A Tap Tempo Control Chip

[smallbearelec]

I thought that this product would open possibilities for many builders, especially given the increased interest in DIY delays, both digital and bucket-brigade:

http://www.smallbearelec.com/Detail.bok?no=795

[Lurco]

Springbreak
and new flowers will bloom soon

great find!

[R.G.]

12C508 PIC for the tap tempo.   

[ianmgull]

This seems like it could get interesting.

OK, I'm going to go out on a limb here...

Most delay circuits that I know of set their delay time by means of a resistance. How do you go about turning this 25mA pulse into a resistance for say a PT2399? or am I in way over my head (I don't yet know how to program a microcontroller).

Ian

[aziltz]

This seems like it could get interesting.

OK, I'm going to go out on a limb here...

Most delay circuits that I know of set their delay time by means of a resistance. How do you go about turning this 25mA pulse into a resistance for say a PT2399? or am I in way over my head (I don't yet know how to program a microcontroller).

Ian

good question.  can that be done?

[trendyironicname]

This seems like it could get interesting.

OK, I'm going to go out on a limb here...

Most delay circuits that I know of set their delay time by means of a resistance. How do you go about turning this 25mA pulse into a resistance for say a PT2399? or am I in way over my head (I don't yet know how to program a microcontroller).

Ian

good question.  can that be done?

I've been playing around with this stuff for a little bit now.  If you vary your pulse width and run it through a lowpass, you can vary your current.  Run that through to an led shining on a LDR and you have your differing control voltages. 

Might not have all my ducks in a row in explaining but having fun with it right now.

[trendyironicname]

there's probably better ways of doing it though.  I just use what I have on hand..

[Andi]

I've been playing around with this stuff for a little bit now.  If you vary your pulse width and run it through a lowpass, you can vary your current.  Run that through to an led shining on a LDR and you have your differing control voltages. 

Can that be accurately calibrated?

[The Tone God]

Interesting find and will be of use for those who want things like sequencers and chopper pedals to work in time but from what I understand its not a drop in solution for analog and digital delays. There still need to be something in between to translate those taps to an appropriate output for the delay.

Andrew

[R.G.]

Interesting find and will be of use for those who want things like sequencers and chopper pedals to work in time but from what I understand its not a drop in solution for analog and digital delays. There still need to be something in between to translate those taps to an appropriate output for the delay.
Andrew

It has never been hard to make a microcontroller sense a switch closing at intervals and output a persistent pulse train at intervals approximating that pulse timing.

As Andrew notes, that ain't what drives a delay. A delay is driven by a quite-high-frequency clock pulse. What you're trying to do with a tap tempo on a delay chip is to sense the time between taps then *compute how many delay clocks to put in* to make the total delay come out to be the time between taps. This is complicated by the fact that the RC oscillators which generate clocks for delay chips are not too stable, and their timing is often dependent on a nonlinear function of the resistance. Then you have the cap tolerance to factor in. On top of that, if you "simply" convert to a resistance, you add any nonlinearities of the conversion to resistance to the errors in timing. When you get left, how close is the delay time to what you tapped? Sometimes it'll be fine. Sometimes, not so fine.

It's not like no one ever programmed a PIC to sense a tapped tempo before...    I'll bet that some people have even run into some of these problems on their breadboards before. 

Not that it can't be done. Just be careful what you think "tap tempo" means, and what pedals you think it applies to.

[trendyironicname]

Ah, didn't think past the pulse to voltage thing when replying.  R.G. and Andrew are some of the main people I get my real knowledge from.  I go to class for hours a day to get the EE degree but, honestly, my working understanding of stuff comes from places like here. 

[aziltz]

smallbear did throw it out there mentioning tap tempo for delay.  will this not make sense at all for that?

[The Tone God]

I'll chip in some more info for the PT239x based crowd.

Unlike clock based delays the PT239x do not have an external clock input. They have an internal clock whose frequency adjusted by an external resistance. The accuracy of the internal clock is not know as Princeton does not publish that information from what I have seen.

Another issue is the relationship of resistance to time for the clock is not very well known. Even with what information that has been published it results in an odd shaped log-ish curve. The idea of using this IC or another clock source with a linear time to output relationship will not work. Neither will using a optocoupler of some sort as it will never match this odd shape curve.

Digital delays do not suffer from these issues as all the math is done in one common part sharing the same clock source so they can be accurate. With analog and multi-IC digital delays this is not the case so there is room for accuracy problems. One of the questions that needs to be asked is how accurate do you want to be ?

If you need absolute accuracy then you will be hard pressed to develop something or atleast something within a reasonable amount of resources. If you are willing tolerate alittle bit of inaccuracy then you can get something working within reasonable resources.

So after all that can a reasonably accurate simple tap tempo be made for the PT239x ? YES! In fact someone may already have done that.

Andrew

[R.G.]

So after all that can a reasonably accurate simple tap tempo be made for the PT239x ? YES! In fact someone may already have done that.

That's true. In fact, convergent evolution being the illegitimate stepchild of Mother Nature that it is, perhaps more than one person may have. 

[ianmgull]

All delay issues aside the "Tap Tempo Volume Sequencer" in the Application Notes seems interesting (and functionally similar to a certain boutique tremolo on the market )

My question is what is the deal with the input/output section? Maybe I'm reading the schem wrong, but are the In/Out jacks connected to the photo resistor portion of the optoisolator? 

[nelson]

Reading the datasheet, it seems its only standalone use is for clocking circuits at the frequency of the taps, or by however many times the control pins allow you to scale the frequency.

To get a tap tempo analog delay you'd have to convert that frequency to a 50/50 duty cycle square wave, at the clocking frequency that corresponds to the delay time of the taps.

I suppose you could do this with a PLL, a 4013 and a CD4049. However, it's a bit much circuitry to support what is already a pre programmed PIC/AVR.

You'd be much better doing all the necessary logic in the code itself.

On the bright side, it's a drop in for tap tempo sequencers.

[MoltenVoltage]

We are happy to see all the interest in our Tap Tempo Control Chip.

To answer some questions, the the signal from the 4017 goes to the LED side of the optoisolator in the App Note schematic.

Regarding a 50/50 duty cycle, MV-52 simultaneously outputs a pulse from pin 3 and a 50/50 duty cycle square wave from pin 2.  These are both scalable "on the fly" by grounding the various control pins.  See the datasheet for more info:
http://www.smallbearelec.com/Projects/MV-52Data.pdf

We have heavily researched the PT2399 issue and have plotted the variable resistances and agree with what is already in this thread, that the resistances are not linear and would therefore require an extensive table which is not possible with an 8 pin microcontroller.  That combined with the inherent inconsistencies between chips makes using MV-52 with the PT2399 impractical.

MV-52 was designed to provide a compact and highly accurate Tap Tempo controller for circuits that rely on a pulse or square wave trigger.  It has proven accuracy of greater than 1/1000th of a second.

Please post any other questions in this thread and we will do our best to answer them.  Thanks!

~ Molten Voltage ~ visionary effects ~

[slacker]

According to Sebastian (stm), who normally knows his stuff, in this post the delay time to resistance ratio is pretty linear for the PT2399. Like he says though it's probably not consistent enough to be able to make a reliable tap tempo.

[StephenGiles]

Look, you can't handle a sample and hold Adaptive Sweep Generator, and yet you think you can fart about with tap tempo

[Mark Hammer]

Is there any reason why a tap tempo control chip cannot be adapted to controlling a parameter other than delay time or tremolo speed? For instance, could one use a momentary off to the side of a 1590BB to allow one to "tap" control the centre-frequency of a resonant section, ore the drive intensity, or some other parameter, using something analogous to the old MXR PWM trick?