MFOS schem questions (VCCS)

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, Joe., analogdigital, infradead, lisa, parasitk, plord, sduck

Post Reply
User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

MFOS schem questions (VCCS)

Post by Orgia Mode » Wed Nov 20, 2019 3:40 pm

I am trying to tune a 13700 in the floating point resistor setup and came across MFOS's Sub Commander schematics which I believe shows an exponential Voltage Controlled current source.:

Image
original


I have my own linear VCCS schematic that seems to work very nicely in my VCA, but I want to know more about the MFOS VCCS.

I can't get it to simulate in LTSpice. EDIT: I mistakenly disconnected the power on one op amp. It simulates now, but the window of current sourced is quite small.
I still have the questions:
1. Does this setup need a tempco (R74) and matched NPN's (Q3 and Q4)?
2. What purpose are R77 and R78? The LM13700 has a maximum rating of 2mA on the Iabc pin and coming from the collector pin of Q5, the source is already a current, right?
You do not have the required permissions to view the files attached to this post.
Last edited by Orgia Mode on Wed Nov 20, 2019 5:01 pm, edited 1 time in total.

jorg
Wiggling with Experience
Posts: 378
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Wed Nov 20, 2019 3:47 pm

This looks like a DIY question.

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Wed Nov 20, 2019 4:33 pm

Yep, the curse of having too many tabs open at once. I chose the wrong one.

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Mon Nov 25, 2019 8:21 pm

Playing around a bit more I was able to widen the resistance range and better calculate each section of my own board using the floating resistor model.

Should be fun to make a fast VC LPF. I just have to determine the different values I need for each section. 5-poles should be a good start.

User avatar
guest
Super Deluxe Wiggler
Posts: 3605
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Mon Nov 25, 2019 9:16 pm

that is a bizarre circuit, and i wouldnt reccomend using it. switch the transistors to PNP (2N3906), and just feed the OTAs directly, rather than going thorugh Q5. R75 will need to go to -12V as well. R77,78 help balance the current into the 2 OTAs, and also limit max current. not strictly neccessary, but a decent thing to do. you might want some temperature compensation, but for a VCA its not really required.
openmusiclabs.com

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Mon Nov 25, 2019 11:23 pm

guest wrote:that is a bizarre circuit, and i wouldnt reccomend using it. switch the transistors to PNP (2N3906), and just feed the OTAs directly, rather than going thorugh Q5. R75 will need to go to -12V as well. R77,78 help balance the current into the 2 OTAs, and also limit max current. not strictly neccessary, but a decent thing to do. you might want some temperature compensation, but for a VCA its not really required.
Thanks guest! I'll update later when I simulate both of the versions, but for now how would you suggest using a current mirror instead of a bunch of parallel resistor on one 3906? For my purposes, I have 6 LM13700's.

User avatar
guest
Super Deluxe Wiggler
Posts: 3605
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Mon Nov 25, 2019 11:29 pm

the issue with 6 13700s, is that you need to decide how much current you drive the OTAs with, and how much headroom you want to have. the 13700 likes 500uA as a max, with 250uA being a more reasonable top limit. but, multiplying these by 6 gives 1.5mA to 3mA, which is a bit much for an exponential converter to run at and still be exponential. it will tend to go flat at the top, which might be ok for a VCA, as it will act sort of like a compressor for high input values. using more transistors in parallel will give more current, but it might make the currents not match exactly to each one.
openmusiclabs.com

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Tue Nov 26, 2019 2:35 am

guest wrote:the issue with 6 13700s, is that you need to decide how much current you drive the OTAs with, and how much headroom you want to have. the 13700 likes 500uA as a max, with 250uA being a more reasonable top limit. but, multiplying these by 6 gives 1.5mA to 3mA, which is a bit much for an exponential converter to run at and still be exponential. it will tend to go flat at the top, which might be ok for a VCA, as it will act sort of like a compressor for high input values. using more transistors in parallel will give more current, but it might make the currents not match exactly to each one.
Oh no, not that way. I don't want to multiply the current; I want to mirror the one current converter across 6 separate OTAs. Kinda like this:
Image

I am using 6x LM13700's as 6x floating point resistors in a VC LPF. If I mirror the current to each section rather than just splitting the current with 12 resistors in parallel, the floating point resistor calculations will be more accurate.
Image
You do not have the required permissions to view the files attached to this post.

User avatar
guest
Super Deluxe Wiggler
Posts: 3605
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Fri Nov 29, 2019 1:33 pm

there are a number of different ways to mirror the current. the quickest path from where you are at, is to take the bases of your PNP transistors, and connect them to the NPN emitters. then collect the collector of the NPN to ground.
openmusiclabs.com

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Wed Dec 11, 2019 2:57 am

Thanks so much for your help. guest. I have limited my Iabc current to 500uA because my simulations seem to like that value, but I can easily drop it down to half that if needed.

I've finished up a rough schematic, without precise values, and am going to breadboard a filter right away so I can fine tune the resistor. Bread boarding has always been a dreadful chore for me. :bang:

VCF QUESTION:
Would it be weird or backwards to have a low pass filter roll off at 20kHz with an input of 0v and a roll off of 20Hz with an input of 10v?
I've looked at commercial filter specs and they basically just say "acceptable input range 0-10v" and nothing else. Some even say "1v/Oct tracking" but again, no specifics beyond that.

jorg
Wiggling with Experience
Posts: 378
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Wed Dec 11, 2019 9:40 am

Orgia Mode wrote:Would it be weird or backwards to have a low pass filter roll off at 20kHz with an input of 0v and a roll off of 20Hz with an input of 10v?
Yes, it would be weird and backwards. For your own use, do whatever you find convenient. But generally, nobody else would find that convenient in a modular.

User avatar
Orgia Mode
Common Wiggler
Posts: 106
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Wed Dec 11, 2019 2:33 pm

jorg wrote:
Orgia Mode wrote:Would it be weird or backwards to have a low pass filter roll off at 20kHz with an input of 0v and a roll off of 20Hz with an input of 10v?
Yes, it would be weird and backwards. For your own use, do whatever you find convenient. But generally, nobody else would find that convenient in a modular.
Thanks!

Post Reply

Return to “Music Tech DIY”