Announcing: Synthesizers.com Manufacturer Sub-Forum

Please check out the new sub-forum by following the link above. It is nice to have another manufacturer onboard and opening another direct communication line to potential and existing customers

Jupiter-4 style oscillators.

Anything modular synth related that is not format specific.
Post Reply
MaxDM
Common Wiggler
Posts: 72
Joined: Tue Feb 15, 2022 12:13 am

Jupiter-4 style oscillators.

Post by MaxDM »

I was looking at the Jupiter-4 service manual and noticed that the oscillators are quite unique, in the sense that they are driven by a 555-generated high-frequency pulse, which is then used to build a sawtooth out of discrete steps.

I'm sure many of you here know this, but this is the first time I hear about it.

I used to own one many years ago, but gave it away because it was not midi, and the sound was a bit too '70's' for what I was doing at the time.

What were/are the advantages to such a design, do you think

Do the oscillators lend themselves to a stand-alone monophonic module, say for mono basses etc?

Has anyone here tried any re-issues of said oscillators?
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

Absolutely valid as a standalone module. As are the EGs.

Notably, the PWM is also 'stepped' in non-obvious and not-normal ways since it's acting on - derived from the stepped ramp-saw.

Primary advantage -at the time- was that it was a way to accomplish the goal of a programmable poly. That *could* be done. It's hard for those who weren't alive at any given time to understand just how limited things were in prior times. We can't escape our own knowledge of the present to see and feel what it was like when the life as we know it didn't exist yet.

Back then there were quite a few different paths being tried for presets and polyphony. Then the P5 came out and the die was cast.

AMSynth was working on a single voice Jup4 monosynth a few years ago. Appears it was sidelined when he did some work for Behringer.
There have been a few other one-offs using Jup4 circuits.
User avatar
cornutt
Fig juggler
Posts: 1683
Joined: Sun Jan 25, 2009 10:06 pm
Location: Rocket City USA

Re: Jupiter-4 style oscillators.

Post by cornutt »

Interesting. I did not know this. I always assumed that the J4 used fairly conventional saw-core VCOs. I'm going to have to read up on this. It isn't a Walsh-function generator, is it? That description sort of sounds like one.
Sequence 15 -- sequence15.blogspot.com
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

Not Walsh functions.

Here's the VCO core.
Where it says VCO out at the left is a pulse train that is V-oct already. This part of the circuit makes a 3 bit stepped saw-ramp from it.


!! Can't upload the drawing. Getting this message still: Sorry, the board attachment quota has been reached.

The image i wanted to upload This is cropped from page 6 -of 44- in the JP4 Service notes PDF. On page 22 of that PDF you can see -in the lower mid-left of the module board schematic- how IC9 is used as a comparator to create PWM. Since the 'saw' is stepped the comparator has 'flat spots' in its response. And so does the resulting PWM output.
Attachments
JP4-VCO-core.JPG
Last edited by KSS on Mon Mar 07, 2022 3:28 pm, edited 1 time in total.
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
MaxDM
Common Wiggler
Posts: 72
Joined: Tue Feb 15, 2022 12:13 am

Re: Jupiter-4 style oscillators.

Post by MaxDM »

Well, here's a couple of pages from the manual.

Unfortunately, the translation into English language is not good, so it's hard to understand. Some of the terms are directly translated.

Just reading it quickly I am under the impression that the main pulse VCO has some kind of feedback mechanism which aids to keep it in tune?? and that would explain the unusual design, perhaps. They needed something very stable, for a poly synth. ??

I can't remember how the JP-4 was, tuning-wise. It was so many years ago.

EDIT I tried to upload some images as well, and got the 'board attachment quota message' as well
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

If you add the PDF page numbers to your post in an edit then we can still converse on the topic.

I believe you're talking about the paragraphs following 1. on page 6 of the PDf. There is also a block diagram on that page just above the saw creation portion I cropped but could not load. Going to upper left on page 22 you will see the actual Module Board circuits used for the V-oct pulse creation. Where a typical dual transistor expo current generator is fed into a VCO created from three gates-inverters of IC4-4069. The 555 IC4 is the one shot in the feedback loop with a couple more of the inverters from IC4. This feedback is added to the original expo gen current and the result is a stable V-oct narrow pulse train.

Which then feeds the pulse-to-saw circuit using IC6 4024 and IC7 4053 and 'summing' resistors to make the stepped saw. It's quite an elegant design and not too difficult to re-create today. The now expensive temp compensated 726 used for expo gen can be replaced with a tempco and dual tranny or the heated 3046 trick using an AS3046.

The entire VCO core and 4016 WF switching circuit is on page 22 -of 44- in the PDF. Upward of co-ordinate R and left of 18.
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
User avatar
cornutt
Fig juggler
Posts: 1683
Joined: Sun Jan 25, 2009 10:06 pm
Location: Rocket City USA

Re: Jupiter-4 style oscillators.

Post by cornutt »

That's an interesting circuit. It could almost be mistaken for a DCO design, although clearly it isn't. I don't think I've seen a "square core" VCO design before.
Sequence 15 -- sequence15.blogspot.com
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

What may be even more interesting is that the PCB layout is on page 23 and seems made to order for a DIY VCO. It's all on the left side of the PCB, and includes the power connections and power filtering caps! And there's even an easily split vertical 'line' that doesn't have a bunch of traces getting in the way. Single sided PCB for easy DIY.

The pads for the 726 are in DIP form, which means making a little translator board -to use something instead of the 726- is easy from stripboard, pad-per-hole or plain perf.

It's almost like Roland were planning to make individual VCO product of some type. Almost. They were actually only foollowing decent design rules for the best isolation of the VCO from other parts of the synth.

The EGs of this synth are also special. Working again using a stream of pulses, this time integrated to make the envelopes. Another fine DIY project.

Here's the Module PCB showing the VCO part. Easy enough to pull up the PCB layout in an EDA or even a graphics-only app and trace it for press-n-peel.
Attachments
JP4-ModuleBd-1024-VCO=RofPkLine.jpg
JP4-ModuleBd-VCO.JPG
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
nuromantix
Super Deluxe Wiggler
Posts: 1010
Joined: Sun Jul 29, 2012 7:34 am

Re: Jupiter-4 style oscillators.

Post by nuromantix »

Is ProMars the same design?
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

Yes. The VCO description for the ProMars is on page 6 of its MRS-2_SERVICE_NOTES.PDF

And the VCO circuit schematic and PCB layout is on page 16 of the PDF. PCB layout is different but still easily DIY'd
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
User avatar
MindMachine
weekend warrior
Posts: 7957
Joined: Sun Aug 02, 2009 1:45 am
Location: Santa Susana Field Lab
Contact:

Re: Jupiter-4 style oscillators.

Post by MindMachine »

KSS wrote: Sun Mar 06, 2022 4:28 pm Absolutely valid as a standalone module. As are the EGs.

Notably, the PWM is also 'stepped' in non-obvious and not-normal ways since it's acting on - derived from the stepped ramp-saw.

Primary advantage -at the time- was that it was a way to accomplish the goal of a programmable poly. That *could* be done. It's hard for those who weren't alive at any given time to understand just how limited things were in prior times. We can't escape our own knowledge of the present to see and feel what it was like when the life as we know it didn't exist yet.

Back then there were quite a few different paths being tried for presets and polyphony. Then the P5 came out and the die was cast.

AMSynth was working on a single voice Jup4 monosynth a few years ago. Appears it was sidelined when he did some work for Behringer.
There have been a few other one-offs using Jup4 circuits.
https://modwiggler.com/forum/viewtopic. ... s#p3692872
User avatar
Prunesquallor
Ultra Wiggler
Posts: 858
Joined: Wed Nov 11, 2009 3:08 am
Location: Valencia, Spain

Re: Jupiter-4 style oscillators.

Post by Prunesquallor »

MindMachine wrote: Sat Apr 02, 2022 11:27 pm
KSS wrote: Sun Mar 06, 2022 4:28 pm Absolutely valid as a standalone module. As are the EGs.

Notably, the PWM is also 'stepped' in non-obvious and not-normal ways since it's acting on - derived from the stepped ramp-saw.

Primary advantage -at the time- was that it was a way to accomplish the goal of a programmable poly. That *could* be done. It's hard for those who weren't alive at any given time to understand just how limited things were in prior times. We can't escape our own knowledge of the present to see and feel what it was like when the life as we know it didn't exist yet.

Back then there were quite a few different paths being tried for presets and polyphony. Then the P5 came out and the die was cast.

AMSynth was working on a single voice Jup4 monosynth a few years ago. Appears it was sidelined when he did some work for Behringer.
There have been a few other one-offs using Jup4 circuits.
https://modwiggler.com/forum/viewtopic. ... s#p3692872
Unfortunately, AMSynth erred on the side of being too authentic and there are no separate waveform outs. A huge opportunity missed.
If at first you don't succeed read the instruction manual.
User avatar
cornutt
Fig juggler
Posts: 1683
Joined: Sun Jan 25, 2009 10:06 pm
Location: Rocket City USA

Re: Jupiter-4 style oscillators.

Post by cornutt »

Now that I've had a chance to look more closely at this VCO circuit, it actually bears a fairly strong resemblance to the Juno 6/60/106 DCO, with the 555 providing the timing in place of the Juno microprocessor. What is the 4052 mux actually doing? It looks to me like all it is doing is providing the octave select function, since its A and B inputs are coming directly from the panel octave switch. That being the case, what's actually building the stepped saw? What am I missing?
Sequence 15 -- sequence15.blogspot.com
User avatar
KSS
Super Deluxe Wiggler
Posts: 10035
Joined: Mon Jan 25, 2016 7:28 am

Re: Jupiter-4 style oscillators.

Post by KSS »

cornutt wrote: Sun Apr 10, 2022 10:48 am Now that I've had a chance to look more closely at this VCO circuit, it actually bears a fairly strong resemblance to the Juno 6/60/106 DCO, with the 555 providing the timing in place of the Juno microprocessor. What is the 4052 mux actually doing? It looks to me like all it is doing is providing the octave select function, since its A and B inputs are coming directly from the panel octave switch. That being the case, what's actually building the stepped saw? What am I missing?
Upper left corner of pg22 of 44 in the JP-4 service notes PDF
The Y side of 4052 IC7 is setting the saw's integrating current -using weighted resistors- at OPA 5a
IC8 pins 3,4,5 are the saw WF switch for the VCO output WF summing at OPA 9b.
The X side of IC7 is selecting the octaves from the IC6 divider which is being fed directly by the CCO of IC4d,e,f. Instead of going through the 555.

The integrating current needs to be increased for higher frequencies due to shortened time to reach the desired peak. The quintessential problem of all DCOs. That of decreasing WF amplitude as freq goes up.

The 'step' in the saw comes from the discrete-quantized bits of charge provided by master freq CLK-VCO IC4d,e,f then inverted by IC4a and quantized by the 555. OPA 5a is receiving these 'packets' of charge to integrate on C8, with Q17 performing the cap reset using the X out of the 4052.
See page 6 for text description and note the two figures there.

Edit: What might be confusing is that the text description refers to the servo-ing -balancing action of comparator 5b and it may not be clear how this is happening. Look again at page 6 VCO feedback loop and at the same time the actual circuit on pg 22. Looking at only one at a time it's more confusing since the separate parts of IC4 are not called out on pag 6. IC4d,e,f is the 'VCO' seen in the pg6 feedback drawing and feeds the X side of the 4052 directly to provide the suqare WF output.

What adds to the confusion is that 5b as a comparator seems not to makes sense with its non-inverting input at GND, and the two things described as being compared on pg6 both showing up on its inverting pin. But if you look at it as if IC5b is 'taking' the current from IC2, pin 9 and is then causing the CCO at IC4d,e,f to adjust its rate of triggering the 555's 'drips' of current to 'refill' the emptied point at IC2, pin9 -via IC4a,b, diode d1and UNlabeled R118** in a feedback loop then you can see that the comparision to 0v-GND makes sense.

**R118 in the pg6 text is *not* seen on the PCB. Nor is it labeled R118 on pg22 schematic There's a jumper you can find just above Cf label at pg23 page coordinate D23.

The servo'd pulse output of the 555 is used to charge the saw WF cap. <--Which is one of *three* timing caps in this JP-4 VCO core. One for the 555, one for the saw integration and one for the IC4 CCO AKA Current Controlled Osc.
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
User avatar
emuarc
Common Wiggler
Posts: 247
Joined: Sat Oct 02, 2010 4:23 pm
Location: Horsham, West Sussex
Contact:

Re: Jupiter-4 style oscillators.

Post by emuarc »

The AMSynths Jupiter 4 modules are alive and well, the AM8400 VCO is in production along with the AM8104 Filters using the new AS662 chips. Oh and a MKS-40 is in development…
Rob Keeble
Owner & Designer
AMSynths
od120
Learning to Wiggle
Posts: 47
Joined: Thu Jul 01, 2021 9:42 am

Re: Jupiter-4 style oscillators.

Post by od120 »

MKS 40???
A rackmount Jupiter 4 clone?
User avatar
emuarc
Common Wiggler
Posts: 247
Joined: Sat Oct 02, 2010 4:23 pm
Location: Horsham, West Sussex
Contact:

Re: Jupiter-4 style oscillators.

Post by emuarc »

Yep, 2U rack mount Jupiter 4 with MIDI and 64 patch presets.
Here is more info, along with pages about the Jupiter 4 VCO, VCF and ADSR (with timings).

MKS-40
Rob Keeble
Owner & Designer
AMSynths
oldgearguy
Ultra Wiggler
Posts: 990
Joined: Thu May 26, 2016 6:42 am

Re: Jupiter-4 style oscillators.

Post by oldgearguy »

So please let us know when you're taking pre-orders/payments. This is such a worthwhile undertaking and I'd be happy to support it.
User avatar
Sysagent
Wiggling with Experience
Posts: 277
Joined: Thu Jan 07, 2021 2:35 pm
Location: The Ends Of The Earth

Re: Jupiter-4 style oscillators.

Post by Sysagent »

Sign me up as well!
"The Most Annoying Man In The World..."
automatique
Common Wiggler
Posts: 53
Joined: Wed Feb 28, 2018 8:13 am

Re: Jupiter-4 style oscillators.

Post by automatique »

I'm curious how far it has gone yet? Checking your blog from time to time...
Post Reply

Return to “Modular Synth General Discussion”