Adding TAP-TEMPO functionality to old/analog effects

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

clintrubber

Well-known member
Joined
Jun 3, 2004
Messages
5,982
Location
The Netherlands
Hi,

I've been thinking a bit about adding tap-tempo functionality
to some old analog effects like S/H-controlled filters, amplitude-modulation-FX etc.

So I'm looking for/want to make an add-on circuit that can synchronize an
LFO to a footswitch-tapped tempo.

I figure that the tap-button we see on digital-FX-boxes will be a simple
effortless requirement; I guess the existing DSP just counts how long
between pulses.
But if one would want to make a stand-alone add-on to an existing circuit,
what would be the best way ?


Have some thoughts about how to realize it,
but so far the circuits seem a bit overly complex.

Some thinking out loud:

* Counters: hardware-realization with off the shelf gates will
likely be too complex; (but I have no experience with a PIC/...-realizations)

* 'Analog' realizations with voltage controlled LFO's may suffer from drift.
Let's see what a stability-requirement could be. Let's just assume a 3BPM-drift over a
180BPM three minute song is tolerable (without intermediate tapping to correct).
Uhh, that's an allowed drift of some figure, but how stable are the musicians
themselves actually ? Without click, likely they won't be much better than
that 3BPM-shift.
Actually a bit hard to define a requirement for the taptempo-circuit here,
putting this one aside for now.


Apart from people adding footswitches to existing circuits,
I haven't seen any DIY-taptempo circuits so far.

Anyone ? (Suggestions for asking this somewhere else also appreciated.)


Thanks !

Peter
 
Peter,

What is the most common way (analogue-wise) for the LFO rate to be set in a normal analogue pedal?

Is it voltage controlled? Does the "Rate" pot usually just have the wiper connected to an IC to set LFO rate?

If this were the case (or similar), it might be fairly simple to program a microcontroller IC to set LFO rate by counting "taps".

So, how are the LFOs controlled in the effects you are thinking of?
 
What is the most common way (analogue-wise) for the LFO rate to be set in a normal analogue pedal?

Is it voltage controlled? Does the "Rate" pot usually just have the wiper connected to an IC to set LFO rate?
Like you say, a pot, but making it voltage controlled wouldn't be hard. To make it working well we need of course to know how that delivered control-voltage translates to the LFO-frequency. Will require a trim of some sort.

If this were the case (or similar), it might be fairly simple to program a microcontroller IC to set LFO rate by counting "taps".
This is where I understand what can be done but where I have no experience of doing it.

So, how are the LFOs controlled in the effects you are thinking of?
Pots forming the variable-R element in RC-relaxation oscillators. To facilitate external control some have a provision the use an external pot, often by putting something instead of the internal pot.

Some provide an additional ready-to-go section awaiting an external pot. Both pot-actions get summed then for LFO-control.

If you will, for illustration of the latter situation there's one of the mothers of all analog FX, the USS-1A Universal Synthesizer System, of Maestro (by Oberheim).
Schematics are already on-line:
http://www.diystompboxes.com/pedals/Maestro_USS-1A.zip
This box has all sorts of effects and a provision to connect external controls (pedal-pots) for speed (of the phaser & S/H) & frequency (filter).


But I'm wondering if the 'add-on' couldn't better deliver a LFO-signal by itself ? So the existing LFO is set aside and in its place this thing that could probably have a more direct translation of tap-pulses to the LFO-rate (and this perhaps even with an adjustable integer multiplication-factor).
Would this 'direct conversion' suit the microcontroller IC approach ?

Thanks,

Peter
 
I had an idea some time ago about a digitally controlled analog tremolo circuit with a PIC controlling the led of a vactrol. I havn't built it yet but I will... some day.

/Anders
 
I think you would really want to go with a micro controller to do this, it would get abit complex and big in the analog world, at least any way I can think of doing it. You would need to replace the LFOs most likely, You most likely want a decent exponential or linear control voltage LFO so the tempo you tap can easily be converted into the control voltage the LFO needs for that pitch. But even that is getting to large to fit into alot of pedals. External LFOs might be the best way to go, just add a switching jack to switch out the internal LFO when you plug in the external. This also allows you to interface them with modular synths or use fun stuff like envelope generators/fallowers and any other CV modual to be used to control the effect.

Here is a pretty good VC LFO with exponential response to control voltages that is stable and should stay in time once set there.

http://www.musicfromouterspace.com/analogsynth/lfo.html

It gets fairly complex fast, but most of this is the waveshapping circuitry, 1 quad opamp and 4 transistors is all that is needed for the core of the oscillator. It may just be best to do the LFO in the micro also if your programming skills are capeable. Although digital LFOs for modulation never sounded right to me, you need to have some error in there, at least with slow sine waves on filters or VCAs. I would personaly go with the analog VC LFO controlled by a micro, but that may just be overkill.

adam
 
[quote author="adamasd"]I think you would really want to go with a micro controller to do this, it would get abit complex and big in the analog world, at least any way I can think of doing it. [/quote]
Same feeling here. It's the field I know, but not the best approach indeed.

You would need to replace the LFOs most likely,
No problem, replacing the existing LFO looks to most tidy way.

But even that is getting to large to fit into alot of pedals. External LFOs might be the best way to go, just add a switching jack to switch out the internal LFO when you plug in the external. This also allows you to interface them with modular synths or use fun stuff like envelope generators/fallowers and any other CV modual to be used to control the effect.
Adding it to existing pedals wouldn't be required, adding that switching jack keeps everything tidy.

Although digital LFOs for modulation never sounded right to me, you need to have some error in there, at least with slow sine waves on filters or VCAs. I would personaly go with the analog VC LFO controlled by a micro, but that may just be overkill.
'Some error', as in a distorted sine (but the very same for each cycle) or you feel the need for cycle-to-cycle variations ?

& thanks for the link, more interesting stuff there.


from Anders:
I had an idea some time ago about a digitally controlled analog tremolo circuit with a PIC controlling the led of a vactrol. I havn't built it yet but I will... some day.

/Anders
Keep us informed ! :grin: Digital control, what kind of method of setting had you in mind ? Tap as well ?

Bye,

Peter
 
I want to make it easy to use, so I though about having all the controls as analog potentiometers feeding DC into the A/D inputs of a PIC. And of course tap tempo as well.

/Anders
 
I had an idea some time ago about a digitally controlled analog tremolo circuit with a PIC controlling the led of a vactrol. I havn't built it yet but I will... some day.

/Anders

Vactrols have alot of variation from part to part and do not fallow a set curve, so you would have to program the micro to the response of the individual vactrol for anything time or pitch sensitive like tap tempo. I also seem to recall hearing that vactrols response changes over time, so they may not be a good route here. They do a great job when tuning is not an issue though, like in a digitaly controled analog tremolo, probably one of my favorite parts.

'Some error', as in a distorted sine (but the very same for each cycle) or you feel the need for cycle-to-cycle variations ?

There is just something in analog LFOs, they are always ignored in the digital synths, low priority. Maybe the problem is how the VCFs and VCOs track CVs though. Even the wonderful Microwave XT sounds bad to me once the LFOs are used, but it has enough other controls to make up for that. Then again it might just be in my head, either way it really can get on my nervers.

adam
adam
 
Then again it might just be in my head, either way it really can get on my nervers.
Better safe than sorry ! Worth the additional (analog LFO) circuitry me thinks, so the analog VC LFO controlled by a micro you mentioned above.


I've also been thinking in another direction, but don't see it yet as a better alternative.
Probably something with one of those looper-boxes could function here. Or having a simple redundant pocketsampler generate the LFO-waveform, based on the right incoming MIDI-note (sampler having mapped out of different LFO-speeds - but lower bandwidth will be a problem).
Again, just thinking out loud, and not really found anything too promising in those directions.


Bye,

Peter
 
Hey,

Just sort i'd throw this in, but if your thinking of using a PIC then have alooksie at the ucapps site - mebbe these threads may spawn some possibilites:

http://www.midibox.org/forum/index.php?topic=3.0
http://www.midibox.org/forum/index.php?topic=3853.0
http://www.midibox.org/forum/index.php?topic=5080.0

Cheers,

D
 
Hi Peter,

If you decide to go for a microcontroller approach, I can give you some advice with assembly language programming if you choose to take that route.

I spent 8 months designing a MIDI foot-pedal board with expression inputs etc. , so I could give you sample pieces of code or advice.
 
Thanks guys for the info & offer. As will be clear I'm a bit (quite :? ) hesitant at going the microcontroller-route, but it seems the only sane way. I'll give it more thought - it's a pity no corners can be cut when going analog but at least good to know that after having completed it in digital form or in overelaborate analog form the chances are small that it could have been way simpler.

W.r.t. simpler, I guess a simple pedal (adjusted to the song by ear during the first section) could already realize the best part of the story - but it ain't perfect of course, definitely not as nice as tap tempo could be.

TBC...

Bye,

Peter
 
[quote author="rodabod"]
I spent 8 months designing a MIDI foot-pedal board with expression inputs etc. , so I could give you sample pieces of code or advice.[/quote]
Sounds like a nice project. Did you finish it yet ?
 
[quote author="clintrubber"][quote author="rodabod"]
I spent 8 months designing a MIDI foot-pedal board with expression inputs etc. , so I could give you sample pieces of code or advice.[/quote]
Sounds like a nice project. Did you finish it yet ?[/quote]

Yeah, I finished it. It still has a small "bug" - actually a noise issue with the analogue expression input, but other than that, it worked great with a POD.

I gave it to my University "project advisor".

Assembly language programming is very easy btw.
 
I can *strongly* recommend the use of a mios device from www.µCApps.de/midibox for such an application.
It would probably only take 1x core, 1x din and 1x aout modules to have a minimum completely midi capable tap tempo enhancement (and various cv possibilities...). It can easily be built up to a full fledged steering unit with scalable LCD support, USB connectivity, BankStick EEPROM UserProgram Memory, midi updatable application OS etc...without reinventing the wheel, mainly changing just a few parameters in the mios OS and MidiCV/AnalogToolbox application header files and doing some editing to tailor the app to your needs (very easy, maybe you don't even have to 'program' a single line of asm really) and doing some testing. The most expensive part would be a good lcd with led background lighting (i prefer BLUE and ORANGE :grin:), the MAX chip for DA (if theres no alternative until now)...and your work.

I hope, this helps you to decide... :wink:

Happy DIY :thumb:

Martin
 
[quote author="smallbutfine"]I can *strongly* recommend the use of a mios device from www.µCApps.de/midibox for such an application.
It would probably only take 1x core, 1x din and 1x aout modules to have a minimum completely midi capable tap tempo enhancement (and various cv possibilities...). [/quote]

Thanks Martin, if I understand it correctly this approach might be both the best way to tackle it and wouldn't require to get my hands too dirty with nifty-languages that I don't yet understand, right ?

( BTW, for others wanting to read more as well, use http://www.ucapps.de )

The taptempo add-on I'm thinking about wouldn't need MIDI as well, but as I understand it the MIDI-functionality comes with only little bit more complexity ?

Now MIDI enters the equation, I'll also check my Beh. V-Amp box for its possibilities. It has a taptempo-button, but I don't know if it can be sent and it actually footswitch-controllable.

Well, let me first read more now at that site you mentioned.

(Hey, it's actually related to or the same as those links of daArry, right ? Two fellow-DIYers can't be that wrong :wink: )


Thanks !

Peter
 
But I'm wondering if the 'add-on' couldn't better deliver a LFO-signal by itself ? So the existing LFO is set aside and in its place this thing that could probably have a more direct translation of tap-pulses to the LFO-rate (and this perhaps even with an adjustable integer multiplication-factor).
Would this 'direct conversion' suit the microcontroller IC approach ?

Peter,
Yes. I think you are right on target with the idea of making a tap-tempo controlled digitally-generated LFO to replace the knob-controlled analog LFO. It would be easy to include a pot input as well so the LFO could be set either way. Digital LFOs can use a look-up table approach to generate sine wave, triangle wave, square wave, hyper-tirangular, semi-log - what ever you want it to.

Actually, I've thought about designing one myself - that tap tempo is a nice feature - and I'm actually kinda suprised no one seems to have made one yet, at least to sell to DIYers. I guess alot of engineering types are analog OR digital, and don't seem to hang out on the cusp between them all that much.

PIC microcontrollers come with support in the form of a free programming and debugging environment, and they make several cheap chips that would probably be ideal for implementing this kind of thing. The chips can be programmed without any fancy equipment - a simple cable plugged into your computer serial port will do it for many of them.

Or if you don't want to go there yourself, you could try and talk somebody else into doing it - like maybe RG Keen, who is a pretty creative and inventive guy. He's already bitten by the PIC bug, too.

Good Luck with your project,
:guinness:
 
[quote author="mr coffee"]

Peter,
Yes. I think you are right on target with the idea of making a tap-tempo controlled digitally-generated LFO to replace the knob-controlled analog LFO. It would be easy to include a pot input as well so the LFO could be set either way. Digital LFOs can use a look-up table approach to generate sine wave, triangle wave, square wave, hyper-tirangular, semi-log - what ever you want it to.[/quote]

Hey, nice coincidence... while typing some stuff about this, a new reply !
Let's first post my ramblings, then I'll be replying to your words.


Stubborn me giving the 'analog route' one last shot...

Thinking of a realization with a PLL & dividers, the BOSS DF-2 schematic
came to mind
(it's here: BOSS DF-2 schematics )

The idea is to lock the PLL on tapped pulses and then upon stopping tapping to
disconnect the Phase Detector of the PLL, essentially leaving the PLL-VCO controlled
by the charged but now no longer updated loopfilter-cap.

Using a freq-divider in the PLL-loop can realize all kinds of ratios between
the tapped tempo & the LFO(VCO)-frequency.
So this would run the VCO at N-times the incoming tap-tempo and that signal
could likely be converted to an usable signal for replacing an existing FX-LFO.

Component-count seems reasonable: 4046-PLL, FET or 4066-switches,
some CMOS-logic for the divider, some additional stuff for proper interfacing etc...

I don't know yet how this would work in practice - it's now used for lower
frequencies (i.s.o. using it for a gtr-signal as in the DF-2).
There might be a conflict between keeping the VCO
reasonably fixed during open-PLL-loop (requiring a larger loopfilter-cap)
and the speed at which the PLL can catch up again (smaller cap desired).
(Guess it's time to blow the dust of that PLL-theory again )

FWIW...
Bye,

Peter
 
[quote author="mr coffee"]
Yes. I think you are right on target with the idea of making a tap-tempo controlled digitally-generated LFO to replace the knob-controlled analog LFO. It would be easy to include a pot input as well so the LFO could be set either way. Digital LFOs can use a look-up table approach to generate sine wave, triangle wave, square wave, hyper-tirangular, semi-log - what ever you want it to.[/quote]
OK, here we are again. The lookuptable sounds practical - though there was the previous remark about sounding 'less right', but I haven't been able to do a side by side comparison.

Actually, I've thought about designing one myself - that tap tempo is a nice feature - and I'm actually kinda suprised no one seems to have made one yet, at least to sell to DIYers.
Indeed surprising ! At least I haven't found anything yet.

I guess alot of engineering types are analog OR digital, and don't seem to hang out on the cusp between them all that much.
True ! My bad...

PIC microcontrollers come with support in the form of a free programming and debugging environment, and they make several cheap chips that would probably be ideal for implementing this kind of thing. The chips can be programmed without any fancy equipment - a simple cable plugged into your computer serial port will do it for many of them.
I haven't really checked how much/what kind of PCBs would be required. I could imagine it wouldn't be requiring that much components but at least for the situation I an now in, an analog circuit would be allowed to be say twice the complexity of a digital one since the latter requires the additional programming learning curve. But I would learn something new along the way of course.

Or if you don't want to go there yourself, you could try and talk somebody else into doing it - like maybe RG Keen, who is a pretty creative and inventive guy. He's already bitten by the PIC bug, too.
When looking for taptempo DIY I did indeed check his site since he has quite some added-goodeis described. I haven't seen any updates for quite while now - good to hear he's still active.

Good Luck with your project,
Thanks ! As you'll have noticed I'm more testing the water for that tap-tempo functionality than that I've fired up a soldering iron or programmer yet. Nice to keep this one in my mind for a little while & see what comes around & then decide. If others have additional ideas/stuff going/etc, please let us know.

Bye/thanks,

Peter
 
Sorry to bother you again with that boring midibox stuff... :wink:
I digged again into that...building a mios based box for exactly your task is even more easy as i thought.

This thread:
http://www.midibox.org/forum/index.php?topic=5080.0
mentioned earlier by daArry, discusses the perfectly suited complete mios application programmed in C (...no asm :grin: ) containing 2x 16 Bit digital LFO's (sawtooth or programmable, example for a sinus available for demonstration...) and an envelope generator, as well as flexible digital processing - downloadable here:
http://www.ucapps.de/mios/mios_bootstrap_loader_v1_1b.zip(Bootstrap Loader, if you want to burn the PIC yourself, otherwise preprogrammed PICs are available for buying...next paragraph)
http://www.ucapps.de/mios/mios_v1_8.zip (the MidiOS)
http://www.ucapps.de/mios/analog_toolbox_v1_1b.zip(the Application)

Hardware (1x core module / 1x aout module)
All PCB's/components/kits with parts/accessories/mios pre-programmed (on request) PICs (NO PROGRAMMER NEEDED, OS and app's upload via midi sysex in-circuit/in-Box) ...can be bought at a german webshop here:
http://www.mikes-elektronikseite.de/shop_englisch/index.htm
as simple pertinax/copper version

or here at an american webshop:
http://www.avishowtech.com/mbhp/info.html
as complete solder masked fr4 version / full kits with complete pictured construction advisory, BOM, documentation links etc.

MAX525 (DA for aout module) available as sample from maxim (not that cheap...), everything else non-critical standard parts available for example from Reichelt or Conrad.

I am in no way affiliated to both, i don't now if it should be in the blackmarket section...must leave it to the webmaster...
(posted because there was a kind of "kit" discussion about this earlier...)

For recompilation / fiddling around, this software is needed:
MPLAB IDE / MPASM(even works under dosemu/Linux):
http://www.microchip.com/stellent/i...E&nodeId=1406&dDocName=en019469&part=SW007002
An appropriate Perl interpreter like
http://www.activestate.com/Products/Download/Register.plex?id=ActivePerl
GPUTILS, like here
https://sourceforge.net/project/showfiles.php?group_id=41924
SDCC 2.5.0 C-Compiler
http://sourceforge.net/project/showfiles.php?group_id=599
and an appropriate sysex upload program like
http://www.midibox.org/mios_tools/sysexbox_18f_v1_1c.zip

I just try to make a complete quickstart into this project possible as the midibox documentation is very extensive (i needed alot of time to get thru this the first time...midibox is much more mature in the meantime, very comfortable toolchain).

This application may be of use in other projects, too.

(Please don't shout me for excessive posting manners :roll: )

Happy DIY :thumb:

Martin
 

Latest posts

Back
Top