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

Filter Set-Up
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Author Filter Set-Up
eljay
I have recently built the COTA and SVF filters, they work very nicely. However I was reading some of Tony's general documentation and came across a line about the Oakley CV/gate Buss that said "Any VCO and VCF can be connected to the Oakley Buss's KCV line, and this will save you having to patch KCV to every module that needs it". Both of the filters have 1V/Oct input sockets but I didn't spot a Buss interface on the PCBs, have I missed something?

When calibrating both COTA and SVF filters I 'ran out' of adjustment on the 20K multi-turn trimmers for setting the V/OCT value. in both cases I very nearly got to the setting I was aiming for but I'd reached the limit of the trimmer with an octave range just short of the right amount. Whilst the filter results are fine I was sort of expecting to find the correct setting somewhere in the 'middle' of the trimmer value and wondered whether I had inadvertently missed the 'sweet spot'. I have tested the values of both trimmers and they are 20k. I only ask because all other calibration settings on all other modules have been achieved as specified in the User Guides.
Synthbuilder
Generally I would expect a trimmer to be set between 25% and 75% of its travel. But if you can't calibrate it at all then either I've made a booboo or there is something else wrong. Usually, if I've made a calibration booboo, Paul at Krisp1 will find it and let me know. Paul makes a lot of boards of mine and that sort of thing would show up quite quickly.

So the first thing to check is that your KCV bus is actually 1V/octave. If that is OK, maybe check R21 (COTA) and R35 (SVF) are both 220K, and R31 (COTA) and R29 (SVF) are both 120K, and R32 (COTA) and R39 (SVF) are both 33K.

Tony
Synthbuilder
eljay wrote:
Both of the filters have 1V/Oct input sockets but I didn't spot a Buss interface on the PCBs, have I missed something?


No, you've not missed anything. Ages ago when the Oakley Buss was first created and before the creation of the Sock4, 6 and 8 PCBs I envisaged all modules having dedicated KCV and gate buss inputs. In practice only the gate inputs to the EGs and 1V/octave inputs to the VCOs were typically used. I think a lot of people prefer the filtering to not follow the keyboard exactly. Indeed, my preferred amount is somewhere around two thirds.

With the introduction of the 1U wide Filter Core module format the KCV to filter connection became even less important. It is expected that you will be using a mixer (most usually the Multimix) to control the various CVs that are controlling the filter cut-off frequency - therefore routing KCV to the Multimix is the most likely scenario now.

So none of the current batch of VCFs support the Oakley Buss natively. However, you can do what we used to do and that is to give your module a tail. A simple connection from the NC lug of the 1V/oct input socket taken back to the KCV buss will work very well. However, this is somewhat problematic when you use the socket board as you may need to cut the copper track leading to the NC lug in question.

Tony
eljay
It's been a while since I made the original post and having checked the resistor values were as specified above I decided at the time that the 1v/oct trim error was small enough not to be a real problem.

However, it still bothers me that I have the same calibration issue on two of my filters (COTA and SVF) and wondered what other factors to consider in achieving the 1v/oct calibration setting.

With nothing supplied to the filter inputs and a 1V/Oct PITCH voltage supplied from the midiDAC, I set the resonance/feedback pot to max and used an oscilloscope to measure the the 24db output. I selected a note to play (A4) and adjusted the frequency pot until I got a self oscillating frequency of about 880hz, I then played the note an octave above (A5) and observed the new frequency. This process was repeated a number of times until I reached the end of the 20k trim pot (approx. 17.6K). At this point I had reduced the 1v/oct error from 99hz flat to 15hz flat.


I'm assuming I could modify the value of one of the resistors to bring the trim pot range up, what should I try?
Synthbuilder
Try making R31 on the COTA and R29 on the SVF both 130K. In theory 120K should work but maybe the resistance of those temp co resistors tends to the low end.

Let me know what you find and I can change the Builder's Guide accordingly.

Tony
eljay
Thanks for the suggestion Tony. I have changed R31 on the COTA to 130k. I have recalibrated using the same method as before and this time I am able to find a setting near the middle of the trim pot that gives the 1V/Oct result I was expecting. I was able to demonstrate that going either side of this point either sharpened or flattened the frequency gap dependent on the direction of adjustment.

So, a resounding success, I will make the same adjustment to the SVF and I anticipate the same successful outcome but will let you know if there is any difference.

Many thanks, I know it's pedantic but I like to get the results you give in the builders guide, otherwise why bother calibrating?
Synthbuilder
eljay wrote:
I was able to demonstrate that going either side of this point either sharpened or flattened the frequency gap dependent on the direction of adjustment.

Thanks for letting us know. It seems that we do indeed more gain around the CV summing amplifier. My prototypes were able to be tuned up with the values calculated by theory. But clearly there is something amiss with my assumptions, and quite possibly it's the nominal value of the temp co.

Quote:
I know it's pedantic but I like to get the results you give in the builders guide, otherwise why bother calibrating?

Not pedantic at all. This stuff should work as advertised so I am always keen to hear from folk who find it doesn't.

I'm going to have a play at some point with a COTA I have here and some newer temp co resistors and see if I can replicate your results.

Tony
eljay
I have now modified R29 on the SVF to be 130K, as expected this change now enables me to achieve the expected calibration for the 1V/Oct setting. Many thanks, an easy change to make.


For your information just in case it makes a difference, I am using metal film resistors throughout and TL072CP opamps from TI. The tempcos are as supplied by yourself.
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Page 1 of 1
Powered by phpBB © phpBB Group