For what its worth, i havent had issues with ceramic thus far. I do use a mylar for the 470nf caps though.
3 Likes
i use only ceramic in all my old ms20 clone filter even 470nf and all is good now i put some film),
but if you can for some strategic cap is better to put a film one
2 Likes
The main issue with ceramics is that thereās a whole bunch of different materials used for the dielectric, and theyāre all good at some things and really bad at other things (unless you explicitly get the ones that are good at more things, but they only come in tiny values). Well-known issues include:
- bad temperature stability, some types can drop 85% of their capacity as your board heats up (on the other hand, the C0G type is very good)
- voltage sensitivity, making them bad for AC signals
- piezoelectric effects, aka āmicrophonicsā. Ceramics are both microphones and speakers; theyāll pick up vibrations, and can also make audible noise.
ā¦and if you just got a bunch of āceramicsā off the Internet, odds are you got something thatās bad in multiple ways. Still good for powerline bypass, but probably not something that should be in audio paths if you can avoid it, especially not filters or oscillators. Also, box capacitors tend to look better 
9 Likes
@fredrik +1 on the box caps looking better. Easy to fit and generally handle too!
These are good things to keep in mind, but not obsess over if you are worried. ceramics are cheaper, and if thats all you got, it will probably be fine, but if its not, you can use these known issues to help deduce it was cause you cheaped out. We all can have different experiences here, but usually if we see problems, its not a result of the quality of component, its usually wiring errors, sketchy soldering, or the like.
I built the strip-board version of this a few times before it functioned properly. The type of components never changed, but each time one problem would go away and another would rear its head. It works fine now!
In your case though, you say:
When you say level, are you referring to the frequency range or the attenuation of the filter, or the overall amplitude coming out is lower?
- Plain english: (Does it seem like you cant sweep it much, that the effect of the filter is weak, or is that shit quiet when it comes out?)
- Plain french: (Semble-t-il que vous ne pouvez pas le balayer beaucoup, que lāeffet du filtre est faible, ou cette merde est-elle silencieuse quand il sort?)
3 Likes
Did you ever get this sorted?
1 Like
Anyone have an issue where the filter just whistles? No sign of the wave from the input, just high pitched self oscillation. Iāve been trying different transistors to replace the bc558. The shop said I could use a nte159, but that Iād have to spin it around. So far Iāve tried both ways with no luck. I hope I havenāt bricked the chips. 
2 Likes
The transistors are part of the exponential converter, and have very little to do with the signal path, so I doubt theyāre involved. You can use pretty much any small-signal PNP transistor there (as long as you turn it the right way, that is).
Iād check the feedback path (including the RV1 trimmer, the resonance pot, and U5A (the first opamp in the TL074)). Does tweaking RV1 change anything?
EDIT: To clarify, the transistors provide the āamplifier bias currentā that tells the LM13700 how much current to conduct, so if they donāt do anything at all Iām not sure any signal would make it through, not even the feedback that causes oscillation. You could measure the voltages across the 10k resistors that lead to pins 1 and 16 on the LM13700 respectively. That reflects the bias current.
4 Likes
Are you talking about if the resonance is high? Filters can self oscillate. To be fair, i think i experienced the same damn thing when i was banging my head with this. I know it sounds silly, but check and double check your grounds.
1 Like
not really. i resoldered everything, replaced all capacitors but it still behaves strange in high pass mode when the cut off is high.
i thought it might have something to do with me using a 5k pot for the cutoff instead of 4.7k, but since i canāt find 4.7k anywhere i will probably keep it as is. not using high pass mode too often anyway.
also in another thread i read there have been issues with the input resistors. maybe try changing their value and see if that helpsā¦
1 Like
Thatās a voltage divider, so the value doesnāt matter much. The original MS20 uses 100k, the Schmitz clone a 47k, and 4.7k might just be a typo in Clacktronicsā layout that Sam based his design on.
Also potentiometers arenāt very precise things, even quality ones are usually Ā±20ļ¼
so designing a circuit where the 6% difference between 4.7k and 5.0k mattered would be madness 
(and 4.7 multiples arenāt really a thing these days, pots use the 1/2/5 series)
4 Likes
Iām glad to hear this. I had loads of 10k pots, so I just threw that in there. I did notice that if I had the res. turned up all the way and tried turning the cutoff up, the circuit would cut out. Like i was overloading something.
I know that mucking things up is a great way to learn, but holy hell is it frustrating.
2 Likes
I forget what I used, but it definitely wasnāt a 4.7k. Pretty sure itās a 5 or a 10. Also I never adjusted the trimmer as the self oscillation kicks on at the perfect spot for me whatever it was set on from the factory
1 Like
yeah. do you have this issue in both modes or also just in HPF? i think when i have all my other modules done i will try experimenting with the input resistors and if that doesnāt help i will keep it as it is
Correct! As far as I know our hear is more developed in that range because, when we were still primitive humans, we needed to hear if something dangerous was coming. In fact, wood cracks, leaves sounds and noises like that are in that range!
The circuit only cuts out if the cut off is turned up in the one mode. Iām pretty sure itās the hpf.
Another strange thing I noticed is that I get almost no voltage on the collector of q1/non inverted input on pin 5 of the tl074 until the cutoff is almost all the way up, then it jumps from almost nothing to 8v in about an eighth of a turn. That canāt be right.
I double checked that Iām using a linear pot. So, Iāve got a bit more grinding to go.
1 Like
What signal are you feeding it with? Sounds like you might be driving it too hard:
EDIT: What does ācollector of q1/non inverted input on pin 5 of the tl074ā mean exactly, btw? Theyāre in different parts of the circuit.
2 Likes
I finally finished the performance filter! It seems to work except for a ground issue. It hums until I touched the jack sockets. It doesnāt matter which input i use. This is my first diy build ever. Where do i have to look for it?
3 Likes
itā seems to be an electrical ground problem, check and re check all your ground connection.
what sort of power supply have you, have you test it with mutilmeter before connect module ?
2 Likes
I use a TipTop uZeus. I will check all connections again 
. Its a learning process. Thank you
3 Likes
Hey folks,
So I finished my first Kosmo Module build last night (1113 RS filter) and I need some help troubleshooting. Pictures attached.
-
The right sided LED lights and stays lit when I plug in the module to my power supply. The left sided LED does not light under any circumstances.
-
Input oscillator signal from my Korg into 1 seems to be passed unchanged to the output (When the switch is set to high pass) but no signal when it is set to low pass.
-
Turning the filter knob does not change the sound quality at all.
-
The level knob on input 1 works.
-
Turning the input knobs on 2 or 3 (the jacks not in use) can lower the volume of the output signal.
-
The CV knobs donāt do anything (donāt have anything plugged in there so that tracks).
Backside:
Closeup of the Jack Socket Wiring - Did I wire these up right? (blue is supposed to be ground and yellow supposed to be signal)
Any ideas you guys have about what I did wrong would be helpful. Iām sure the polarity on the diodes, LEDs, electrolytic caps are all correct (triple checked). The rest of the solder joints look solid but tomorrow Iāll probably try and re-flow everything.
I didnāt check the leds before I installed them, could one be burnt out and could that do it? Trying to get some current on it to test it directly.
Did I do something dumb like wire up the signal of the jack sockets to the ground circuit or something?
The ICās are the correct way round. My TL074 is a TL074CN, not sure if thatās a problem or not.
I donāt think it matters which orientation the switch is soldered in, but if it does I could.
I donāt think messing around with the trimpot is gonna solve this but Iāll have a go Saturday.
Thanks in Advance for any advice.
2 Likes