ARP Odyssey Oscillators

patroche · Joined: Feb 26, 2012 Posts: 16 Location: montelimar

ok andy , give me your mail and i give you all my work !
Just a picture of my project
It don't run yet , because i have many vérification , before, on the sequencer.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Sebo · Joined: Apr 27, 2007 Posts: 564 Location: Argentina

WOW, the sequencer looks awesome!!!

isak · Posted: Mon Jun 09, 2014 11:30 pm Post subject:

thats really looks good man!
amazing work!

isak · Posted: Thu Jun 19, 2014 4:00 am Post subject:

hi guys.

i did the Vpp boost suggested by Andy on page 8.

well, works great, i replaced the 100k to 51k and 50k trimmer so i can get it back to 10Vpp, that went out great.

as you see the signal is inverted, i want to invert it back to the original phase, i did a small scheme, someone here can please tell me if i did ok?

cheers,
Isak E.

elmegil · Posted: Thu Jun 19, 2014 6:16 am Post subject:

I've more commonly see 10k or 20k used for the re-inverting op amp, but the layout is correct.

isak · Posted: Thu Jun 19, 2014 6:53 am Post subject:

Thanks

So 10k to 20k instead 100k?
Ok, I'll take that in mind.

Another thing..
When looking at the scope after the Vpp boost I've noticed that the waveform is no longer from 0 to 5V, it became, (+)(-), i guess it should be like that?
I'm wandering if I wasnt connecting the -15V would it still be (+)(-)?

elmegil · Posted: Thu Jun 19, 2014 7:04 am Post subject:

2 x 10ks, or 2 x 20ks. both resistors have to be the same to have unity gain.\

It seems strange that an audio signal would be 0 - 5V; 5Vpp would normally be -2.5 to +2.5. I haven't gone back to look at the prior schematics, is there a DC blocking cap at the end of the prior stage?

In any case, don't try to run your op amp without the -15V rail. And a TL07X is not going to like simply trying to substitute ground there. You'd need a different op amp capable of doing single-sided supply. There are some LM amps (e.g. LM358) that can do that if you really feel you need to.

isak · Posted: Thu Jun 19, 2014 12:16 pm Post subject:

hi elmegil

elmegil · Posted: Thu Jun 19, 2014 12:54 pm Post subject:

Sorry for the misunderstanding. Seems like a good chance of it working as you say....

isak · Posted: Thu Jun 19, 2014 11:32 pm Post subject:

elmegil,

I don't think there is a DC block.
I guess the the -15V and the 300k is for center it around zero.

furio · Posted: Tue Jul 08, 2014 5:19 am Post subject:

isak · Posted: Tue Jul 08, 2014 5:55 am Post subject:

furio · Posted: Tue Jul 08, 2014 9:08 am Post subject:

isak · Posted: Tue Jul 08, 2014 9:17 am Post subject:

furio · Posted: Thu Jul 10, 2014 12:19 am Post subject:

isak · Posted: Thu Jul 10, 2014 2:47 am Post subject:

kanji · Joined: Aug 17, 2012 Posts: 7 Location: Finland

Hello all. I've build several DIY synths and decided to build Odyssey using Dual Oscillator PCB.

First of all I want to thank Andy for his amazing work and dedication in this thread. Even with issues my oscs have, they sound really powerful and your work has helped a LOT to get where my project is now Smile

I've build VCOs, Yusynths Arp filter and Yusynths VCA. ADSR/AR circuits I build from original ARP schematics, same with Mixer section, LFO and S/H.

I was happy with the raw nature the instrument has, but always thought that square waves arent as clean as they seem in odyssey videos in youtube.

I don't have scope, but I checked both oscillators using visualanalyzer and it seems that squarewaves aren't right. Saw sounds nice and looks "ok" in analyzers screen. Squares seem to droop right after their edge has risen to peak voltage. Peak voltage doesn't stay up, instead droops before cycle starts its second half.

I'm running this on 12V and powersupply is checked and tested three times. Same goes with the PCB itself. I also tested changing 274k resistor that affects PWM but that didn't do the trick. My board sets squares dutycycle to 50/50 with original components, but squares shape is the issue.

I've used CA3146 instead of CA3086 and 2N5459's I replaced with 2N3823. All other components are like in original BOM.

If some one has any idea where to start looking solution, help would be awesome.

Thanks in advance Smile

Synth.jpg

Description:

Filesize:

309.71 KB

Viewed:

419 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

waves.jpg

Description:

Saw and Square wave in Visual Analyser

Filesize:

382.49 KB

Viewed:

423 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

furio · Posted: Sun Aug 17, 2014 7:01 am Post subject:

Sebo · Joined: Apr 27, 2007 Posts: 564 Location: Argentina

The squares look like that because you are using a sound card to get the audio into the computer for analizing it. The sound card have a high pass filter to block DC, and that HP filter is deforming the waves.
If you use a real oscilloscope or get hold of a sound card without the high pass filters (there are a few of them, MOTU for example) you would see an almost perfect square.

furio · Posted: Mon Aug 18, 2014 1:12 pm Post subject:

Sebo · Joined: Apr 27, 2007 Posts: 564 Location: Argentina

The converters have nothing to do. The problem are the High Pass Filter that blocks the DC at the input of the sound card.
Just look at the waves with a proper oscilloscope and you will see the diference.
It's funny, this subject is popping out again and again. Software oscilloscopes aren't good for some tasks because the DC blocking filters of the sound cards. You can't measure DC offset (to calibrate a VCA for example), you can't properly see an LFO, and waveforms of VCO will be deformated, specially in lower frequencies...

Broadwave · Posted: Tue Aug 19, 2014 1:22 am Post subject:

I agree with Sebo - I had a similar result using software. It wasn't until I bought a cheap Hitachi Scope, that I was able to monitor waveforms directly from the PCB... they were correct Wink

I'm now using a DSO201 updated with BenF's firmware (recommended). It may not be the most accurate scope out there, but it does the job well and I can take it anywhere.

dso201-diy-kit.jpg

Description:

Filesize:

18.6 KB

Viewed:

25307 Time(s)

kanji · Joined: Aug 17, 2012 Posts: 7 Location: Finland

Thanks guys for the help and info. What drow me to take a look more closely to the waveforms was an interference/sidenoise that could be heard on both oscillators. I thought that its a side product of wrong shaped wave. Like square mixed with little bit of saw wave. This side noise reacted to pitch so I thought it must have something to do with the shape of square wave.

Later I noticed that I forgot to solder in 10uf decoupling caps and when those didn't eliminate the issue, I added 100n caps to powerrails aswell.

Funny though, while I googled info about oddys square waves and known issues, I stumbled on this Very Happy

http://www.gearslutz.com/board/electronic-music-instruments-electronic-music-production/778218-arp-odyssey-wonky-square-waves.html

EDIT : Got it working.. problem was most embarrassing.. Oscillators sounded great when I tested those out from the synth, but when I installed PCB back in, side noise came back. I started to hook modules one by one to powersupply until sidenoise appeared. Problem was that my powersupplys caps were bit too small for the load all these modules take. I boosted in total more than 10000uf more to my PSU and now everything works great.

Thanks again Andy for sharing your work Smile

Here is a small clip how my oddy clone sounds. All sounds (apart white noise) made using these oscillators.

https://soundcloud.com/tiergrinder/clockwork-odyssey

krainov · Joined: Feb 26, 2015 Posts: 20 Location: Nowhereland

Hello!
Could anyone tell me what is VCO 1 Audio/LF Switch on the pcb?