Synthesizer Build part-68: VC DELAY by BMC.

This is BMC 83 the voltage controlled delay using the PT2399 chip. This is a eurorack friendly project.

Dispite the fact I built over 68 projects I never build a digital delay or reverb, except for project 11 but that was a ready made effects unit. This project takes care of that. It can deliver delays of up to 4 seconds, but not in high quality. The PT2399 wasn't made for such long delay times but shorter times sound really good and with the long times you get some cool distortion, sort of a bitcrush effect.

This was quite an easy project to build. You can find the original article on the Barton Musical Circuits website. There are audio demonstrations on that website so you can hear what the delay sounds like. I also made a demo video myself which you can find at the bottom of this article.

This circuit will work fine on both a dual 12V or a dual 15V powersupply.

The finished delay module

Here's the schematic I used to make my layout from. I changed the opamp numbering to match that of the layout.

I didn't use the 10 Ω resistors in the powerrails as shown on the schematic. But if you have problems with hum you can include them. You could put a 10 Ω resistor from K-3 to I-3 and then lead the red wirebridges from there and the purple wirebridge could be replaced with a 10 Ω resistor for the negative voltage rails.

The diode and 1M resistor in combination with the 100nF cap and the top transistor with collector to pin 6 of the PT2399 make up an anti latch-up circuit that presents a high impedance to pin 6 in the first 400mSec after you switch on which gives the internal oscillator time to warm up and prevents the chip from latching up and crashing which can happen if the resistance between pin 6 and ground is less than 2K at start-up. After start-up this resistance can be much lower but not a straight short to ground. In this module the resistance is then controlled by the second transistor which is opened up by the time control potmeter or external CV input. This resistance controls the delay time.

So there are voltage controls with level potmeters for the delay time and the return amount and the module has an audio output that outputs just the delayed signal and a mixed audio output which mixes the original signal in with the delayed signal controlled by the 'Mix' potmeter. There's also a tone control potmeter which also influences the return time I noticed (see demo video below)

The delay time range goes from 60 milliseconds to 4 seconds but like I mentioned earlier the audio fidelity drops quite a bit with longer delay times, mostly at times longer than 2 seconds but that doesn't have to be a bad thing. It has quite a cool distortion effect. With the longest delay times you do get some clicks and artifacts mixed in the audio but it's not much. The delay times are controlled by the two transistors forming a voltage controlled current sink. The 47 Ω resistor at the emitter of the bottom transistor determins the shortest delay time while the 330K in parallel from the collector to ground determins the maximum delay time.

In my own build I did notice quite some dead space at the beginning (ccw side) of the 'Time' potmeter but lowering the value of the 47 Ω resistor didn't do anything. 

I urge you to download the PDF accompanying the original project. It has a comprehensive description of how the circuit works and what all the components do. 

Here's a block diagram of how the delay works. This is also from the PDF that comes with the build instructions on the BMC website. 

Audio in 2 is the Return input and it has the Direct Output normalled to the socket switch. So if you take the direct output into an external effect module and take the output from that module and connect it to the return input you can have an external effects loop going, creating all sorts of possibilities. You can, for instance, lead the direct output into a lowpass filter and have the VCF out connected back to the return input.

HOW TO PATCH UP THE MODULE:

To get the best out of this module you need to make a synthesizer voice in your modular synthesizer where this delay sits behind the VCA at the end of the signal chain. You can also patch it up so that the delay sits inbetween two VCA's and have the second VCA opened by an ADSR with a slow Release time. That way you get more control over the Delay time, but it's not necessary. The minimum Delay time is 60mSeconds so it won't be able to create flanging or chorus effects. But you can mix in the effect with the clean signal by using the Mix control and the Mix output.

LAYOUTS:

Here are the layouts I made for this project. They are verified as always. I used them to build my module. This was almost another hole in one. I made one little mistake. I had all four non inverting inputs of the TL074 grounded only the last opamp with the direct output must not be grounded. Once I corrected that the circuit sprung to life. Pins 5 and 10 of the TL074 are connected through the strip underneath the chip. The 'Tone' control potmeter has pin 1 not connected. It's important to wire it the way you see in the layout or it won't work properly.

Wiring:

Stripboard only:

Cuts and wirebridges. You know the drill, mark the cuts on the component side using this guide and then stick a pin through the marked holes and mark them again on the copper side. Then cut the marked positions with a hand held 6- or 7mm dril bit.

Don't forget to cut position P-8 underneath the ground wirebridge.

Here is the Bill of Materials. 

It might be a good idea to use a logarithmic 100K potmeter (A100K) for the return potmeter. A lot of changes happen quite early in the throw of that potmeter. However I used a linear 100K myself and that works fine too. But the log type would be more convenient. You could use other value potmeters for all but the Tone Control. That has to be a 10K linear potmeter. The other potmeters are just voltage dividers in this circuit.

PICTURES:

Here are some pictures from the build proces:

I left out the two short wirebridges that connect all 3 ground strips at the eurorack connector together. Instead I soldered them together with some extra solder bridging the gaps.

Stripboard all wired up for testing. I normally only wire things up when I have the panel ready so I can keep the wires as short as possible but with this module I had to be sure first that everything worked. Anyway, it made mounting the board behind the panel easier coz no need for soldering and I was able to stuff all the wiring underneath the stripboard out of harms way.

This is the panel with the waterslide paper applied ready to receive a final thick coat of clear lacquer. The panel is 14hp wide (7CM). the width I normally use because it allows me to mount the stripboard flat behind the panel keeping the depth to a minimum.

Here's the panel design I made in Photoshop just in case you want to use it. It's in A-4 format 300pix/Inch resolution.

Module on the test bench:

The rear of the module. It's 3.8 cm deep so it will fit any Eurorack case.

VIDEO DEMO:

Here's a little demo I recorded showing the module in action.

Here's an interesting look at the inside workings of the PT2399 chip: --- click here ---

Okay, that's it for this one. Hope you like it.

If you have any questions or remarks about this project please put them in the comments below. Remember comments are moderated so they don't appear straightaway. Only after I read them.

You can also post questions on the special FaceBook group for this website.

If you like this content and would like to support these projects and the upkeep of the website which is presented to you free of costs, you can buy me a coffee. There's a button for that underneath the main menu if you're on a PC or Mac. You can also use this PAYPAL.ME link to donate and cut out the middle man. All donations go straight back into new projects. Thank you!

Synthesizer Build part-67: KASSUTRONICS PRECISION ADSR.

The best version of the 7555 based ADSR's on this website. This one uses precision rectifiers to eliminate the problems the previous versions have. This project is small enough for Eurorack and runs fine on dual 12V or 15V and is easy to build even for beginners.

This is another version of the two 7555 ADSR's you've already seen on this website. The previous ones by Yusynth and Rene Schmitz had the problem that, because of diode voltage drop, the envelope wouldn't get down to 0 Volt after each cycle. The diode in series with the release potmeter would stop conducting when the voltage dropped to the threshold of 700mV in case of a 1N4148 and around 300mV for Schottky diodes.

This ADSR eliminates that problem.

By using precision rectifiers made up of a diode inside the feedback loop of an opamp, you solve the problem of the 700mVolt remaining after the release cycle and so the ADSR not returning to 0 Volt after each cycle. The opamp now has that voltage drop in the feedback loop and compensates for it, effectively creating a perfect diode.

I tried to address the voltage drop problem in the Rene Schmitz version by using Schottky diodes that have a very low voltage drop of about 0.3 V (300mV) and that already helped a lot. This version lowers that even further although on my oscilloscope I could still measure a tiny bit of voltage left over but the majority of that was due to the capacitor I was using. It was about 90mV. I used a normal electrolytic capacitor for testing. I then tried a Tantalum capacitor and that lowered the offset to around 10 to 20mV. That's almost the noise floor so really no problem what so ever. It's 35 times better than using a 1N4148 in the Yusynth ADSR. The reason for this is transistor resistance. The Gate voltage, if switched by a transistor, never reaches zero because of transistor resistance. But this is such a low voltage that you can totally ignore it. So please don't go fretting about 20 thousandths of a Volt. 20mV is equal to 0V!!

Use a 1µF Tantalum capacitor like the schematic says. The slowest risetime you can create with 1µF is 1.2 seconds. If you want longer risetimes you need to connect two caps to a switch so you switch between low and high speeds. That's up to you. I didn't include that option in this project but it's very easy to implement. 

If you want to read more about this ADSR then here's the link to the Kassutronics webpage.

SCHEMATIC

I made some changes to the design of this ADSR. For one, I don't like the high value resistors on the gate input. I always get problems with the gate pulse not getting through. So because the Rene Schmitz version works so well I copied the Gate/Trigger section from that ADSR and put it in this one too. It's practically the same circuit but with different resistor values.

I also changed the inverted output to an attenuverted output. I find that much more useful because you can play with the attenuverter while you're feeding the ADSR signal into the CV input of a filter and get all sorts of cool sounds from it. You can turn it into a normal output if you need an extra output. Much more versatile I think. The schematic below has all the changes I made included.

Eventhough I used BAT43 Schottky diodes for D1 and D2 in the layouts below, you can just put in 1N4148 diodes. The voltagedrop isn't important here and both diodes have the same switching speed of 5nSec. They are just used here as reverse voltage protection.

LAYOUTS:

Below are the layouts I made for this project. As always they are verified. I used them to build my ADSR and it worked rightaway. An other hole in one. 

I alterred the layouts a little one day after posting this article in so far that I added a transistor to drive the LED to avoid pulling down the envelope voltage. 

Wiring: (All potmeters viewed from the backside!) As you can see the potmeter wiring is a bit complicated looking so be accurate when wiring up the pots! 

Stripboard only:

If with testing you notice that the envelope doesn't come up when the Attack potmeter is fully closed then use a 330 Ω resistor in series with the Attack potmeter (R8) instead of the 100 Ω in the schematic. This is something I had to do with my build. 

Cuts and wirebridges seen from the component side.

You know the drill, mark the cuts on the component side, stick a pin through the marked holes and mark them again on the copper side and then cut on the marks.

Here's the Bill of Materials.

Not every component is numbered exactly as in the schematic but most of the resistors are. Order a Tantalum capacitor for C3 1µF/35V. You can use any type of 7555 timer chip. I used the ICM7555. Don't use a normal NE555 though. It might work but they're not ideal. It needs to be a CMOS type.

As usual I didn't put any decoupling caps in but if you want to include it, there's room enough on the stripboard. You can put two 100nF caps; one from plus to ground and one from ground to minus. If you feel you need extra stabilization put some 10µF caps over the power rails too. That's up to you, the ADSR works fine without them.

PICTURES and test results:

This ADSR has a very fast risetime. I measured risetimes of 550µSeconds! The output amplitude of the envelope has a maximum voltage of 8.4 Volt when you run this on a +/-12 Volt power supply. Maximum Sustain level is 8 Volts. This is determined by pin 6 of the 7555 (Threshold) which stops charging the capacitor at 2/3rds of VDD. (+8V). The timer stops and the capacitor is discharged through the Decay potmeter and U2-D and D4 to the Sustain level. The output will stay at the Sustain level until the Gate input stops. Then the capacitor will discharge through the Release potmeter, U2-A and D3 to 0V. As I mentioned before, the maximum risetime of the Attack phase is 1.2 seconds with a 1µF cap. If you need longer times you can put a 1µF and a 10µF on a switch and connect that to the stripboard, so you have a choise. The fast times sound amazing though when used on filters (especially the 303 filter).

Here are some pictures of the finished product. They were taken in the test phase so some components that are on the layout are not in these pictures (like the LED driver transistor for instance).

I used the same faceplate as I used for my previous 7555 ADSR's. I just exchanged the stripboard for this one and wired everything up again.

Here are some screenshots from the oscilloscope. The first one shows the extremely fast risetime of this ADSR/ With Attack set to zero you can get risetimes of 550µSeconds. This with a Tantalum cap and a 330 Ω resistor in series with the Attack potmeter, instead of 100 Ω in the schematic.

Below is a screenshot of the quickest pulse I could get with all potmeters closed. You can see the risetime is the same as above, about 550µSec and the releasetime is faster because it only has 100 Ω in series. It's about 400µSec.Total time is 992µSec. So you could create a waveform with a top frequency of 1kHz with this ADSR.

Below here is the normal and inverted output. The voltages indicated by the scope are a bit lower because I had the LED connected straight to the output. I now have the LED driven by a transistor which means no voltage pull-down so the real maximum voltage is about 8.4 Volt. Max sustain voltage is 8 Volt as is the case with all ADSR's that use a 7555 and are run on +12V because that voltage is 2/3rds the voltage of the positive powerrails. If you run it on +/-15V it would be +10V.

The sustain is actually very stable because of the use of precision rectifiers. There is no leakage of voltage from the sustain stage.

Well, that is pretty much all I have to say about this. It was a pretty fast build, done on a sunday afternoon and because I re-used the faceplate and potmeters etc it wasn't that much work. I hope you enjoy building it. This is without doubt the best one of the three 7555 ADSR's on my website.

TIP: using your ADSR as a VCO. Send the squarewave output of a VCO to the Gate input of the ADSR. Now your ADSR acts as a VCO and with the Attack and Decay you can shape your own wave. It's a trick used in the Psy-Trance. This ADSR is fast enough to do this. I tried it and it sounds pretty cool when you then input it into a filter.

There is one more ADSR design that tries to really come down to zero volts after each cycle and that is the ADSR PRO by Davor Slamnig. You can visit his website by clicking here.

If you have any questions or remarks about this project then please put them in the comments below. Comments are moderated and don't appear straightaway!

You can also post your questions on the special facebook group for this website.

If you like this content and you feel you learned something and you want to make a gesture of appreciation you can buy me a coffee. There's a button for that underneath the main menu if you're on a PC or Mac. You can also use this PAYPAL.ME link and cut out the middle man. All donations go straight back into new projects and the upkeep of the website. Thank you!