Tap tempo in a Rebote

[taylorchuck15]

I know how everybody who builds pedals builds the rebote 2.5 sometime in their lives. I would, but if I was to have a delay, I would want it to be a tap tempo delay like the Boss models. I saw somebody on here who put a tap tempo in a tremelo. Is there a way to put a tap tempo in the same way as the tremelo or is it way over the DIY line?

[blanik]

this as been talked about somewhere... i remember it was quite complicated for padawans (like me) 
do  a search with the subject title, it should work...

R.

[jlullo]

TC,
Hey there... it is possible....

http://www.diystompboxes.com/smfforum/index.php?topic=50185.0

there is also a link to andre's tempo indicator LED design (which i am in the process of giving a shot... thanks andre!)

[taylorchuck15]

I read it, and the indicator LED is awesome, but I'm not sure if anybody figured out the tap tempo secret.

[The Tone God]

Tap tempo will pretty much need a microcontroller. I'm sure there are some threads on the topic in general so do a simple search.

Andrew

[R.G.]

The tap tempo secret is ... use a microcontroller.

Think about what you want a tap tempo to do. You want to tap a switch with your ... OK, some body part, and have the pedal adjust its modulation to suit what you tapped.

There are some interesting issues there that are not apparent.

First, your human range of tapping and the effect's range of modulation must overlap substantially or you get nothing particularly useful. In the case of the referenced thread, the max delay is about 1/3 second maximum. So a tap tempo effect, even if perfect, can't make the echo time more than that. From 3 times per second, how much faster can you tap your toe? Can you tap five times in a second? Ten?

Then there's the whole mess of what's a valid tap? How many times do you read time periods to get a tap period? Do you use one interval? Average a couple? Continuously average? When do you stop counting taps? Did the guy just quit tapping or this a verrrrrrrry long tap interval? And from the pedal's point of view, I just got two taps 57mS apart. Is that my owner having a great night and a fast toe, or did the switch bounce?

Then there's the issue of what the tempo is when the player asks it to do something it can't. Say the slowest delay is 330mS, 1/3 of a second. I get taps 1 second apart. Does the pedal do nothing, because this is an invalid circumstance, or does it attempt to quadruple the period and give 250mS delay, four per tap interval? Or three or six per tap invterval?

Hey! Here's an interesting thought. I have here a tremolo, phaser, and delay. They all have PERFECT tap tempo features. How do I synchronize them? Tapping two or three times on each one? What if the internal arithmetic is not perfect and they start wandering apart in the song? What if I WANT them to wander a bit? Can I do the tapping just enough off to get them to wander?

I think a lot of people have succumbed to "tap tempo" as a new icon of musical goodness without much thought about how it really works in a pedal and on stage in a playing situation.

[Fp-www.Tonepad.com]

very interesting, RG.

I've faced a couple of different tap tempo implementations in 'bought' gear, some take an average of several taps and some will work by using the last interval between taps, in these the tap tempo is set by just using a 'start' and a 'stop' taps, if you take the average out of the equation it seems a little simpler...

How about a circuit that reads only TWO taps and outputs a current relative to the time it took between the too taps, then you could use that current to drive an LED/LDR setting the delay time of the pt2399. The way I see it is a circuit capable of ramping the current to the LED that starts with the first tap, current increases(or decreases) constantly until the next tap (the final tap) and then it holds the current output constant until next cycle begins (with a new tap). I guess one could preset a max/min current value to the relative delay time available, and have it reset if the current reaches a preset ceiling.

Is this too crazy to do with opamps/ttl/clocks? some of my thoughts even went as far as using a second pt2399 to do the timing... maybe a little out of my league. But a fun problem to think about in non conventional (and non microcontroller) solutions. What do you guys say, possible?

[R.G.]

That will absolutely set some tempo. The problem is linearity, or lack of it, and open-looped-ness.

That's very similar to the way that old analog TV sets did horizontal and vertical sync. They generated a ramp of voltage that was reset by the incoming vertical and horizontal synch pulses. However, the TVs got a continuous series of sync pulses, like a click track in studio terms.

An LED puts out a light intensity that is almost but not quite linear with respect to the amount of current. An LDRs resistance is not linear with respect to the light that hits it, and it drifts in light value as it adapts to the light over a period of seconds and minutes. If you had a perfect ramp and hold circuit (which is easy) and a voltage-to-current converter to set the LED current, then the LDR would still not give you equal increments of time per change in LED current.

Then there's open-looped-ness. Even if the LDR and LED were linear, nothing particularly says that the current per ramp level will generate a time response from the delay oscillator that matched the time it took to do the ramp. Even with tweaking, to set the delay for a 1/3 second delay for a 1/3 second interval between taps, any thermal or voltage drift will drift it off timing.

The simple analog solution isn't likely to be very good, and the good analog solution is going to require a lot of stuff to get it right.
In my opinion, by the time you got anything to work, you would have a simpler, cheaper, and more practical circuit if you used a microcontroller. Even using a second pt2399 for timing is something of a loser, as a PIC is cheaper than a 2399.

It's not a question of whether it can be done analog - the radar sets of WWII were comparable - but rather what do you have to do to get there, how much time and money does it cost, and how well does it work when you get it done.

I consider it impractical, but then so is playing golf, and a lot of people like that.

[taylorchuck15]

So does anybody have a schematic or possibly a layout for how to do this?

[Fp-www.Tonepad.com]

We're theorizing, you can join in.

[slacker]

You can supposedly control the delay time using a tranistor current sink instead of an LED/LDR combo http://www.sdiy.org/destrukto/vc-echo.html. I haven't tried this but I guess it could be converted to 9 volts and it might be more accurate than an LDR. Still got the poblem of generating a voltage proprtional to the tap time though.
Does anyone actually use tap tempo delays though or are they just marketing mojo? I had a multi effects processor that had tap tempo and I always found it easier to preset the delay time and then as a band we'd sync to the delay rather than trying to tap the delay to sync in with the band. Personally I reckon an expression pedal would be more useful, at least that would let you tweak the delay time without bending down mid song or whatever.

[R.G.]

So does anybody have a schematic or possibly a layout for how to do this?

The short answer is - no. At least none have been presented publicly here.

I have programmed tap tempos with microcontrollers, and I once posted the pseudocode for programming one here at the forum. My tinkering with that is how I came up with the set of interesting but not apparent issues with tap tempo.