Question about circuit modelling in SDK

Shroomz~> · Post by **Shroomz~>** » Sat Jun 03, 2006 7:16 am

Hi,
If I were to want to model an analogue or hybrid circuit component by component in the SDK, how would I go about it? Is it possible?

I've read that Flexor modules were built using basic building blocks which are available. Has anyone else got their teeth into this who is willing to share info on the concept?

Thanks

at0m · Post by **at0m** » Sat Jun 03, 2006 7:34 am

http://www.musicdsp.org has plenty links. If you understand the maths used there, you can apply it to Scope. I'm not saying Adern uses these, but reading there will give you good insight on how filters and oscillators are made on a low level...

Shroomz~> · Post by **Shroomz~>** » Sat Jun 03, 2006 7:41 am

Thanks Atom, I've had a good look through that site before. To be more specific, what I'd like to get my head round is how to achieve the realisation of a specific variant in a specific circuit in the SDK. The circuit I have in mind seemingly uses digital inverters instead of op-amps & this varient alone gives the circuit a very specific sound. Any advice on the implementation of such an idea in Scope would be fantastic.

Thanks

astroman · Post by **astroman** » Sun Jun 04, 2006 3:01 am

nah, it's not that complicated...

...The common source MOSFET is the basic circuit in Anderton's Tube Sound Fuzz. A similar circuit is used in Fender's Stage Lead, and EH "Hot Tubes" pedals. This circuit can produce very convincing tube-like distortion if it is carefully designed. The commonest way to do this circuit is with the CD4049 or CD4069 CMOS logic (yes, logic) IC. It can be misapplied by biasing it into its linear region and to function as an amplifier.

or the 4007, that's featured in my Nobels DT-SN pedal, see a filter app <a href=http://www.uni-bonn.de/~uzs159/vcf4007.html>here</a>

cheers, tom

<font size=-1>[ This Message was edited by: astroman on 2006-06-04 04:15 ]</font>

Shroomz~> · Post by **Shroomz~>** » Sun Jun 04, 2006 4:12 am

Thanks for the info.

I was kind of hoping an SDK circuit modelling guru might drop us all a little building tutorial based on an example circuit implementation or something

I know it's maybe unlikely, but hey, you never know

astroman · Post by **astroman** » Sun Jun 04, 2006 11:27 am

what a nice opportunity to throw a can of gas into the fire...

you probably remember that a certain kind of 'disrespect' from my side towards the 'open source scene' was based on a lack of originality and own ideas, so to say...

no need to personify this hypothesis by looking for a blueprint in which you can simply replace some variables with your own items

...and hope to get along without even having 'thought' about your project in detail.

otherwise you would have noticed that circuit modelling is hardly applicable, as the respective part is operated beyond it's regular specs.
The acoustic result is a smooth transition into (close to perfect) symmetric clipping, so almost no even harmonics are generated - but a significant amount of noise.

It's not depending on variations of the circuits operation - it's either there or not.

A ladder filter (f.e) would be a different case, as it depends (mainly) on non-linearities of it's transistors.
Even then it's not the transistor itself that's modelled, but it's significant parameter(changes).

The trick is to find out which are the significant items and to which math rule they obey - that's all you need to know to start circuit modelling.
Of course a quality signal source plus a (non-SFP) scope and a multimeter will help a lot on more complex items that are not covered by a (single) manufacturer's doc.

yet programming is more about detecting operation patterns and rules - not so much about hacking examples

it's up to you to decide if you want to face the challenge or become one among the average.

cheers, Tom

Post by **irrelevance** » Wed Jun 14, 2006 6:53 am

It's an interesting thought though. I've not looked too deeply into the low level stuff, but it would be quite a good lesson in music based electronic circuit building if it were possible, kind of like being able to run electronics workbench through scope

Thanks for the link Astro

Shroomz~> · Post by **Shroomz~>** » Sat Jun 17, 2006 9:26 am

It sure would be cool to have all the necessary components as modules called actual electronic component names with variiable value settings for each. I think J9k has been thinking & looking at the SDK in this way to a certain extent, considering some of his modular module designs.

symbiote · Post by **symbiote** » Sat Jun 17, 2006 12:57 pm

just pickup any analogue/digital electronics book, and start by building an ideal/practical mathematical model of the circuit you are trying to recreate. an ideal model will be simpler, so start with that, then you can start to add the practicalities and slowly build on this.

mathematical models in electronics are usually very precise, once you have something close it'll mostly be a matter of tweaking your model until it gets close enough (or sounds good enough compared) to the original. for a really precise model, you'll need pretty precise (and also pretty expensive) measurement equipment. some of it can be rented though, for the final tweaking.

also it might be a good idea to model the circuit in a circuit simulation program like spice or electronic workbench or some such first, so you can run simulations on the whole/parts of the circuit, in order to better understand it's behavior and what is going on.

sadly it won't ever be as simple as just dropping a few blocks and getting a perfect recreation, it'll take some time, patience and hard work, but it's definitely do-able.

Shroomz~> · Post by **Shroomz~>** » Sat Jun 17, 2006 3:00 pm

Thanks Symbiote!

cheers

PS. anyone else who can add something of worth is more than welcome.

<font size=-1>[ This Message was edited by: Shroomz on 2006-06-17 16:02 ]</font>

Post by **irrelevance** » Sun Jun 18, 2006 10:38 am

An interesting link:

http://graffiti.virgin.net/ljmayes.mal/ ... hmodel.htm

j9k · Post by **j9k** » Wed Jun 21, 2006 6:01 pm

one of the things that will kill you right off the bat is x+y=(x+y)-1 above zero in the mix2 mdl. you have to correct for it if you want to do any sort of integration(like the trapezoid mdl). but this only comes into play in a few circumstances.

j9k

Shroomz~> · Post by **Shroomz~>** » Thu Jun 22, 2006 12:26 pm

This does indeed kill me right off. If you can help with that by posting a more detailed description I might just grasp the point

wolf · Post by **wolf** » Thu Jun 22, 2006 1:40 pm

well, it already starts with shit raining from heaven, as soon as you want to do additions/substractions with sync signals exclusively. Afterwards the death is at least a variation

at0m · Post by **at0m** » Thu Jun 22, 2006 7:40 pm

Shroomz,

I think j9k means that in the total 2^32 numbers, there's 2^31 negative numbers, 0, and then "2^31 - 1" positive numbers.
There's no center position here in 2^32 or 32bit land, but the 0 is the first positive number. Try doing some proper maths with that

Shroomz~> · Post by **Shroomz~>** » Fri Jun 23, 2006 3:38 am

Got it, thanks

j9k · Post by **j9k** » Sun Jun 25, 2006 2:36 am

1+1=1
2+2=3
10+10=19
2147483647*1024=1023