| Author |
Message |
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
 Posted: Fri Oct 17, 2008 11:08 am Post subject:
|
  |
|
| Exactly all I needed to know - thanks for taking the time to answer! |
|
|
Back to top
|
|
 |
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Fri Oct 17, 2008 12:21 pm Post subject:
|
|
|
| One more question - I notice 4 FM inputs - is this 1 exp and 1 linear for each osc? If so, which one is the exp input and which is the linear for each osc? |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Fri Oct 17, 2008 4:43 pm Post subject:
|
|
|
| The pink dot FM inputs all look exponential (notice the 100k resistors are connected to the same node for each VCO). |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Fri Oct 17, 2008 4:50 pm Post subject:
|
|
|
Yeah, I couldn't tell the difference between them either. So I'm guessing each osc has 2 exp FM inputs?
Thanks! |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Fri Oct 17, 2008 8:27 pm Post subject:
|
|
|
Sounds right, check out the Odyssey it has 2 sliders for 'FM' per VCO (with selectable modulation sources).
 |
|
|
Back to top
|
|
|
peng
Joined: Feb 23, 2006
Posts: 111
Location: Chicago
Audio files: 2
|
| Posted: Fri Oct 17, 2008 11:51 pm Post subject:
ARP Odyssey Oscillators |
|
|
| AndyR1960 wrote: | | To be honest, I had no idea that Tim Servo was also working on these VCO's (as we say in England - you wait all day for a bus, then two come along at once).Andy. |
Umm... I just built a clone, too...before I even knew this thread was going! It turned up in my Google search for "Odyssey selected FET" !
| AndyR1960 wrote: | | There is something "special" about the Oddy VCO/RM combination |
I agree. That is one reason why I built this clone. The other reason being I had two Odysseys here to repair and one with a non-working VCO. Thought it would help in troubleshooting (and some fun).
I've built three and all are different. These are on the breadboard and may be suffering from interference of some kind. I'll just have to do a pcb for this and see.
I subbed the 2N5910 with 2N3906
Subbed 2N5459 with 2N5458
Didn't bother with a tempco...yet.
The waveform offsets in the Odyssey are maintained when using SH Mixer to modulate VCO 2 or the VCF. The signals are AC coupled pre-VCF so level shifting should have little effect on the output sound.
I plan to do pcbs with:
2 VCO circuits with 4011 "ring mod"
portamento
noise
The goal is to make an Odyssey work alike with 3 modules - VCOs, VCF and LFO, and Mod sources (EGs, SH, etc). I'll ditch the 3080 VCA in favor of something better.
Oh yeah.
The FETs are marked "selected" in the Service manual.
Selected for what?
Any guesses?
phil |
|
|
Back to top
|
|
|
Broadwave
Joined: Feb 16, 2007
Posts: 347
Location: Manchester UK
Audio files: 6
|
| Posted: Sat Oct 18, 2008 1:48 am Post subject:
|
|
|
Numbertalk - The FM inputs are linear. I don't think ARP used Log FM on anything they made, maybe apart from the 2500.
Peng - The FET is "selected" for a low pinch off value. Any 2N5457/8/9 seems to work fine.
I didn't bother level shifting the VCO outputs to +/-5 volts, as I wanted a true clone (plus the fact that I wanted a small PCB footprint).
Andy.
_________________
Kronos 2-88, Kronos 61, Studiologic Sledge V2/SL, Broadwave ARP 2600EX, Broadwave 18U ARP based Eurorack Modular, Broadwave Minimoog Clone, GEM S2 Turbo.
Synth DIY Projects
Musical Doodlings |
|
|
Back to top
|
|
|
peng
Joined: Feb 23, 2006
Posts: 111
Location: Chicago
Audio files: 2
|
| Posted: Sat Oct 18, 2008 2:18 am Post subject:
ARP Odyssey Oscillators |
|
|
Here are a couple samples. Two VCOs raw saw output. No VCF.
You'll here the mistracking most in the high range. This is without tempcos.
| Description: |
|
Download (listen) |
| Filename: |
od_raw_saws1.wav |
| Filesize: |
3.06 MB |
| Downloaded: |
1311 Time(s) |
| Description: |
|
Download (listen) |
| Filename: |
od_raw_saws2.wav |
| Filesize: |
1.91 MB |
| Downloaded: |
1163 Time(s) |
|
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Sat Oct 18, 2008 11:05 am Post subject:
|
|
|
| AndyR1960 wrote: | | Numbertalk - The FM inputs are linear. I don't think ARP used Log FM on anything they made, maybe apart from the 2500. |
Can you clarify this please? The FM inputs are going through the exponential converter, so they are exponential FM inputs, right? 
Or did I miss something?  |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Sat Oct 18, 2008 11:18 am Post subject:
|
|
|
Never mind, I think we just have a semantics issue here. Exponential and linear response to the input versus linear and logarithmic controls to the input.
The Odyssey provides an exponential response to a linear input, right? |
|
|
Back to top
|
|
|
peng
Joined: Feb 23, 2006
Posts: 111
Location: Chicago
Audio files: 2
|
| Posted: Sat Oct 18, 2008 1:17 pm Post subject:
|
|
|
| urbanscallywag wrote: | Never mind, I think we just have a semantics issue here. Exponential and linear response to the input versus linear and logarithmic controls to the input.
The Odyssey provides an exponential response to a linear input, right? |
Yes. Expo response. All the FM inputs sum at the base of the PNP of the expo converter. Works with 1V/Oct keyboards, etc.
The faders are different taper (not that doesn't matter much):
two FM inputs - Audio taper
Pulse Width - Linear
PWM - Audio
p. |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Sun Oct 19, 2008 6:48 pm Post subject:
|
|
|
| Looking at how close some of these resistor values are to each other, are any of them crucial enough to require .1% or is 1% okay all around? |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Sun Oct 19, 2008 7:06 pm Post subject:
|
|
|
| Which resistors are you looking at? |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Sun Oct 19, 2008 7:13 pm Post subject:
|
|
|
120K & 121K are actually the only ones I seem to see that could crossover within 1%. Does that matter for these particular ones? Just guessing why not use all of 1 of those values instead of some 120K some 121K if it didn't matter, but I'm not sure.
Thanks. |
|
|
Back to top
|
|
|
peng
Joined: Feb 23, 2006
Posts: 111
Location: Chicago
Audio files: 2
|
| Posted: Sun Oct 19, 2008 9:10 pm Post subject:
ARP Odyssey Oscillators |
|
|
| numbertalk wrote: | 120K & 121K are actually the only ones I seem to see that could crossover within 1%.
Does that matter for these particular ones? Just guessing why not use all of 1 of those values instead of some 120K some 121K if it didn't matter, but I'm not sure.
|
120K for both are fine (or 121K). The reason why they didn't pick the same value is probably because they had 120K 5% and 121K 1%. Back in those days 1% metal film resistors were significantly more expensive than 5% carbon film.
For what it's worth, any clones I work on I approach this way:
Build the exact circuit (as much as possible).
Resistor values that I don't have are made with series resistors.
Once the circuit is going I start swapping out resistors in favor of values I already have. I also start swapping out opamps, transistors, etc.
See what is critical and what is not and WHY.
What improvements can be made and still be true to the original's sound?
Many original circuits are probably stripped down versions of their original self. Meaning Moog, ARP, or whoever would opt for cheaper parts whenever possible as long as it didn't compromise the circuit TOO much.
Feel free to do the same (maybe with better parts). Share your results.
What would be really interesting is if we could figure out what exactly makes the Odyssey VCOs interesting to our ears. For instance, I love the sound of Odyssey VCO sync. Why does it sound different than other saw core VCOs? It would be nice to implement this in other non-clone VCOs if possible. Or even change the circuit to make the sync more pronounced.
Cloning the exact circuit is well and good. But if I don't gain a deeper understanding of it, then it is not a good return on my time. I seek discoveries that will carry over to the next project, and the next...
p. |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Sun Oct 19, 2008 9:28 pm Post subject:
|
|
|
| peng is right, the 120k is specified at 5%, the 121k is 1%. Looks like the 121k resistor is part of the PWM circuit, I guess its a more critical component. |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Mon Oct 20, 2008 6:47 am Post subject:
|
|
|
Thanks for such a detailed response! Make sense.
I've learned a lot since I started building but there's still so much to learn. I should do the same more often, regarding switching out parts once I've finished a successful initial build. Especially with the ICs and transistors, since I always socket all of those. The resistors and caps I could see myself not doing since I'd have to desolder them. Though once I knew a circuit better and knew where a resistor or cap change could/would make a difference I could socket those specific ones. |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Mon Oct 20, 2008 7:15 am Post subject:
|
|
|
| peng wrote: |
The faders are different taper (not that doesn't matter much):
two FM inputs - Audio taper
Pulse Width - Linear
PWM - Audio |
So the PWM pots are 100K Audio (but the manual PW pots are 100K Linear)? |
|
|
Back to top
|
|
|
Broadwave
Joined: Feb 16, 2007
Posts: 347
Location: Manchester UK
Audio files: 6
|
| Posted: Mon Oct 20, 2008 9:11 am Post subject:
|
|
|
I don't usually bother with log pots unless they're absolutely required - lin are fine, you just loose a bit of control at the start of rotation (does that make sense??).
Yes, the FM inputs are logarithmic (muddled brain due to too much Tolexing and glue fumes - apologies)
I've just completed installing everything into a 6U rack, including both the 4023 and 4035 VCF's (with a second rack on it's way)
Andy.
_________________
Kronos 2-88, Kronos 61, Studiologic Sledge V2/SL, Broadwave ARP 2600EX, Broadwave 18U ARP based Eurorack Modular, Broadwave Minimoog Clone, GEM S2 Turbo.
Synth DIY Projects
Musical Doodlings |
|
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007
Posts: 317
Location: sometimes
|
| Posted: Mon Oct 20, 2008 9:42 am Post subject:
|
|
|
Nice, I like the Odyssey MKIII colors.  |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Tue Oct 21, 2008 12:44 pm Post subject:
|
|
|
Sorry if this is a dumb question, but this project is a good example of something I've wondered about. When only 1 pad is given for each jack and pot and there are no extra ground pads on the board, where do you connect the ground lugs of each of these off-board components? Even if chained together on the jacks and pots themselves, where do you then ultimately connect to the common ground of the circuit?
EDIT: I especially wonder about this given the initial postings you made about problems related to using the same ground for both oscillators. If the answer to my question is to just drill a hole/holes somewhere in the ground trace, for this project, should the jacks and pots for each oscillator be connected to each osc's respective ground trace?
Great looking panels, by the way!
| AndyR1960 wrote: | | I don't usually bother with log pots unless they're absolutely required - lin are fine, you just loose a bit of control at the start of rotation (does that make sense??). |
Makes sense. Thanks! |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Thu Oct 23, 2008 6:39 am Post subject:
|
|
|
Is it really *that* dumb of a question? 
I'm just curious what others do to ground their off-board components when there aren't pads provided for them. Do you add pads yourself when you etch your board? If so, or if you just carefully drill small hole(s) somewhere in the ground trace, for this particular circuit, given Andy's initial problem with a ground loop between the 2 VCOs, does it matter what I connect to what part of the ground trace here?
Any help for someone admittedly still learning much appreciated.
| numbertalk wrote: | Sorry if this is a dumb question, but this project is a good example of something I've wondered about. When only 1 pad is given for each jack and pot and there are no extra ground pads on the board, where do you connect the ground lugs of each of these off-board components? Even if chained together on the jacks and pots themselves, where do you then ultimately connect to the common ground of the circuit?
EDIT: I especially wonder about this given the initial postings you made about problems related to using the same ground for both oscillators. If the answer to my question is to just drill a hole/holes somewhere in the ground trace, for this project, should the jacks and pots for each oscillator be connected to each osc's respective ground trace?
Great looking panels, by the way!
| AndyR1960 wrote: | | I don't usually bother with log pots unless they're absolutely required - lin are fine, you just loose a bit of control at the start of rotation (does that make sense??). |
Makes sense. Thanks! |
|
|
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004
Posts: 24452
Location: The Netherlands, Enschede
Audio files: 297
G2 patch files: 320
|
| Posted: Thu Oct 23, 2008 6:50 am Post subject:
|
|
|
| numbertalk wrote: | | Is it really *that* dumb of a question? |
Not dumb at all ... just not so easy to answer in a general way ... when the PCB designer did not bother to think about this it implies some laziness ... and wrong grounding can be a cause of trouble ... when no return paths are provided I'd just solder up something and when it behaves badly I'd try to find a better ground ref. and only then maybe drill a hole ... the hole doesn't have to be in a trace BTW, just beside it might be better as it probably can be a bigger hole then without damaging the trace.
_________________
Jan
also .. could someone please turn down the thermostat a bit.
 |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Thu Oct 23, 2008 8:29 am Post subject:
|
|
|
| Blue Hell wrote: | | numbertalk wrote: | | Is it really *that* dumb of a question? |
Not dumb at all ... just not so easy to answer in a general way ... when the PCB designer did not bother to think about this it implies some laziness ... and wrong grounding can be a cause of trouble ... when no return paths are provided I'd just solder up something and when it behaves badly I'd try to find a better ground ref. and only then maybe drill a hole ... the hole doesn't have to be in a trace BTW, just beside it might be better as it probably can be a bigger hole then without damaging the trace. |
Thanks for the reply. Sorry if my question implied any laziness on the part of the designer - not intended at all.
So when you say solder something up without drilling holes, what are some examples of where you might make these connections then? Thanks for the good advice regarding how to drill without breaking the trace. I usually use a smaller bit but this makes more sense (sometimes the bit is so small it makes it a slight challenge getting the wire through).
Andy, given the ground-related problems you initially had had and Blue Hell's comment on how wrong grounding can be a source of problems, any advice on grounding off-board components on this board? |
|
|
Back to top
|
|
|
Broadwave
Joined: Feb 16, 2007
Posts: 347
Location: Manchester UK
Audio files: 6
|
| Posted: Thu Oct 23, 2008 9:45 am Post subject:
|
|
|
| numbertalk wrote: |
Andy, given the ground-related problems you initially had had and Blue Hell's comment on how wrong grounding can be a source of problems, any advice on grounding off-board components on this board? |
I now tend to take the ground directly from the 0V pad on the PSU connector on all my modules, so far no problems.
I'm also glad that I'm not the original designer of this circuit : )
Yes, ARP did cut a lot of corners on their designs... certainly not enough decoupling for a start.
BTW, just posted the info for the 4035 VCF Module.
Andy.
_________________
Kronos 2-88, Kronos 61, Studiologic Sledge V2/SL, Broadwave ARP 2600EX, Broadwave 18U ARP based Eurorack Modular, Broadwave Minimoog Clone, GEM S2 Turbo.
Synth DIY Projects
Musical Doodlings |
|
|
Back to top
|
|
|
|
Forum index » DIY Hardware and Software
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
