MonoMiK

Finally...

for nearly a year, i just didn't have the time to continue on this project. I had the idea of a simple polysynth with which i played around a bit, up to the first layouting of the VCOs, some firmware tests, looks promising. But as i have too many open projects, i set the PolyMiK on ice for now. Back to the MonoMiK. Since about 2-3 weeks, i'm working on the overall routing and control of this synth. Meanwhile, i switched to SMT for most of my projects, but the MonoMiK will continue to use THT at least for the last filter module and the second backplane. Those might be the last veroboard designs of this synth, as SMT gives me the possibility of much more features per space, so the modulation functionality will again be sort of full-blown, which in my eyes is what this synth deserves.

As i'm too lazy to solder stuff together currently, i'm working on the user interface. Besides the real analog controls on every module (8 modules, 4 knobs each), there will be 4 rotary encoders with push function and a 160x80 monochrome touch screen for the rest. Currently, i'm working on the firmware for this user interface, the current state is a prototype of the VCO configuration.

It's still a long way to go to have this synth playable, but at least the project is not dead :)

So, time to optimize signal count coming from the backplane. One important bunch of signals are the control voltages from the CPU. One idea is that you can store patches, even with those front panel knobs, which are true analog to the the attached electronics. There are some tricks that the CPU is able to scan those CVs in a simple way and then lock out the front panel knobs to provide the values from the CPU. To keep complexity low, only all panel knobs can be turned off together, so the software has to be able to help the user dial in the desired values when a stored patch has to be modified.

Besides those knobs, not much CVs are really needed. Each VCO pitch and VCF cutoff gets a separate CV line from the CPU, which controls keyboard tracking and larger detunes by software. The front panel knobs will be there for full realtime control without using the CPU, if front panel usage is desired. Older synths of the 80s use a similar way, but they usually have the CPU between the knobs and the synthesizer, so there is no infinite resolution for those controls. Here, it is not for stored patches, but it is for real front panel control.

As space on the backplane is very limited, i had to use SMD components here. The resistors you see in those photos are 0805 sized components. Not funny to solder them by hand to a veroboard with 100 mil hole distance. For size comparisons, the resistors are 80 mil long and 50 mil wide. 100 mil are 2.54mm, so you can have an idea about the size. During the soldering, one of the resistors fell down on the floor - no way to find it :)

Yesterday, i completed the next module - the VCO mixer 2 module.



It's nearly identical to the first mixer module, but doesn't contain the sub oscillators. Instead, it has a noise generator for white and brownian noise included. I will test this module today, but it will not result in a new video, as this module is nearly the same as the last one.

After a break for most the time of my vacation, i got back to development of the MonoMiK this week. Yesterday, i started soldering the next module, the second VCO mixer, nearly identical to the first one, but noise instead of the sub oscillators. The second sub oscillator channel is just the VCA stuff, while the signal comes from the backplane. The ring modulator will find it's place there, i think.

More to follow - i think, i will complete the module today.

Okay, my holidays started and i found the time again to continue working on this synth. So i completed the VCO mixer module, which has TWI-controlled analog switches (okay, standard 4053 analog switches with PCF8574 TWI/I2C I/O ports :) to provide the waveform switching. Two sub oscillators, one per VCO, are also on this module, finalized by 4 VCAs, which are controlled by the front panel knobs but are also able to get modulation signals. The 4 VCOs output the main and sub signals of the 2 VCOs.



To test everything, i needed to create a more sophisticated setup with a microcontroller switching the waveforms, while the selection vomes in via MIDI control changes. I also needed the wiring between the dual VCO module and the VCO miver module, so i decided to just add a bit more and include the VCF and the ADSR modules and peek into the current state of the development by listening to it. I think, the result is quite fine:



(The left one is part 1, the right one part 2)

Another new piece of electronics has been added. Those funny things are the small LED modules which go to the front panel. They are connected to modules and the backplane via adapter boards. Details will be seen in the allover construction photos which will follow in the future.



Schematics and my veroboard layout can be found in the Wiki article about the MonoMiK

All this was done in 2008, but since 01-Jan-2009, i have a new hosting plan again, which enables me to re-activate the website. This is one of the first posts on the new server with lots of more space :)

Completed yesterday, and today you get the photos and video, as promised.





Schematics and my veroboard layout can be found in the Wiki article about the MonoMiK

Finally, i found the time to complete my ADSR modules. They were the first module working perfectly directly after soldering them together, i didn't forget any wires or so. Maybe due to the very compact layout i just took more care on this. The only imperfection was a bug in the schematics, 2 swapped resistor values. This is the reason why i always build the modules and test them before releasing the schematics.

I'm very happy with the performance, attack time can go down to fantastic 30 microseconds, the official range starts at 100 microseconds - 0.1ms. Due to the exponential control of the timing parameters (attack, decay, release), the front panel knobs go from 0.1ms to 6s on the attack time (starting at 0V) and are quite comfortable to adjust. Using the CV inputs from the backplane, the time range can be extended a lot. Extremely long times can lead to accuracy problems due to leakage currents.

The design uses a very small timing capacitor (1nF) to get the quick attack times using a standard OTA, and exactly this capacitor is quite small for longer times. If you build the circuit, feel free to replace the capacitor by a larger one if you don't need the super-fast timing. The rest of the circuit doesn't need to change for this.

Tomorrow, demonstration video, photos and schematics will follow. The board layout will also be there, but this is just for completion. It might be very difficult to route this layout on a 2-layer PCB, but i needed to save space there because of the overall construction of the synth. I never take an enclosure and build the synth according to the size of the enclosure, this is unsuitable for my featuremania :)

So, my vacation is over. I didn't do a lot on the MonoMiK during the vacation, i was around a lot and besides this, i restored a Roland CR-78 and repaired my new Dynacord Echo King, besides fiddling together a replacement for the missing power transformer of the Lexicon Prime Time i a currently restoring.

I hope to get back building the ADSR modules and developing the SV filter module the next days.

Okay, found some time to continue the project the last days. Fascinating how i get more and more routine developing the modules with the time. I scribbled the ADSR on a sheet of paper sitting on the couch and watching TV. Calculated some resistor values and so, put everything together - my first ADSR.
I was some sort of fascinated that this circuit did everything i thought of when being tested on the breadboard. Some minor optimizations and completion of the functional tests were on Monday.
Yesterday i did the board layout and started the next module, the state variable VCF, to get the sizes of those last 3 modules (the dual ADSR comes twice), as they will not be full-length, i need some space at the rear for the connectors. Everything did work perfectly, so now i also know the size of the backplane and can continue with some mechanical planning.

By the way - the ADSR is working quite fine. Attack times down to near 0.1ms are possible, on the other end 30s are also no real problem. It's completely voltage controlled and uses an OTA to create the exponential charge/discharge curve of a capacitor. This design was a bit inspired from the Yamaha CS80.

As usual - schematics follow when everything is built and tested, which again might take a while.

Finally i found the time to create the audio comments on the video snipplets i collected during build of the ladder VCF module. Everything needed to get compresses to the 10 minutes limit of YouTube, but i hope you enjoy it.