vocoder speech in

This section is very similar to JH FS-1 so I used the same values. It seems that everyone else who has completed this section copied the first suggested values using THAT 1510 instead of SSM2019. I had some leftover SSM2019 from building my FS-1.

The values I used are as follows

Tone:

TONE Pot 50k linear

U2101 = SSM2019

Sp Level Pot - 10k log

Mic Gain Pot - 5k lin however I will get a 5k reverse audio for my panel

R2105 / RG2 = 10r

R2106 / RG1 = link

R2107 / RG0 = nothing (left open)

Limiter:

R2116 / RG3 = 22k (based on advice in e-m thread)

R2136 / RLIM = 6k2 (based on advice in e-m thread)

There was note that the U2101 would become unstable at certain settings and a large capacitor was used in place of RG2 but i didn't have that issue on first test. 

Mods (not yet implemented):

I wish to have an aux input on the synth in. If you read texts on vocoders they suggest that either the synth or vocal input should be prerecorded, only one input should be live. Whether or not this is gospel is up to the critics but i would like the opportunity to insert a line level (or modular level vocal input  - or perhaps a feedback input). I will follow the circuit used in the FS-1 input and send an input into U2102B. I'll document that once I have made some more progress on these sub-modules.

I am happy with the output thus far. The limiter light is coming on when it should ( i think ) the sp out is not distorted so time to move on.

jh vocoder progress

Finally some time to work on the vocoder has opened. At this point I have finished the panel layout for panel components and I have created a layout to mount the sliders. This slider pcb is being etched for me at the moment. All connectors and final components are in stock.

The path to the finish line looks like this:

• test pcb
• psu
• synth in
• exc in
• all other modules
• one channel
• complete panel design
• aesthetics (fonts / linework etc)
• mount holes for case and pcbs
• case / faceplate fabrication
• wire up

Proud to say today I have powered up the board and the psu is functioning. I had a bit of a stumble trying to use a switching psu - I mistakingly bought a single supply instead of a dual. Not sure how I slipped there since I was specifically looking for a dual. Annoying. Luckily I had a 15v torroid in stock so I used the onboard psu. As far as I'm aware I believe I am the first to use this onboard psu. 

Here is the heatsink with the TIP3055 and mica pads

Next up I will test the modules

vocoder continued

Well this is turning to be a marathon pcb population. i've almost completed resistors and i've populated some other components outside the filters. After so long staring at resisters i think i have finally memorised resistor colour codes :)

And i reached a special milestone - i filled up fetta tin with resistor legs - i guess this was the project to do it

I have reworked my panel design also - this now includes a 20x20 ghielmetti patch matrix. 

Vocoder Begins

jurgen haible living vocoder

this project is finally coming together

most parts are on their way and i have done a preliminary panel design

having a huge internal debate whether a patch matrix is necessary and i haven't included it in my current design. I have a thought of having it on the back of the case. as you can tell the design is still in flux.

i think i need to borrow a nord modular and test it out

here is the current panel design draft

im reasonably happy with it - i think ill adjust the signal flow space so it is a little tighter. with this panel design i'll need a bit of extra circuitry including 

• speech input amp - line input
• speech input amp - led bar graph 
• excitation input - led bar graph

im also bringing almost every option onto the front panel except the patch matrix.... mmmm

the plan is to make a panel pcb - even if i end up etching one. i think i am coming around to the idea of panel pcbs and no wiring. it will take a lot of effort but hopefully it will make the unit easier to service and live longer. 

so waiting on parts and should begin once they arrive. i will build the unit and test some of the mods before builiding the panel. most likely i will get the front panel fabricated as part of the case. no idea how to get the decals on as it is way too big for my normal lasertran technique.  

Jurgen Haible FS1

it has been a while between posts - travelling the world and having fun back in a house in my home town and the studio is back together working on some new toys currently hoping to complete the jurgen haible (rip) FS1 frequency shifter here is my design


edit:

Here are my dome filter picks

I have added to jurgen's speadsheet to measure the time constant different between ideal and measured. My resuts were ok - not as good as I would really like but I cant be bother ordering more components.

I found that there is a lot of play when choosing values - it helps to have a lot of resistor values on hand to match together. I would recommend buying capcitors that are not 1% and make sure you get a few of them. The capacitor value add magnitude to your final result so in cases where you have a good resistor match but the capacitor value is large it can bump up the final error. Having a bunch of capacitors at slighly different value gives you more options of resistor combos.

edit 2:

I managed to get it to fit well into a case. Time to wire :)



Neural Agoniser PCB block diagram

I am doing a some servicing of my synth before i hand it over to a friend to look after while i go overseas. There were a couple of noisy sections in my NA which i wanted to resolve. I created this diagram to help me work out what is what. I hope it helps others also..