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

exp fm vs lin fm on a vcf?
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Author exp fm vs lin fm on a vcf?
numbertalk
Hi Tony,

Could you help clarify when I might want Lin FM versus Exp FM on a VCF? I notice your Super Ladder VCFs have both. Any patch examples would be great. I know for a VCO this comes into play when using audio-rate FM for those sorts of effects, but not sure about linear FM on a VCF.

Thanks!
juice721
Want to hear too! Mr. Green
Synthbuilder
numbertalk wrote:
Could you help clarify when I might want Lin FM versus Exp FM on a VCF?


It simply gives you another type of sound. Whether that sound is useful or not is going to be very subjective.

One could argue that using expo curved EGs through a linear input will sound more natural than the double expo curve one gets with a expo EG going through the exponential convertor on the standard V/oct input. Remember that most EGs are have expo curves and that we typically use these with linear VCAs.

I guess that we are so used to hearing expo EGs sweeping expo filters that we expect to hear the double expo effect.

What I can say is that I don't think the linear option has enough merit to warrant its own socket on the standard 1U panel. But if you have the extra space on the larger panel then why not.

As for sound samples - all my VCFs are in the 1U format so I'm afraid I'm going to have to leave that for someone else to do. Sorry. Dead Banana

Tony
ndkent
Does this filter self-oscillate? If so then you would have a great advantage with having both kinds of inputs. You could use the expo to make the filter track your controller and then use the linear to create classic tracking FM timbres you wouldn't get if you just expo inputs.
numbertalk
Thanks for the explanation.

Also, interesting that does this filter self-oscillate comes up - it does, but I wanted to mention to Tony that I could not get it to track anywhere near 1V/Oct - is this ok? Is there something possibly wrong with my filter? I just left it as-is because I don't really use my VCFs as oscillators but I wouldn't mind it being a little closer for other types of keyboard tracking when being used as a filter. I tried calibrating this with octaves over the entire range of the filter (up until it got too high or low for my tuner, which was really high/low to be very useful for a filter anyway) and I wasn't able to get any of them very close.
Synthbuilder
numbertalk wrote:
but I wanted to mention to Tony that I could not get it to track anywhere near 1V/Oct - is this ok?


Yes, it should track. It does tend to go a little flat at the higher octaves though which probably means any precise FM tones would be hard to create reliably.

Do you have the 1U filter core version? If so this is does not track 1V/octave natively from the the front panel FREQUENCY socket. The sensitivity of the input socket is around 0.5V/octave to allow for big EG sweeps. It is expected that any frequency tracking adjustments are done in the external CV mixer.

The 3U version with it's dedicated 1V/octave input should track just fine unless there is a problem with either your module (or my calculations...)

Is it going too flat or too sharp as you move up the keyboard?

Tony
numbertalk
Hi Tony,

I'll have to check tonight when I get home if it's going flat or sharp - don't remember offhand, this was last week when I finished it.

I built the 3U version. It's possibly, though doubtful, I connected the Keyboard CV jack to the wrong pad on that row of CV inputs. I followed the schematic closely and it should be right, but possibly the issue. Will report back after I get a chance to look into it.

Thanks!
numbertalk
Also what frequency range would you say over which it will track better or it makes the most sense to use to calibrate the scaling?

Thanks.
Synthbuilder
numbertalk wrote:
Also what frequency range would you say over which it will track better or it makes the most sense to use to calibrate the scaling?


I would concentrate on the usual keyboard range. eg set for two octave spread at A 220Hz and A 880Hz.

If you can't get it to track try altering R54 to firstly 180K and if that doesn't work then 220K.

Tony
numbertalk
Thanks Tony. Planning to look at this over the weekend. Will update you when I do.

Synthbuilder wrote:
numbertalk wrote:
Also what frequency range would you say over which it will track better or it makes the most sense to use to calibrate the scaling?


I would concentrate on the usual keyboard range. eg set for two octave spread at A 220Hz and A 880Hz.

If you can't get it to track try altering R54 to firstly 180K and if that doesn't work then 220K.

Tony
numbertalk
Hi Tony,

So it's sharp - as far as I can trim it, it's still a full step and then some sharp. I don't have a 220K resistor here but I do have a 180K - swapped it out for R54 and it's better - now it's .5 steps sharp, but again, the trimmer can't go any further. Should I try 220K? It seems I got closer by lowering R54 - should I try an even lower resistor value there?

Any idea why I might be having this issue? I checked and all the components are correct. Not quite so concerned about getting it to track perfectly, more concerned with it being SO off that there's something wrong with my filter and that it might not be giving me 100% in some other way if this is the case. It *seems* fine otherwise, but I'd like to know why it's so far off from tracking properly and make sure it's not effecting something else in the filter as well.

Thanks!
Synthbuilder
numbertalk wrote:
It seems I got closer by lowering R54 - should I try an even lower resistor value there?!


Yes. Perhaps 150K.

But that is way lower than the theory or practice should dictate. To be on the safe side I rechecked the maths on Friday but it really should work with R54 as 180K.

What are you using as the temp co (R65)? Temp co resistors do sometimes vary quite a bit from their nominated value.

You could also try a new BC549/BC550 in Q9.

Is your 1V/octave input connected to the pad labelled Key-CV?
numbertalk
Synthbuilder wrote:
numbertalk wrote:
It seems I got closer by lowering R54 - should I try an even lower resistor value there?!

Is your 1V/octave input connected to the pad labelled Key-CV?


No, it's connected to pin 1 of the CV Inputs set of pads. I'm not at home at the moment to be able to check but is there a pad actually labeled "Key-CV" that I should connect this input to? If so, this is probably the issue!
Synthbuilder
numbertalk wrote:
No, it's connected to pin 1 of the CV Inputs set of pads.


Yes, that's the correct one. Pin 1 is the Key-CV input pad.

So I'm not sure what is wrong with your scaling.

Try one more thing, with the power off, measure the resistance of R65. It should be around 1K.

Tony
numbertalk
Thanks Tony - I'll do that this evening.

Could very well be Q8/Q9/Tempco - I used thermal adhesive on the transistors and bonded the tempco on top of them as well. I've done this a million times, even a few times like this one, bonding to 2 transistors where the tempco needs to be extended and bent a bit, but this time it was a little weird...a little adhesive got onto one of the transistor legs. I checked them and the transistors seems ok, but maybe something funny is going on there.
numbertalk
R65 measures exactly 1K. However, Q9 is the one that got a little of the thermal adhesive on it. Not sure if that could really have an effect, I've always read to be really careful about getting it on component leads. It's not caked on or anything, but it's there. Otherwise Q8 & Q9 seem to test out ok with the diode tester on my DMM. Any other way I could tell (besides it not tracking 1V/Oct smile ) if either of those transistors aren't doing what they should be - any readings I could take from there? In the meantime I'll try a 150K resistor too.

Thanks!
numbertalk
Update - tried 150K, did not work. Funny thing, though, it was too flat, but then wouldn't trip up far enough either. So I tried replacing R54 with a trimmer but still could not get it to trim to 1V/Oct.

Even though all seems ok there, since I had a little trouble with the adhesive, should I try replacing Q8 & Q9? Any other measurements or anything I can check there to see if they seem ok?

Also, I noticed another thing I wanted to verify if it sounded normal to you or not - the filter oscillates fine, however, if I set the frequency (to a frequency high enough to oscillate) and leave it with no FM at all, and I turn the resonance pot up, it will not break into oscillation. To get it to oscillate I have to leave the resonance pot up and then turn the frequency pot up then down before it will start. I've never seen a filter do this - normally just turning the resonance up to the point of oscillation will make it ring. Is this how this module is supposed to behave?

Thanks for all the help. Haven't had this much trouble with a build, especially a straight-forward one, in a long time. very frustrating
Synthbuilder
numbertalk wrote:
Update - tried 150K, did not work. Funny thing, though, it was too flat, but then wouldn't trip up far enough either.


Sounds like the combination of R54 and the trimmer need to be around 175K. Which I'm sure should be wrong if everything else is fine. Right now I don't have a clue why its behaving the way it is.

Maybe try 30K in R64 and 180K in R54.

numbertalk wrote:
should I try replacing Q8 & Q9?


It maybe worth a try. I'm going to have a play with my Superladder and see if I can get some voltage readings you can verify against.

numbertalk wrote:
To get it to oscillate I have to leave the resonance pot up and then turn the frequency pot up then down before it will start.


Nope you shouldn't need to do that. Maybe TWEAK isn't set high enough?
Synthbuilder
One thing you may want to check if you have a decent voltmeter. The voltage across the temp co should be decreasing 18mV for every 1V on the key-CV input (ie. -18mV per octave). It's probably easier if you measure the change over four octaves which should give you around 72mV.

This should check the gain of the CV summing circuit and the temp co divider.

Tony
numbertalk
Synthbuilder wrote:
One thing you may want to check if you have a decent voltmeter. The voltage across the temp co should be decreasing 18mV for every 1V on the key-CV input (ie. -18mV per octave). It's probably easier if you measure the change over four octaves which should give you around 72mV.

This should check the gain of the CV summing circuit and the temp co divider.

Tony


It's really close - ~-/019V. Hmm. So sounds like things are ok there, right? Does this mean the transistors are probably fine? Any other ideas?

Also, resonance is weird - I was able to trim where it would oscillate by just turning the resonance pot up, but the range is super-limited - only oscillates for high frequencies. So re-trimmed it to have a fuller range but now will not oscillate simply by turning up the resonance pot. Any other ideas for this one?

Thanks again!
numbertalk
And yeah, tweak was (and is) all the way up with described behavior - good full range of oscillation but have to move the frequency knob to get it going.
Synthbuilder
numbertalk wrote:
Any other ideas?


Without it here in front of me I can't really suggest anything else.

I'm thinking that both your problems could be related to one thing. In this case it would be perhaps the ladder transistors themselves - but transistor problems are pretty rare these days. If you have a spare THAT300 try swapping it over with the one in the board.

13700s can be a bit variable too - but the TWEAK control should cover all eventualities.

Badly matched capacitors can also have an effect on resonant behaviour - but again I wouldn't expect to see such a problem with modern devices.
numbertalk
I hand-matched the caps, and they're polypropylene. I did a diode test on the 2 parts of each transistor on the board and they seemed ok but I can check again.

Is there not a resistor value I can tweak for the resonance issue? I'm guessing not, since it is oscillating and the range is fine, it's just exhibiting that strange behavior. Maybe I'll swap out the 13700.

Thanks.
numbertalk
Swapped out the 13700, no difference. I've noticed not only do I need to turn the frequency knob to get oscillation but I have to turn it higher. Ugh.
MUFF WIGGLER Forum Index -> Oakley Sound Systems  
Page 1 of 1
Powered by phpBB © phpBB Group