I found a readable SAJ110 datasheet in a 198-page document called "ITT Integrated Circuits for Consumer Applications 1977/1978" somewhere online. I didn't find a readable SAJ110 datasheet anywhere else but I didn't look that hard. I did find a pretty-muchy unreadable one.
You'll need a 12P2T switch to switch the extra divide by 2 stages in or out. I would cannibalise one of those massive ones from an old PC console or printer A/B switch.
Now, if SAJ110 are too hard to find you might be able to use more modern chips, as long as they're happy to run on +10V (e.g. CMOS not TTL). Maybe a 4520, but that's only a dual so you'd need six (one per M000420?). Or you might be able to use three quad flip-flop configured cleverly with some extra circuitry.
Yes, start off by building up the divider board like you suggest. Get it working on the borrowed power and using inputs from IC4 pins 2 and 4, then figure out how you're going to do the actual octave switching (the fiddly part).
In case you're wondering why I didn't suggest using the K and Y connectors on the M000420 board directly, they look like they are 0V/+27V which would be a pain to work with (coming from the old school high voltage AMI TOG chips IC2/IC3 on board M000419). I would measure with an oscilloscope to confirm. T5 and T6 convert the 0V/+27V signal down to 0V/+10V which is easier to work with. If you duplicated those voltage converters on your divider board you could probably use the K and Y wires though, hmmm, that might be overall easier actually! The only question would then be whether or not the +10V output from your divider would be enough to switch T5 and T6. At worst you might need to change a few resistors if it isn't.
So not a project for a beginner, maybe a 6.5/10 difficulty rating
Sorry, some of this is stream-of-conscious and not necessarily 100% easy to follow. I should draw some schematics instead!
Cheers,
A. |
|
Back to top
|
|
|
Boerge
Joined: Sep 02, 2009 Posts: 80 Location: Germany
|
Posted: Tue Jan 21, 2014 6:57 am Post subject:
|
|
|
Hi,
the problem I see in your plan is: if you shift the complete keyboard by one octave you should know, that the filters will act different than: lowpass filters will have lower output level if you shift up and vice versa. That means: you will not only change the octave, but also the sound! Don't know how the filters in your Elka work, but in my old Vermona-organ there is one LP filter for each half octave of the keyboard to get nearly sine waves, one octave shift would result in too much harmonys (when shift down, more squarewave-like) or too less output level when shift up.
Good luck!
Boerge _________________ ...ich will doch nur löten... |
|
Back to top
|
|
|
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
Posted: Tue Jan 21, 2014 12:59 pm Post subject:
|
|
|
Thank very much aladan, that is very thorough! Looks like I would have to try and find some SAJ110's or equivalent before anything else. They are socketed by the way.
I'm not one of those people who can picture things in their mind (never been able to understand what people mean when they say that actually!) so I'll have to work through your post and make diagrams as I go. However, having just read Boerge's post (thanks Boerge) I'm now wondering if this will be very successful. I'll have a look out for those SAJ110's first I think. Not much luck here in the UK on my initial search, will look into alternatives like you mentioned.
Thanks very much again, wasn't expecting such an in-depth reply, much appreciated. |
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
Posted: Tue Jan 21, 2014 3:45 pm Post subject:
|
|
|
Boerge is probably right
Looking at the table of capacitors (even numbered from C30 - C46) on Sheet No.3/8 it looks like these provide the tuning of the filter circuits for each key. If you pitch down like I suggested you'll get more harmonics on each key; a more 'buzzy' sound. This may or may not be acceptable to you
However, looking at how the filter circuits work has helped me understand the key mechanism better. I now think your best option would be a MIDI retrofit, so you can play the Elka from another (MIDI) keyboard that has an octave up/down capability. Note that you won't get an extended range doing this - the Elka will still only respond over the existing range of 49 notes. I'm not sure if that's what you're after or not.
One way to do this would be 3 MIDI-chained Highly Liquid MD24 MIDI-to-trigger interfaces with two of them covering two octaves of notes, and one covering the 49th note - the final top C. Each output on the MD24s would drive a transistor switch and when that note was triggered via MIDI it would act the same as pressing the corresponding note on the Elka's keyboard by shorting that note's K or Y input to the M000420 board to +27V. Since the key switch voltage range is 0V to +27V you can't use a switch IC like a 4066 (as they are limited to +15V.) So you need a board with 49 transistors and 98 resistors.
Two of the spare outputs on the third MD24 could allow MIDI control of the Violin/Cello and Strings switches, and maybe the volume and sustain controls (but that would probably be a bit tricky as you'd need to use the MD24's PWM mode to generate a CV to drive a voltage controlled resistor).
There are probably other MIDI interface boards out there that would do this job, maybe even better. I only mention the MD24 because that's one I have personal experience with so I know how it could be made to work.
Cheers,
A. Last edited by aladan on Tue Jan 21, 2014 5:30 pm; edited 1 time in total |
|
Back to top
|
|
|
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
Posted: Tue Jan 21, 2014 4:17 pm Post subject:
|
|
|
Thanks aladan. It is the octave shift that I really wanted to do. Adding MIDI would be an interesting thing to do, and it so happens that I have an MD24, but not three of them. I'll have to give it some thought then as to whether it's worth doing the octave shift if there's the possibility that it won't sound that great. I haven't had any luck with the SAJ110's either.
Thanks again though for all your help. |
|
Back to top
|
|
|
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
|
Back to top
|
|
|
aladan
Joined: Aug 13, 2011 Posts: 52 Location: Australia
|
Posted: Sun Mar 16, 2014 4:37 pm Post subject:
|
|
|
tonewill wrote: | There is one other thing that I wanted to do with this keyboard that I'm sure will be far easier. I would like to be able to disable the modulation via a switch. I know there are 2 modulation circuits, what would be the best way to do this please? I've attached the overall schematic of the modulation section. Is it a case of grounding the signal somewhere?
Many thanks for any help. |
Sorry for the delay on this response.
I'm really not sure how to give you what you want - I don't completely understand the modulation circuitry. I think all LFO pitch mod is purely done by the BBD circuitry - looks like there's no modulation at the TOG which I would have expected to be much easier to implement. But I'm not completely sure. Maybe try varying the "depth control" trimmer on SPM61-8 (page 5/ and see what affect that has? If it does what you want then you could bring it out to a panel switch.
If only I had a 490 of my own I might be able to be more helpful!
Cheers,
A. |
|
Back to top
|
|
|
tonewill
Joined: Aug 21, 2009 Posts: 135 Location: England
|
Posted: Mon Mar 17, 2014 3:32 am Post subject:
|
|
|
Thanks a lot aladan, much appreciated. |
|
Back to top
|
|
|
|