MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

VCO Style CV summer/tempco for DUSG/VCS 1V/OCT Input?
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author VCO Style CV summer/tempco for DUSG/VCS 1V/OCT Input?
TOPLEL
The 1V/Oct input on the DUSG/VCS is just a voltage divider with "dubious" tracking:




Can i substitute this section with a proper VCO V/Oct section? For example this one:






It looks like both of these just divide down the 1V into the 17mV ballpark, the second one just do it better.

Is it even worth it doing this or the tracking of the DUSG/VCS is flawed in other parts of the circuit too?
cygmu
I discussed this a while back with a friend and decided it is probably a good thing to try if you're putting together a new board or anything like that.

Someone had mentioned that the key to good tracking with a DUSG was to use transistors with high gain (hFE), which I didn't understand until I realised that lower hFE will mean more current drawn at the transistor bases which will load the input CV. Putting an op amp in the path will resolve that. There's also the CV at the other two transistor bases to worry about of course.
TOPLEL
cygmu wrote:
it is probably a good thing to try if you're putting together a new board or anything like that.


I am doing that, i am testing out JLCPCB and this circuit is probably the best candidate to have 5-10 PCBs of.

cygmu wrote:

Someone had mentioned that the key to good tracking with a DUSG was to use transistors with high gain (hFE), which I didn't understand until I realised that lower hFE will mean more current drawn at the transistor bases which will load the input CV. Putting an op amp in the path will resolve that. There's also the CV at the other two transistor bases to worry about of course.


Yeah now that i checked the schem it looks like i need 3 of these V/OCT circuits which might be an overkill. Maybe putting a 2k tempco instead of the 2k2 voltage divider resistors connected to the 2 bases can be helpful.

I have matched pairs of BC547Cs and 2N3906s but these pairs are not matched to each others, i hope that it is not a problem.
cygmu
You probably only need to temperature compensate the V/Oct input. The other inputs can be thought of as reference voltages for the V/Oct, sort of, and though there will be temperature effects from the components around there, they are the type of effects that are ignored in almost all synth designs as far as I am aware, so you will get away with it! It would be worth buffering those other voltages though. In the existing design there is an input buffer for the CV but then a passive summing stage with the voltage from the pots -- that's where I would insert another op amp.

If the PNPs are matched and the NPNs are too, you should be fine. The pairs are used independently, one pair for rise, one for fall, so they don't all need to match.

(I think!)
oldenjon
TOPLEL wrote:

Can i substitute this section with a proper VCO V/Oct section? For example this one:




You'll need to either change this to non-inverting or use an additional inverting stage to get the response you want. Otherwise, the response will be the opposite of what you want (increase in voltage = decrease in frequency).
TOPLEL
oldenjon wrote:

You'll need to either change this to non-inverting or use an additional inverting stage to get the response you want. Otherwise, the response will be the opposite of what you want (increase in voltage = decrease in frequency).


How does this work in VCOs? Is there an additional inversion happening in the matched pair?
oldenjon
TOPLEL wrote:

How does this work in VCOs? Is there an additional inversion happening in the matched pair?

The easiest way to explain it is that the differential pair has two inputs, inverting and non-inverting, just like an op-amp. Typical VCOs only use one of the inputs, but the DUSG is not a typical VCO. You can sum the 1v/oct input with the rise and fall inputs, but it will be less efficient.
TOPLEL
oldenjon wrote:
TOPLEL wrote:

How does this work in VCOs? Is there an additional inversion happening in the matched pair?

The easiest way to explain it is that the differential pair has two inputs, inverting and non-inverting, just like an op-amp. Typical VCOs only use one of the inputs, but the DUSG is not a typical VCO. You can sum the 1v/oct input with the rise and fall inputs, but it will be less efficient.


Would this work or it can be optimized further?
(i did not wanted to do a non-inverting oopamp with gain below 1 because i've heard different opinions about the stability of that)
(btw i think i can leave that 1k8 resistor off)

oldenjon
TOPLEL wrote:

(i did not wanted to do a non-inverting oopamp with gain below 1 because i've heard different opinions about the stability of that)
(btw i think i can leave that 1k8 resistor off)

You can't attenuate with a non-inverting amplifier, but you can buffer the input and use the original resistor values for the divider (with the addition of a trimmer). I don't think non-inverting would pose any problems as you're not summing the 1v/oct signal with anything. The 1k8 resistor isn't necessary if you go the inverting route.
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group