Author |
Message |
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Mon Sep 26, 2016 5:34 pm Post subject:
Expo Conversion for a Filter |
|
|
I am working on a filter and I am wondering if someone could point me to a typical expo conversion schematic that would be suitable for this purpose. It doesn't need to be as precise as an oscillator, but it should track reasonably well.
Can I get away without using tempcos? _________________ JacobWatters.com |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Tue Sep 27, 2016 2:11 am Post subject:
|
|
|
It would depend on the filter .. like .. a(n OTA based) current controlled filter can use the typical voltage to current converter (as often seen in VCOs). _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Fri Sep 30, 2016 7:15 pm Post subject:
|
|
|
It is an OTA based filter. So this one should work?
http://schmitzbits.de/rbecomp.gif
Any recommendations for the transistors? Do they need to be matched?
Also, that schematic doesn't really tell me where the input is - I assume to the left of the 33k resistor? _________________ JacobWatters.com |
|
Back to top
|
|
|
Feifer
Joined: May 16, 2012 Posts: 54 Location: San Diego
|
Posted: Sat Oct 01, 2016 1:52 am Post subject:
|
|
|
If you don't ever plan on using it in self oscillating mode and having good control over the tone, then you don't need a tempco or matched transistors. But if you think you might sometime down the line it's good to future proof it, I think. 2n3904 , 2n2222 , etc works just fine. You're correct that to the left of the 33k would be the voltage input. The open wire on the right is the exponentiated current source.
Good luck.
Also, that René Schmitz tutorial is great. |
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Sat Oct 01, 2016 6:00 am Post subject:
|
|
|
I didn't matched all my trannies in my previous builds which needed matched pairs and they worked well or the VCO's tracked nicely over 4 oktaves or even more... I only used trannies from the same batch glued them on the flat side together and on the side i glued my KTY sensor. I think that with todays industry standards the trannies are mostly in a close enough range... I also never used a tempco... i only used those KTY temperature sensors Rene Schmitz mentioned on his site...
Matched trannies and a tempco would improve many things, but if you have some OTA's driven with that it's maybe more likely that these have different specs and doesn't behave closely similar... noticed that on some of my builds with LM13700... Then changed the OTA and got better results... _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 7:26 am Post subject:
|
|
|
Thanks for the info Feifer and wackelpeter. I will give it a try today. _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 8:24 am Post subject:
|
|
|
Rene mentions the resistor at the bottom is 1M, but he doesn't mention the one at the top.
Also, the capacitor value is not mentioned.
What is the reason for two 2.34k resistors instead of just one 4.7k? _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 8:47 am Post subject:
|
|
|
I did it as best as I could on breadboard, but it is not working. Sending it voltages range from -12V to +12V gives and output ranging from -0.85V to -0.61V.
Does anyone have a complete expo conversion schematic with all values in place? A real idiot proof schematic. _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Sat Oct 01, 2016 9:31 am Post subject:
|
|
|
A Juno style VCF? Is it like most Roland Filters based upon 4 OTA stages?
A few months ago i build a Roland System 100 LP/BP filter and noticed that Roland mostly used the same principle on all their VCF's just with different OTA's and buffers for the output...
Build mine with LM13700 and 2SK30. _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 9:37 am Post subject:
|
|
|
Yes it is four LM13700 OTA's and then one more in the feedback loop for resonance control.
Thanks for the help. _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 10:26 am Post subject:
|
|
|
I have the TH one kinds of working. The audio completely cuts out if I send a positive CV - it has to be negative for some reason. My filter works from -12V - +12V, so I need that full voltage range. _________________ JacobWatters.com |
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Sat Oct 01, 2016 10:52 am Post subject:
|
|
|
well Thomas has some offset trimm on his OTA's if this would work with the roland-style VCF with 2 OTA's it would be okay, but adding 4 trimmer for each stage... should work but a lesser parts solution would be better... you could run the control for the bias for the OTA each through a separate resistor or send them combined through one resistor... don't know if this would work with the offset control... guess no as each stage maybe needs somewhat very different offset voltages to work properly...
edit: in Roland's VCF design 2 OTA's are one filter stage... The first being the BP and the second the LP... so probably you can go with 2 Offset trimmers one for each stage, connecting the Offset voltage to the positive input of both OTA's in that stage...
Please come up with some positive results, as i maybe plan to modify my VCF too as i'm a bit unhappy with it's range or general behaviour with CV and pot rotation... it's somewhat too sharp acting for my ears...
Next edit: the offset maybe seem to be more likely for the signal rather than the CV... a simple inverting amp with unity gain after the expo circuit wouldn't work. right? _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Sat Oct 01, 2016 11:20 am Post subject:
|
|
|
The voltage measured at the output of the expo converter is not very relevant .. as it is a current source, or sink rather and not a voltage source.
When you want to test it, either:
* Connect a resistor (like 30 k or so) from the converter's output to Vcc+ and then neasure the output voltage, also calculate the current then from Ohm's law.
* Or alternatively use a current meter, minus input connected to converter out, plus input to Vcc.
With either method make sure the current can not go over the maximum Iabc of the OTA .. which the data sheet can tell you - I forgot, but either 1 or 2 mA or so.
When that is all OK you can connect the converter to the Iabc inputs of the OTA's.
To get the control range OK you will likely have to add a summing opamp before the converter, and I've often seen a negative voltage bias being applied at the input to get he range OK.
Do NOT add an opamp after the converter, you will likely blow out the filter OTAs. Do use resistors there.
Oh, and yes, you do need two resistors from the output one for each OTA, trimming is probably not needed. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Sat Oct 01, 2016 11:56 am Post subject:
|
|
|
Blue Hell wrote: |
Do NOT add an opamp after the converter, you will likely blow out the filter OTAs. Do use resistors there. |
oops... hopefully cynosure made it into the bed meanwhile and hasn't tested that... _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sat Oct 01, 2016 4:52 pm Post subject:
|
|
|
Thanks Jan and wacklepeter.
I have it working now using the TH schematic, but it is inverted. I am testing the voltage that I send into it (before R20 in the TH schem). The cutoff gets higher as the voltage is lowered - to opposite of what I thought.
Also, the control is all in one side of the pot - nothing happens when the voltage is below 0. Is that normal behavior for a filter?
After the TH expo converter I have a 33K resistor going to each OTA.
I am going to keep messing around with it. Any suggestions for _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sun Oct 02, 2016 11:36 am Post subject:
|
|
|
The MFOS one just plain did not work. The TH one works, but it also adds some frequency modulation - I don't know where that is coming from.
Still nothing that I can use _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sun Oct 02, 2016 11:44 am Post subject:
|
|
|
The frequency of the modulation is 60Hz, which makes me think that it is coming from the power supply. It wasn't there before doing the conversion. Any suggestions for getting rid of it? I have bypass caps on every chip and at the power regulators. Could is just be because everything is on breadboard? _________________ JacobWatters.com |
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Sun Oct 02, 2016 12:07 pm Post subject:
|
|
|
some beads or 10-22ohm resistors into the power rails where they go into the circuit? Guess that you don't have them on breadboard... and then some electrolytes 10-20uF... then the hum should go, if not it's maybe coming from somewhere else... maybe catches this over air from some nearby electrical circuit (like a lamp or something else) or some semi-optimal grounding in your circuit... _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sun Oct 02, 2016 1:26 pm Post subject:
|
|
|
I always have a small amount of 60Hz hum when I breadboard, but it goes away when I solder the circuit together. If I understand the expo converter correctly, it looks like the initial voltage range is reduce drastically. There are 100k resistors going into the first opamp, but the feedback resistor is only 2k. That would make my CV reduced to a level that is close to that of the 60Hz hum. I will continue assuming that all is working fine and hope that it goes away when I solder it up on perfboard.
Thanks for all the help! _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Tue Oct 04, 2016 3:20 pm Post subject:
|
|
|
I am still trying to get this to work correctly. It drastically limits the range of the filter. The highest cutoff frequency is just over 3k when using the expo conversion from the TH schematic. Any idea why the range is limited? I was able to increase it slightly by changing R11 to 4.7k, but that only gave me about 100Hz more.
EDIT: I reduced R12 ti 1k and managed to get the cutoff up to 4.9kHz. That is better, but still way too low. Any other ideas? _________________ JacobWatters.com |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Tue Oct 04, 2016 5:45 pm Post subject:
|
|
|
I finally got it working by changing the caps on the filter. It is weird - with just a pot as a voltage divider going to the OTA's the filter worked great across a wide range. Adding the expo circuit severely limited the range and I had to edit the actual filter.
Maybe I am still not doing it correctly? _________________ JacobWatters.com |
|
Back to top
|
|
|
bubzy
Joined: Oct 27, 2010 Posts: 594 Location: United Kingdom
Audio files: 64
|
Posted: Fri Oct 07, 2016 9:08 am Post subject:
|
|
|
Hi jake, i know i've arrived at this party too late, i would totally recommend thomas henry's "21st century synth book" and the "making music with the 3080" book too the latter has some amazing expo circuits in it! _________________ _Richard_ |
|
Back to top
|
|
|
|