Been wondering about a way of generating a fairly steady sinewave from a tap-tempo switch ?
I was thinking some kind of anti-log current source integrating into a capacitor and that driving a LFO sinewave generator - but I wonder on the accuracy of such a scheme ...
Any ideas from the digital realm ???
If you can generate a triangle, then you can feed that to a triangle to sine converter (an overdriven CA3080 works fine, in my experience).
But how to get the triangle? (which has to be of a constant amplitude).
If you have an integrator that ramps up & down as the counter runs up to the stored count and back to zero, the amplitude will be proportional to the period. Now if you build an analog divider, and divide the triangle waveform by the peak voltage, you get the triangle corrected to a fixed level. (trust me, it isn't as complex a concept as it sounds, though it might be a bit tricky making the analog divider... you could find a hint in the AD633 data sheet, then try to hack a cheaper one up from a CA3080 multiplier.)
Quote from: Paul Perry (Frostwave)If you can generate a triangle, then you can feed that to a triangle to sine converter (an overdriven CA3080 works fine, in my experience).
But how to get the triangle? (which has to be of a constant amplitude)..)
that's a great idea, easy to choose a ramp speed within required range ...
I can see another way, a mommentarily "strt" switch setting a starting point in the ramp - and the stop switch could activate a one-shot that samples the current ramp voltage via a pair of FET samplers ... if you hook up a sine-out VCO that is linear enough in its input range, then you should be able to calibrate the ramp current and capacitor to match the delay time between strt and stp switches to the VCO sinewave period ... provided the VCO has enough kick at those low frequencies ... it's a long shot 8)
Thx Paul ...
give PICs a chance... :wink:
sigh...
(done in a John Lennon voice)
Give PIIIIIICS ... a chance
Give PIIIIIICS ... a chance
pure, you're right. I've been singing this song for long enough that I think I'll just quit.
After a long time of telling people to use PICs, I feel like I've been trying to ride a dead horse.
And I am a trained *expert* in riding dead horses. Here's a paper on the uses of dead horses from my several decades in the corporate world:
===============================================
How To Ride A Dead Horse
Lakota Sioux tribal wisdom says that when you discover you are riding a dead horse, the best strategy is to dismount. However, in business we often try other strategies with dead horses, including the following:
1. Buying a stronger whip.
2. Changing riders.
3. Changing to another dead horse.
4. Saying things like "This is the way we always have ridden this horse."
5. Appointing a committee to study the horse.
6. Arranging to visit other sites to see how they ride dead horses.
7. Increasing the standards to ride dead horses.
8. Appointing a tiger team to revive the dead horse.
9. Creating a training session to increase our riding ability.
10. Comparing the state of dead horses in today's environment.
11. Revisiting the performance requirements for horses to determine if dead horses might not be acceptable for the immediate needs after all.
12. Changing the definition of "dead" so that the horse does not qualify under it, making the horse not technically "dead".
13. Hiring contractors to ride the dead horse.
14. Harnessing several dead horses together in series (or parallel) for increased speed or power.
15. Hiring or appointing a team of expert consulting riders to see if a highly focused team approach can get the dead horse to work better.
16. Declaring that "No horse is too dead to beat."
17. Providing additional funding to increase the horse's performance.
18. Doing a CA Study to see if contractors can ride it cheaper.
19. Purchasing add on products to make dead horses run faster.
20. Declaring that the horse is actually "more cost effective" if it's dead.
21. Forming a quality circle to find uses for dead horses.
22. Revisiting the performance requirements for horses.
23. Saying that the horse was procured with cost as an independent variable.
24. Promoting the dead horse to a supervisory position.
RG, *I* know tht you should use a PIC.
I also know that most DIYers aren't programmers (hell, some of us are just learning to solder!).
I wish someone would sell pre-prog pics for diy applications!
(and, form a buisness angle, it is impossible to pirate a pic with the security code on.. I know, because I lost the code for something made here & now it has to be developed again :x :x )
QuoteRG, *I* know tht you should use a PIC.
I also know that most DIYers aren't programmers (hell, some of us are just learning to solder!).
I wish someone would sell pre-prog pics for diy applications!
Ok, Ok. Sorry. :wink:
I forget sometimes.
Some day, I'll get my list of stuff whittled down to where I can do that. The thing is, the web supports a PIC programming culture more lively than the effects culture. I learned from the web in a few months.
So let me just go back and say
There are some things that are just too clumsy to do in analog. Tap Tempo anything is one of them.
I tested out my idea today with a two-op-amp ramp-inverter oscillator instead of a sineVCO ... the two switch-over voltages in the circuit (Vmax and Vmin) are set by the startup voltage and sampled ramp voltage (buffered of course) as described above and suggested by Paul - the ramp was a simple current mirror and cap with a reset pull-down switch ... the circuit worked fine and stable in period except the tapping sets the half-period duration ... I agree pic is easier, but I like the challenge of seeing how well it can work in analog ...
Quote from: Eb7+9... the circuit worked fine and stable in period except the tapping sets the half-period duration ...
Well, you can tap twice and use a divider!! but, I guess there is an easier way... then again, it's traditional to do 4 beats into, make that TWO dividers :P
...!...
.
.
.
tuning
.
sampling accuracy
.
hold accuracy
.
switch impedance
.
hold capacitor leakage
.
temperature sensitivity
.
threshold drift
.
complexity
.
cost
.
.
.
...!...
I agree, RG... but, this is DIY!! so it means people can do 'crazy' things.
I'd look on an analog tap tempo as being a kind of 'show off your electronic chops' kind of thing! my hat is off to anyone who pulls it off.
Anyone who manufactured something like that... now THEY would be insane.
is this another crossover into the DIG/DSP-forum thread
which has been overlooked?
Paul: I`ll let you know, if it worked out... :icon_wink:
Quote from: Paul Perry (Frostwave) on February 25, 2005, 06:16:02 PM
I agree, RG... but, this is DIY!! so it means people can do 'crazy' things.
I'd look on an analog tap tempo as being a kind of 'show off your electronic chops' kind of thing! my hat is off to anyone who pulls it off.
Anyone who manufactured something like that... now THEY would be insane.
Unless it is a point-to-point, hand soldered, all tube tap-tempo lfo... you know, they sound better.
:icon_lol:
Around here, just when I think the horse is about dead, someone finds a way to revive it with a new twist.
Time to call the used animal truck. :icon_eek:
I've always wanted to ask Mr. Perry if the Fat Controller would work with Ploytech's 34one midi clock pedal. This setup would be fine in the studio but a bit large for a pedal board.
The vanishing point or something similar would take up far less space. So I'l be trying to figure out how to tap tempo the sequencer. I have some Moogerfoogers and my reasons for wanting to build a sequencer differ from what some others are trying to attempt (I think).
As far as this acting like an LFO - if the signal is going from high to low and passes through something like a lag processor it will behave like an LFO. In the case of the Vanishing Point, there may be a way to make the octocoupler respond more slowly to the changes in the pot settings; then this would be acting like an LFO or at least a slowly modulated control voltage.
Because the Moogerfoogers are voltage controlled, my plan is to run a control voltage through the sequencer's octocoupler and then into the moogerfoogers. This would allow me to modulate and synchronize multiple effects.
forgive me for not being able to participate in the conversation using the appropriate technical jargon. Learning about his stuff takes time, and I have many months to go before I get through figuring out PICs, schmitt triggers,nand, nor, etc, and I've got to post here before this thread joins the horses.
Also, If this is a dead horse that has been well worked out in other areas, can someone direct me to the most helpful resource?
Lastly, I've also wanted to know for some time - as long as the signal is analog what difference does it make if some controlling aspect is digital?
So, I've searched here for a couple of hours, and I see that a few folks have said that they've come up with some solutions and code, but I don't see any schematics for hooking up PICs.
Fat controller......reminds me of Thomas the Tank Engine books!