electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Articles  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links  |  Store
Forum with support of Syndicator RSS
 FAQFAQ   CalendarCalendar   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   LinksLinks
 RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in  Chat RoomChat Room 
 Forum index » DIY Hardware and Software » Thomas Henry designs
TH ADSR analysis...questions...
Post new topic   Reply to topic Moderators: Scott Stites
Page 1 of 1 [11 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Thu May 03, 2012 10:36 pm    Post subject: TH ADSR analysis...questions... Reply with quote  Mark this post and the followings unread

After breadboarding the XR VCO (still need to look at my sine behavior, but I'll get there) I figured to really hear what I had, I needed the VCF/VCA and an EG. So I've breadboarded those as well.

Before hooking up the EG, I wanted to test it and be sure I'm getting the expected output. Which, of course, I'm not. Smile

Test setup: hooked up the MTS-100 variant I built a while back, with a small keyboard for generating keypresses. Both Trigger & Gate outputs are hooked into the EG and confirmed.

Results: a "high" output (this was a couple nights ago now, so I think it was steady at 5V), dropping low as soon as I hit the key, staying low while I held the key, then charging back up (classic RC curve) when I released it. Rate of charging back up, and floor for the drop both dependent on the setting of the release pot.

Obviously that's nearly the opposite of what's expected. I'm not seeing the pulse come through the trigger circuitry either.

At the point I was testing, I hadn't really analyzed the circuit to understand correct functioning, but I have now, I think. I wanted to spit out what I think I understand and see if I've goofed anywhere in my understanding--I'm still rusty with my EE skills and don't trust myself entirely. At this moment I think I really need to double check that my power / ground rails are actually what I think they are, because that's the only way I can see the release behaving the way it is. The wiring looks right, but there are enough components hooked in that a bad one could cause weird things to happen.

So, to the analysis (this assumes you have a copy of the schematic, which I'll not provide here...it's part of the Analog Synth for the 21st Century book available through Magic Smoke): Key piece is C7, which is the timing capacitor for the 555 timer. The set up is sort of "Pseudo" monostable; I say pseudo because the discharge pin of the 555 isn't connected, instead we rely on the ADSR pots in combination with the 4016 switch to charge and discharge the cap. And we're getting our waveform off that cap rather than off the normal output of the 555.

Normal steady state (no gate or trigger signals) is for switches 3/4/5 and 6/8/9 to be open and 1/2/13 and 10/11/12 to be closed. This should be holding the timing cap to ground and giving a zero output through op amp IC3, which is configured as a voltage follower with a divider on the output.

Q3 is not conducting, which both holds the 10/11/12 switch closed, and causes Q2 to be conducting to ground, which holds pin 4 of the 555 in reset.

A Gate signal coming in causes Q3 to conduct, taking pin 12 of the 4016 to ground, opening the switch, as well as cutting off Q2, which raises the voltage on the reset pin and allows the 555 to go.

The trigger coming in goes through Q1 which inverts it and generates a trigger pulse to flip the 555 output on. This in turn closes switches 3/4/5 and 6/8/9. 3/4/5 charges the timing cap through the Attack pot. 6/8/9 pulls 13 to ground and so opens up 1/2/13.

The timing cap charges, reaches the 10V threshold (2/3 of Vcc) and resets the output back to ground. This opens 3/4/5 (stopping the charging of the timing cap) and 6/8/9, which in turn closes 1/2/13.

The cap discharges through the Decay pot to the level set on the Sustain pot, and stays there until the Gate is released. At that point, switch 10/11/12 closes and the cap discharges through the Release pot back to ground.

So if I've got that right, I need to check my power/ground rails and find why the output is inverted, and then I need to confirm that I'm getting the trigger pulse correctly through Q1 (in testing the other night I didn't see any pulse on the other side; however, if my power is scrambled somewhere it could simply be that I never get Q1 to conduct and it may work once I have it straight).

Does that all sound right?

Thanks....
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Fri May 04, 2012 3:08 pm    Post subject: Reply with quote  Mark this post and the followings unread

The good news: I know what was wired wrong to invert my output; somehow I connected pin 11 & pin 12 of the 4016.

The bad news: having corrected this, I am still not getting any sign of a pulse through C1. I've removed the transistor and the diode, so all I'm dealing with is signal -> cap -> 10K to ground and 100K onward. Just measuring at the other side of the cap, I get zero volts consistently.

I've replaced the cap twice; first time I replaced a ceramic with another ceramic, and then I replaced it with the only other 470p cap I have, a C0G from a batch I just got in last week.

So now I'm completely baffled as to why I can't see the pulse through the cap. I expect to see a short spike at least....and I'm getting nothing.

Anyone have any clues?
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Fri May 04, 2012 8:11 pm    Post subject: Reply with quote  Mark this post and the followings unread

I've spent some time playing with SPICE and to get any visible pulse at the collector of Q1 I have to increase C1 pretty significantly.

I see Thomas' own comments in an old thread (http://electro-music.com/forum/topic-21864.html):

Quote:
How did I pick the value? In a nutshell: rule of thumb. I didn't calculate anything here.

A keyboard trigger is on the order of 5mS. I wanted something much narrower than that to fire the 555. C1 and R5 set a much shorter time constant, guaranteeing the 555 fires pronto.

The circuit should work well with the values shown with just about any keyboard. I suppose it's possible your keyboard may generate a pulse width that made the larger value necessary. And as you say, it's also possible the original cap was cruddy.

If it were me, and given how cheap caps are, I'd try the original value, or maybe a 0.001. But the 0.01 sounds kind of big to me.


I'm not sure I understand why it needs something narrower than that, since the trigger is supposed to fire on the negative going edge of the pulse. But even to get something ~1ms I have to use a 100nF cap there (in the simulation). If I go as low as 1nF as suggested and crank down the transient analysis steps, I can *barely* make out a negative pulse.

This is on a breadboard, is it possible I'm just not making sufficient contact to be able to use the 470pF?
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Fri May 04, 2012 9:11 pm    Post subject: Reply with quote  Mark this post and the followings unread

Ok, back to the real world, using the right scale on my scope I can see a pulse with the 470pF C0G. It's on the order of .1us, but it's there. I do see where the trigger can override the cycle time of the 555, but I'm still unsure about a longer pulse being a problem, when even my fastest key hit is usually on the order of 100ms.
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Sat May 05, 2012 9:09 pm    Post subject: Reply with quote  Mark this post and the followings unread

Rebuild again...

This time I've rebuilt it on a larger breadboard, and given what I understand of the theory above I've built it in pieces and tested as I went along. I did confirm that my original 4016 was toasted by putting it in its own circuit and testing the response to switch voltage changes.

First: with just the input transistors hooked up, the trigger does generate the very fast pulse on the collector of Q1, and the value at the collector of Q2 basically follows the gate, except from 0 to 15V (gate is 0 to 5V).

Second, with the 555 and 4016 in place (can't really charge the timing cap without both), now the collector of Q2 is normally at 7.5 volts, not 15. Poking around, if I disconnect pin 12 of the 4016 (the release path) then I go back to the previous behavior. This is of concern because it doesn't drive pin 4 (reset) on the 555 correctly..

Even with pin 12 removed, however, I'm still seeing no pulse on the output of the 555. I suspect perhaps that my trigger pulse isn't wide enough. But without that output going high, I never charge the timing cap, and it all just sits there.

I'm about 4 hours more work away from scrapping it and trying a different EG (say MFOS, or the one here: http://www.experimentalistsanonymous.com/ve3wwg/doku.php?id=synth_dual_adsr)


bad-gate-pulse.jpg
 Description:
 Filesize:  99.14 KB
 Viewed:  95 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

bad-gate-pulse.jpg



good-gate-pulse.jpg
 Description:
 Filesize:  104.5 KB
 Viewed:  99 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

good-gate-pulse.jpg


Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Sat May 05, 2012 10:36 pm    Post subject: Reply with quote  Mark this post and the followings unread

Amazing how much better a circuit works when you have power actually hooked to power instead of ground. Embarassed And aside from that goof, I really need to find a better way to get power onto my breadboards, the mta-100 connectors aren't staying in very well either.

I've only been observing the timing cap right now, I haven't yet hooked up the TL071, but I am definitely getting classic ADSR waveforms now.

The range of settings on the pots seems odd to me; if I can trust my scope, my release is a cliff when set to minimum (as expected) but a very short turn of the pot gives me a 1 second release, which seems long. And while I can see the point of longer releases, I would expect to have more range at this end. I may try swapping in smaller pots or using an audio taper rather than linear. I'm going to listen to it first though.
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Sun May 06, 2012 6:23 am    Post subject: Reply with quote  Mark this post and the followings unread

Aaaand with the op amp in place, measuring the actual output, looks very nice. I haven't tried twiddling the knobs any further, but just looking at it, the drop on release is 1/4 second now, so that appears to have likely resolved my concerns about timing. We'll see what I hear later....


adsr.jpg
 Description:
 Filesize:  130.23 KB
 Viewed:  102 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

adsr.jpg


Back to top
View user's profile Send private message
Tim Servo



Joined: Jul 16, 2006
Posts: 924
Location: Silicon Valley
Audio files: 11

PostPosted: Mon May 07, 2012 4:14 pm    Post subject: TH ADSR analysis...questions... Reply with quote  Mark this post and the followings unread

Wow, you've had quite a weekend! Thanks for sharing all your experiences with this. The cure for the last problem (envelope times not proportional to pot settings) was a very good one for people to pay attention to. My guess is that without the output buffer amp in place, the timing cap was loaded down too much. Anyway, thanks for sharing!

Tim (another lost weekend) Servo
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Mon May 07, 2012 6:14 pm    Post subject: Reply with quote  Mark this post and the followings unread

There was an additional wrinkle Smile

Yesterday I went to try and make it a dual using a 556, a TL072, and an additional 4016... I tested it first, and it looked good, but after the first change (swap in the TL072, that seemed the lowest impact) I started getting a profile that wasn't decaying...it would start at about 2V and after release it'd be over 3. You'd see the basic outline of the waveform but it wasn't discharging the cap correctly for some reason. Even after going back to the original just-tested parts I got the same thing.

Turns out I must have zapped or something, Q2 was no longer working. I would have rather done without additional problems, but it was a good troubleshooting exercise to tear it back down to the basics. I swapped in a new transistor and all was right with the world, so I'll be trying to do the dual stuff later this week....
Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Wed May 09, 2012 8:21 pm    Post subject: Reply with quote  Mark this post and the followings unread

I don't think this yet qualifies fully as a build picture, because it's just a panel with parts mounted on it, but I'm curious about what folks think of it as an early effort even so.

So: on left, modified MTS-100 using Arduino and omitting a few things (see previous posts of mine about that in more detail). This is actually a working module. On right, mocked up panel for dual ADSR from ASft21C. The idea to set it up as ganged came from warrenmaximus' website (though I am using stock TH circuitry). It isn't actually wired that way on the breadboard as yet though...

Some thoughts for next time (don't really want to scrap this one unless something really isn't going to work out)....

Slightly off center at the bottom. Need to figure out a good way to reliably get the label centered. Still isn't clear coated...

I had wanted to do/been doing Photoshop designs using 1" spacing, but clearly if I want to put two columns on a 1U panel, I'm going to have to use 3/4". Again, I don't really want to redo this one unless it's terrible when it's time to assemble for real, but the edges of the N/C jacks at the bottom are definitely at the edge, and I could see that being an issue down the line if I do it more.

BTW I like using Photoshop because I'm comfortable with the layer functions, and that makes it easy to print just the crosshairs or just the artwork, while I can use bigger graphics to visualize the end product on the screen. I think if there's interest I could probably work up a tutorial about how to do the same thing with The Gimp (Freeware similar to photoshop).

Biggest potential problem that I see...where to mount the soldered-up board Smile. I think I have room above the gates and between the knobs (and yes, btw, I know I'm two knobs short Wink ) to put some mounting hardware, or possibly just above or below "Envelope". But I think it will be highly dependent on what I use to mount it and what kind of board I use (I'm still oscillating between trying to learn enough Eagle to do a "real" PCB or just do another protoboard like I've done so far).

Thoughts? Advice? Denunciation for being a naive fool? Smile


dual-adsr.jpg
 Description:
 Filesize:  68.85 KB
 Viewed:  2388 Time(s)

dual-adsr.jpg


Back to top
View user's profile Send private message
elmegil



Joined: Mar 20, 2012
Posts: 1535
Location: Chicago
Audio files: 14

PostPosted: Tue May 22, 2012 9:31 pm    Post subject: Reply with quote  Mark this post and the followings unread

Making headway on this, have it soldered/wired up. Ended up going with proto boards again, difficulty printing with toner and spousal aversion to nasty chemicals being the key factors Smile. Waiting on tests. I did make one stupid mistake but found it in routine go-over; I crossed the outputs (EG A wired to pot for EG B and vice versa). As problems go, pretty minor. We'll see if that's all once I start testing.

Not really impressed with my own panel wiring job, but it's better than the MIDI->CV so I'm making progress.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: Scott Stites
Page 1 of 1 [11 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software » Thomas Henry designs
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
e-m mkii

Please support our site. If you click through and buy from
our affiliate partners, we earn a small commission.


Forum with support of Syndicator RSS
Powered by phpBB © 2001, 2005 phpBB Group
Copyright © 2003 through 2009 by electro-music.com - Conditions Of Use