controlling an analogue Ross Phaser by MIDI

[Guest chris.pickett]

Hi,

I think this is the right forum for this kind of question ...

I would like to build a Ross Phaser

http://www.runoffgroove.com/ross3.html

but modify the design so that I can control it with CC's (or even NRPN's for smoother resolution).  I don't think the guitar people can help me with this.

Can the MIDIBox software and chip help me do this?

I guess I need some way to take a MIDI signal and control a knob or button (does it need to be converted to CV first?  In which case am I better off buying a MIDI-->CV converter and then worrying about interpreting a CV signal?)

I've never done this kind of thing before, and actually I won't even be the one building it, but I'm designing it.  My brother is going to build it at school, so I can communicate with him if there's problems.  I have some first year college experience with circuits, and don't mind learning.

So ... I'll read FAQ's and tutorials and stuff if this is already covered somewhere before, but I wanted to check that I'm asking in the right place -- because I'm not sure I even need MIDIBox at all (maybe it would be more appropriate to manage a stompbox or something).

I would be extremely grateful if someone could let me know where to get started on this, whether it's possible or whatever.  Ultimately I'd like to build a whole bunch of analogue effects and have them completely controllable by MIDI and have the order of the effects switchable.  Of course I'll make any designs available so that other people can build similar things once it works.

Cheers,

Chris Pickett

P.S.  Of course it would be nice if the controls on the phaser sent MIDI as well ...

[TK.]

Hi Chris,

this is one of the easiest jobs what a MIOS core can do... ;-)

The current MIDIbox 64 already provides 64 CV inputs and 8 CV Outs. 64 CV Outs are planned (I'm waiting for some ordered sample&hold chips to test this possibility).

Maximum Resolution of the CV Ins: 10 bit

Maximum Resolution of the CV Outs: 12 bit

NRPN is not provided by the MB64 application by default, but could be implemented in a special application dedicated for CV

Best Regards, Thorsten.

[Guest chris.pickett]

Hi Thorsten,

Thanks for the reply.

So, I can get some CV outs / ins.  That's great!  How then do I modify the controls on the effects pedal to send and receive CV (more importantly receive)?  Is this relatively simple?

(if there's a document I should read about all this in particular, feel free to point me to it -- I can't seem to find the answer anywhere)

Cheers,

Chris

[TK.]

Hi Chris,

you can control the effects from the MIDIbox, means: just use 10k pots and connect them to the analog inputs, and use an AOUT as CV out. The MB64 application just forwards the CV Input to the CV Out, and it allows you to send and receive MIDI and to store/restore settings (unfortunately only 16 setups with each BankStick, since it's optimized for 128 MIDI entries per bank, but you could also change the software for storing 128 setups and more if less number of entries are required)

I just noticed that this phaser has no CV inputs, but requires "digital pots", which are similar to a DAC (and identical from the software point of view), but which provide a variable resistance instead of a variable voltage. You will find some infos concerning digital pots in older articles (use the search function of this forum)

Btw.: I guess for such Fx gear 7bit resolution is more than enough!

Best Regards, Thorsten.

[Duggle]

hi guys,

one problem with using digipots for effects can be the digital noise getting into the signal if there is some signal gain involved (usually the case).

Another approach is to use a CV to bias a FET transistor to achieve a variable resistance! This is done with a couple of resistors an a low cost feild effect transistor.

It should be straight forward to use analog techniques to provide taper characteristics other than linear (another limitation of digipots).

Let me know if this sounds interesting to you, I may be able to find some links again (its been a while since I experimented with this).

cheers

[d2k]

Hey Duggle...

I'm interested in this method, so some more info would be great :)...

The idea is to eventually create a mios controlled analog desk :D

Cheers

Dan

[Guest chris.pickett]

i'm also interested duggle ... especially if the digipot thing has that problem.  *edit*  then again, the phaser was originally built with them, and it sounds great, so whatever.

thanks again though thorsten.

i think a midi controlled analog desk ... could be interesting.  i'm not even aware of such a thing even existing already on the market.

techniques should be similar to what i want to do in any case, at least at the "turning a knob" and "sliding a slider" level :)

if you run out of CV i/o, is there some way to chain chips together?

[Duggle]

Ok,

I had a look at the Ross Phaser circuit and have also located the article that I used for my previous experiments.

http://graffiti.virgin.net/ljmayes.mal/comp/vcr.htm

As you see in the article the simplest application involves 2 resitors and FET but one side of the "variable resistor" is connected to ground.

this is ok, because many circuits function with one side of pots connected to ground eg vol pot.  It would be possible using an op amp to implement a floating ground that would create a more general "varaible resistance" but it would add some complexity.

Back to the Ross Phasor the "recycle" pot is grounded so the simple FET resistor can be used in conjunction with a fixed resistor to control regen via CV.

Interestingly, the phasor itself uses a very similar FET configuration to control the characteristics of filter stages. An internal CV is generated to modulate these.

U4B is an LFO which provides this CV.

My recomendation is to inject a MIOS CV at this point in the circuit  ie. at RV4 and delete the LFO from the assembly.

FETs work and sound great for instrument effects (they exhibit subtle distortion similar to valve). However for general mixing application they probably lack fidelity.

Automation is a challange for analog circuits of you want pristine audio. Logarithmic motor faders could be used for volume and linear ones for pan and eq but it starts to get expensive and complex (how many MIOS cores would you need?).

Digipots do not have any noise if there is no digital data be transmitted through their shift registers. It may be possible to transmit data to them to implement a "program change" or "scene change" concept. In between  changes there would be no digital noise.

Digipots also exhibit nonlinearity which would make the purist audiophile cringe (does this really apply to instrument effects and synths?).

cheers

[edit] The reason for replacing the internal LFO by a MIOS derived CV is that it would allow for tempo synced sweep  and other complex behavour of the phasing sweep. (It also makes the circuit a bit simpler!)

[Captain_Hastings]

Hey Duggle, that is great schtuff!

I just finished building a Big Muff and connected it to an Analog Devices AD5204 Digital Potentiometer controlled by MIOS... it picks up a lot of noise and hum indeed. Especially the sustain pot in the first amp stage is very sensitive.

I'll be trying shielded cables between the analog and digital circuit (the Muff and the digipots are on seperate boards) which should help a little bit...

If that doesn't help i'm going to have a look at the method you suggested!

[Duggle]

Yes Capt'n, the bigmuff circuit has a lot of gain so noise will be an issue regardless of control method.

Seperate boards is probably not the best in this case.

However, you should be able to improve it

Sheilding and elimination ground loops is a must to minimize hum.    

The problems I had were definely digital signals.

If you stop data to the digipots does the noise reduce?

[Captain_Hastings]

Seperate boards is probably not the best in this case.

However, you should be able to improve it

Sheilding and elimination ground loops is a must to minimize hum.

Well my reason for the seperate boards was that i intend (*if* all goes well  ;D) to connect multiple digipots to multiple effect boards. Some effect require a combination of 10K, 50K and 100K therefore i thought this was the best way to go...

I also wanted to seperate the digital stuff and the analog stuff (thought that i would otherwise have even more noise problems...)

The Muff is running off a 9V battery for the moment, so digital and analog psu/ground are seperated.

The problems I had were definely digital signals.

If you stop data to the digipots does the noise reduce?

No the noise pickup is there all of the time, varying if i touch the wires that connect the sustain pot.

If there's only audible clicks when controlling the pots i don't mind, i don't plan to use it that way, just to be able to store/retrieve/edit patches is what i aim for...

[moebius]

Other and more general suggestions, maybe:

Later versions of Ross phaser uses OTA's (13600), those are CURRENT controlled amplifiers, used in synths ect. Remember your (not so) boring basic electronics cource: U/R=I

(R25, http://www.runoffgroove.com/ross_phaser_black.png), so any OTA based design is controllable by varying voltage. (But remember to read OTA datasheet for max Iabc)

Many classic phasers (and some filtering fx) use LED/LDR devices. These can be voltage controlled, by supplying leds with proper voltage.

GeoFex has a useful circuit for modifying phaser LFO circuits accepting VC:

http://www.geofex.com/circuits/ldrlfo.htm

Using Vactrol as variable resistor.

Bye, Moebius

[Duggle]

The  LDR/LED technique has the advantage that the control and signal circuits are completely isolated. If you mount the LDR in the effect you can have longer leads to the control circuit.

Capt'n, the concept youve devised requires running wires between digipots and effects. High impedance parts of the circuit that are not intended to be connected to wires will mean that bad noise is highly probable. In fact the noise your experiencing now is probably masking the underlying digital noise. If you only want patch change then ok(as you point out).

Digipots are not isolated and therefore must be connected (dc coupled) you may be better of with the digipot ground actually connected to analog signal ground (and connect the shields of your wires to this ground!). Just make sure the voltage rails of the bigmuff are less than the voltage of the digipots bias rail(s).

The ones I used were dallas semi and had +/- 5V bias inputs so I was able to control parametric filters with +/- 5V split rail. You should be able to run your bigmuff off 5V no problem. I haven't seen the datasheet of your digipot so this is just general advice.

[d2k]

Hey Guys...

How's about implementing a control system on a pultec eq clone from gyraf: http://www.gyraf.dk/gy_pd/calreq/calrec.htm

Could the pots and switches be replaced by digital ones and controlled via mios? (guessin a digital neg log pot might be hard to find tho ;p)

Great stuff

Dan

[Duggle]

d2k, this could be done using digipots and cmos analog switches, or with LDR/LED and relays. Either way there would be a rather significant design effort/experimentation and likely some performance degradation from the original circuit.  

[moebius]

Hi d2K..

I think that best way to control that kind of stuff is using ("shaft thru body") normal pots and behind those MIOS controlled motor driven pots (alps), so that shafts are connected. I hope that someone understood what I meant.. no ASCII this time ;)

And switches could be CMOS or relays as Duggle mentioned.

This method preserves most circuitry intact and probably

wouldn't introduce much noise.

I know: expensive and mechanically difficult method, but that's _the way_ when messin' around with (well, cloned) pro audio gear mostly found in places, where it's more likely to see 72 Flying Faders than 8 ;D

Bye, Moebius

[Duggle]

I agree with mobius, in fact if you used if you used signal quality sheilded relays mounted "in" the PCB and motor pots there should be no sound quality degradation  from the original at all.

cheers