Sunday 10 October 2021

Synthesizer Build part-44: AD/AR ENV. GENERATOR.

This is a simple AD/AR envelope generator by Ole Stavnshoej.  With the little alterations I made it's a very useful AD/AR for use with filters.

CHANGES I MADE:
After re-doing the layout for the 7555 AD/AR I decided to try an other design and I found this one online. It's a simple design and easy enough to build. I intend this particularly for use on the CV inputs of filters and after I built it I made a few changes to make it better suited for this role.
First of all I made the electrolytic capacitor smaller so that the circuit would react faster and be more accurate. Later I added a SPDT switch (ON-Center OFF-ON) with some other electrolytic caps soldered to the switch so you can choose how fast you want the envelope generator to be. I also changed the 1M resistor coming of the 200K trimpotmeter to a 820K to make that trimmer more effective. Later I included the attenuverter potmeter and I changed the trigger input capacitor from 10nF to 3,3nF to make the circuit react better to fast playing. That change made all the difference. It now works much better and is very responsive. This is now my go to AD/AR for use with filters.

The trimmer sets the zero volt line or offset voltage but the strange thing with this circuit (if you use higher value caps for C4) is that if you turn the Release potmeter you also change the offset voltage. If you turn it all the way open the offset can be as high as +5V. So on long release times it never returns to zero. That's why this design is really only suitable for use to excite filters where you have very short envelope times. If you connected it to a VCA it would stay open all the time with long release times. The attenuverter mod however could help a bit in setting this straight but I haven't tested that.
However if you use a 1,5µF cap for C4 it doesn't give this problem. It starts with higher value caps that's why I put in the time range switch with different value caps. I talked about it with Ole and we both think it's due to residual voltage in the capacitor C4.

Instead of the TL084 you can also use a TL074 or an LM324 or really any quad opamp with the same pinout. Same goes for the TL082.

LAYOUTS:
THE LAYOUT FOR THE AD/AR WITH THE ATTENUVERTER MOD IS BELOW THESE FIRST TWO LAYOUTS!
Here is the layout I made. It's verified as always. The inverted output is grayed out because I didn't use it and with the attenuverter mod you get both at the same time. But I leave it up to you whether you want to include it or not. I sent the normal output to a second opamp buffer, which I originally put on there for testing, and connected that to a second output. Always useful to have more than one output I think.
Oh and this circuit runs equally well on +/-12V as on +/-15V but the output envelope can reach almost the positive voltage rail so if you need lower amplitudes install the potmeter in the attenuverter mod. (Schematic + layout is further down the article)
Wiring Diagram:

Print only. (The green wirebridges indicate connections to ground):


Here's the schematic drawing. You will note that some values on the schematic are changed on the layout. I did this after testing so follow the values in the layout. Like I mentioned earlier I changed the 10nF trigger input capacitor to a 3,3nF one because the circuit was too slow if you play fast on the keyboard. 


Here is a Falstad simulation of the above schematic. I had to change the timing cap to a bi-polar one of lower value and the 1M offset zero resistor was changed to a 500K. Try changing values of components (right click on them) and see what happens. The circuit is triggered by a 1% duty cycle squarewave of 7V. If you want a Gate input then change the duty cycle to 50%


Bill of materials for the version without the attenuverter:



ATTENUVERTER MOD:
Here's the version with the attenuverter modification, which I of course only discovered after completing this build. But it's easy enough to implement so I changed my module later and included this option. It only requires a potmeter on pins 8 and 12 of the TL084. It did mean that I had to find a spot on the panel to put the potmeter. I ended up putting it between the output sockets. A bit awkward but at least it works very well.


Here's the layout, adapted to include the attenuverter potmeter. The only other change is that resistor R19, the 47K from pin 12 to ground, is removed.


Print only:


The way the attenuverter is wired up in the layout, you will get the uninverted output if you turn the potmeter fully clockwise and inverted output fully counterclockwise. (On my panel I had it the other way around. ^^) And wow does this make a difference! If I send the inverted signal from this AD/AR into the CV IN of the Steiner Parker filter you get almost a flute like, very clear sound if the filter is set a certain way. I loved it.

CUTS AND WIREBRIDGES:
For extra clarity, here are the layouts showing just the cuts and the wirebridges, which you should do first before soldering on the components. These layouts are the same for the version with or without the attenuverter.

Cuts and wirebridges, component side:

Cuts only, COPPER SIDE!
 
Bill of materials for the version with the attenuverter. 



Here are some pictures of the print and panel. I didn't have the attenuverter connected yet in these pictures:



A look at the panel. You can see the Time Range switch I added to the left of the Trigger/ Gate switch.
I used a dual pole switch with a middle off position so I could have three ranges, short, medium and long. The Short setting (in the middle) uses just the 1,5µF cap on the print. The Medium setting adds to that a 2,2µF cap and the Long setting adds a 4,7µF cap to the one on the print. You can see the attenuverter at the bottom crammed in between the in and output sockets.



Here are some images from the oscilloscope:

Fast squarewave on input and then turning release up. 

Again opening Release on a fast pulse train with a little bit of attack. Note how the pulses go a bit below the zero volt line here. (Not a big deal for use with filters):

Fast Attack, tiny bit of Release:

Turning the attenuverter from negative to positive output. I'm running this module on +/-12V and you can see the maximum output is +/-11 Volt! Just beware that the output voltage can be quite high, especially if you decide to run this on +/-15V..

Just turning the Release knob without any input using a 4,7µF cap for C4. Note how the voltage on the output changes. This should be a flat zero Volt line. If you use the Time Range switch and set it to longer times and you turn up Release, it doesn't come down to zero volt anymore. Again, it doesn't do this with the 1,5µF cap only with higher values.

And finally a little demo video of how the attenuverter influences the filters while turning it from positive all the way to negative envelope output. Note the tuner on top of the case. It's connected to one of my Thomas Henry 555 VCO's and it's rock solid in tune! I really love those 555 VCO's. If I could I'd marry one :p


Doesn't this synthesizer sound great?? It's a real bass monster if you want it to be. I bet it can rival a MiniMoog in that respect.

Okay, that's an other article done. If you have any questions or remarks, please put them in the comments below or put your question to our awesome little community of DIY synth nerds on the EDDYBERGMAN Discussions Facebook Group.


4 comments:

  1. I'm not sure if it's a full explanation but part of what happens in an envelope generator like this is once the cap voltage drops to ~700 mV the diode conductance drops very low and restricts the discharge current so it takes tens of seconds for the capacitor voltage to fall the rest of the way from 700 mV to (near) 0 V. There should be no fixed "offset" (if you don't put one there) but it can look like it since the end of the release takes so long. There's a Kassutronics ADSR that fixes that problem by putting the diode in the feedback of an op amp which guarantees the diode will be in conduction all the way down to 0 V on the capacitor. Here's the link: https://kassu2000.blogspot.com/2015/05/precision-adsr.html . I built one with mods for retriggering and looping, it works well. A similar idea could be used in a simpler AR/AD EG like this, though since you've used four op amps here you'd need to add a TL071. I don't know if it'd fix all the odd behavior but it'd probably help.

    ReplyDelete
    Replies
    1. Thanks for that explanation Rich! That explains it a bit better to someone like me who just builds it but doesn't know all the ins and outs of the working of a circuit like this. (That diode idea sounds good.) I actually had this one go into negative voltages when testing. I think with a simple design like this you're bound to get some weird behaviour. I thought I'd write about it anyway so people know what to expect from this circuit. Thanks for the extra info! It's very much appreciated!

      Delete
  2. Keep up the good work!

    Looking forward to your take on a BBD delay, analogue quantizer or analogue phaser using lamps or LEDS.

    Take your pick or build them all!

    ReplyDelete

Note: comments are moderated and do not appear straightaway. Your first comment is not allowed to contain any links.