Author |
Message |
elektrouwe
Joined: May 27, 2012 Posts: 146 Location: Germany
|
Posted: Fri Dec 13, 2019 3:25 am Post subject:
tap tempo circuit Subject description: 3 chip Lunetta metronome without microcontroller |
 |
|
in times of 50cent 8 pin microcontrollers this circuit comes really 40 years too late. But hey, if you try to repair a defective ATTINY tapper in 10 years, you will be in deep sh*t with your present win10 toolchain
This circuit works well in simulation. I designed it just for fun, but would like to hear about if someone will build it.
how it works:
the circuit has a digital ramp generator - CD4040 with R2R DAC clocked by start/stop oscillator U4,5 - and an analog ramp generator around C2.
Both oscillators are controlled by state flip flop U1,3. 2 momentary push switches TAP1,2 are used to tap in tempo.
1st state starts with hitting TAP1 :
DAC voltage restarts @ 0V and, by count-on/off control from comparator U6, then tracks the analog ramp.
BTW: ramp voltage does not rise linear (RC charge), so because DAC voltage tracks ramp, it is also nonlinear. This has no effect on tap tempo !
2nd state begins by hitting TAP2 :
4040 clock is stopped and DAC voltage freezes. This, now constant, voltage will be used as upper threshold for the analog ramp generator :
C2 is reset, whenever ramp voltage reaches the stored DAC voltage. Thus all consecutive ramps will have the exact shape (within 1 DAC step)
and timing as the first ramp between TAP1 and TAP2 time.
Description: |
|
Filesize: |
451.61 KB |
Viewed: |
498 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |

|
|
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1308 Location: Netherlands
Audio files: 30
|
Posted: Fri Dec 13, 2019 11:53 am Post subject:
|
 |
|
I can’t help wondering what a 4040 will cost you in 10 years  _________________ my synth |
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5777 Location: Moon Base
Audio files: 708
|
Posted: Fri Dec 13, 2019 2:43 pm Post subject:
|
 |
|
I haven't really studied the circuit yet but that R2R is unfeasible in practise. Resistors simply won't be accurate enough
(or they get very expensive). You might get to 6 bits or even 8 but I doubt you'll be able to get much more than that.
You could use a precision DAC if you can find a suitable one.
Grumble wrote: | I can’t help wondering what a 4040 will cost you in 10 years  |
Stock up now so you can sell them in 10 years on ebay (or whatever the equivalent is at that point) to lunettaphiles
But yeah, I sometimes wonder how long all that stuff will still be available with microprocessors getting so damn cheap and
fast these days. Personally I also like the challenge of coming up with a circuit instead of writing some code, but if that
requires 20 IC's while I can do it with something like a nano the choice usually isn't very hard. Having things work
asynchronous can be an advantage though and if you need a lot of in/outputs you will need some extra chips anyway,
also the wider voltage range is very usefull. Of course it's all about demand, they can't keep factories running just because
a couple of lunatics like to listen to digital bleeps and bloops. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
 |
elektrouwe
Joined: May 27, 2012 Posts: 146 Location: Germany
|
Posted: Sat Dec 14, 2019 11:43 am Post subject:
|
 |
|
PHOBoS wrote: | ... that R2R is unfeasible in practise... |
I would agree if this would be a "linear DAC" application like audio DAC or precision voltage generation.
But in this app. DAC linearity is not so important, because the DAC is in a feedback loop with the comparator. Probably 0.1% 0805 resistors (15 cent/pcs.) are already an overkill, some DMM selected 1% Rs will do the job. But a resolution of 11 bit is required if we want a 30..250BPM range with 1BPM resolution :
between 250 and 251 BPM there is a difference of about 1ms. 30BPM equals 2s which is about 2**11 * 1ms.
re. general logic gate vs. microcontroller discussion :
you can't argue with the price of a handfull of CMOS gates vs. price of a microcontroller.
It's not about the price of the chip, it's about the price (money & lifetime) to keep the infrastructure alive.
I often changed my microcontroller infrastructure : from 6502 to 8031, then PICs, ATMEL and now STM32. In the last millenium this was really expensive and time consuming. And even in this decade I 've wasted lots of money because my old winXP hardware (programmers, dev-kits,...) will not run on my current PC without parallel & serial port and win10. And I spent thousands of hours in learning operating systems and programming languages.
Now, when I have to repair e.g. my 8752 MIDI-board from the 80ies, what shall I do ? I still have the source, written in C with asm ISRs, but unfortunately did not keep some programmed chips. So I have to trash the whole thing, because the infrastructure is gone.
So I'd rather pay 2$ for a 4040 in 10years which would be a 1000% rise, than spending weeks to migrate a current Arduino project to some future 64bit-128core-DSP-FPGA-AI-thingy, even if this chip is cheaper. |
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5777 Location: Moon Base
Audio files: 708
|
Posted: Sat Dec 14, 2019 3:47 pm Post subject:
|
 |
|
elektrouwe wrote: | I would agree if this would be a "linear DAC" application like audio DAC or precision voltage generation.
But in this app. DAC linearity is not so important, because the DAC is in a feedback loop with the comparator. |
Ah I see what you mean, yeah than it is a bit less important. Although I don't understand the circuit 100% yet (might have to do with M1)
I do get the gist of it. Nice idea to do it this way. I have thought about a tap circuit before but hadn't thought about doing it somewhat analog.
edit: actually I might get it now. I was looking at U7 as being an integrator (because of C3) producing the ramp voltage but it isn't is it ? _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube Last edited by PHOBoS on Sun Dec 15, 2019 10:07 am; edited 1 time in total |
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1308 Location: Netherlands
Audio files: 30
|
Posted: Sun Dec 15, 2019 4:02 am Post subject:
|
 |
|
To say something in favor of the arduino knockoff boards: I made a tap-tempi module myself, that can simultaneously output a 1x, 2x, 3x and 4x HIGHER tempo as the one tapped or clocked as well as a switchable output to choose one ofthe tempi
Try that with some 4000 cmos chips  _________________ my synth |
|
Back to top
|
|
 |
elektrouwe
Joined: May 27, 2012 Posts: 146 Location: Germany
|
Posted: Sun Dec 15, 2019 1:12 pm Post subject:
|
 |
|
PHOBoS wrote: | .. I was looking at U7 as being an integrator (because of C3) producing the ramp voltage but it isn't is it ? |
U7 is only a comparator( with a small capacitive hysteresis). It compares the frozen ramp peak value from the measurement cycle ( time between last TAP1 and TAP2) with the analog ramp (C2 charged by R3). when this old peak value is reached again, C2 is discharged by M1. cycles with same timing follow, until a new DAC value is produced by a fresh TAP1,TAP2 cycle.
A digital solution, e.g. with 2 counters and magnitude comparators is possible, but not with 3 chips |
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5777 Location: Moon Base
Audio files: 708
|
Posted: Sun Dec 15, 2019 3:09 pm Post subject:
|
 |
|
ok, yeah then I get how it works
I don't think I've seen a capacitor being used to add hysteresis before.
Counters and magnitude comparators was indeed the direction I was thinking.
For the resistors I wonder if it is worth while to use arrays of 1 value and either use 150K with 2 in parallel for 75K (probably easiest)
or 75K and 2 in series for 150K. I guess with this amount of resistors you might as well get the values seperate. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
 |
elektrouwe
Joined: May 27, 2012 Posts: 146 Location: Germany
|
Posted: Mon Dec 16, 2019 1:25 pm Post subject:
|
 |
|
PHOBoS wrote: | I don't think I've seen a capacitor being used to add hysteresis before |
I bet you have all saw core VCOs use this to generate a constant width reset pulse (to make sure C discharge is complete)
re. resistor pairs: 150k+75k or 200k+100k are standard E24 values. |
|
Back to top
|
|
 |
dk
Joined: Feb 12, 2019 Posts: 115 Location: Europe
|
Posted: Wed Jan 01, 2020 9:18 am Post subject:
|
 |
|
Cool idea! Just a thought from the user interface side of things, but would it not be more comfortable to throw a flip flop in so you could use just one button instead of 2? |
|
Back to top
|
|
 |
|