Maker Pro
Maker Pro

1V audio burst every OTHER pushbutton

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
Here is a suggested starting point for the optical sensor circuit. When light is blocked to the sensor the collector resistor raises the output HI. The sensor outputs are each fed to a Schmitt trigger to give a clean transition. The Upper Sensor transition is high-pass filtered to give ~10 μsec negative pulse to trigger the monostable which generates the speed limit pulse (pulse width is set by varying the resistor in the 555 RC time constant - not shown). If the Lower Sensor signal appears before the limit pulse terminates, then a trigger pulse is generated during the overlap interval. Note that no monostable pulse can be generated unless the lift pipe is first raised high enough to activate the Upper Sensor.
EP-69.png
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
The brass tube shown is not pre-made, but I wanted to see how difficult it would be. It took an hour in the garage shop to drill the holes and cut the slits. Essential to hold the tube tightly in a drill press vice; the tight fit wood dowel kept the tube from deforming. Used a centering drill bit to avoid drill wobble, and cut the slits with an X-acto razor saw. Still not sure of the best way to make the extensions for mounting the LED & phototransistor, but I'm thinking that enough solder can make up for a lot of imperfections. If I had a proper machine shop it would be so much easier.
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
The brass tube shown is not pre-made, but I wanted to see how difficult it would be. It took an hour in the garage shop to drill the holes and cut the slits.

You did just fine. In fact you went above and beyond the norm. Many of us will spend time spicing a circuit for a member, including me .. but I've never gone into my shop and done metal work for anyone.

Chris
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
Rather than attach smaller brass tubes to hold the LED/phototransistor I tried using a larger 11/32" tube and drilling a 9/32" hole to accept the 9/32" brass tube. Ordered some Vishay BPW77NA phototransistors, so when they get here I will cut the tubes to the proper length and prepare sleeves to hold the sensors in place.

Completed the functional schematic using two TLC556N dual precision 555 timers, and one CD4093BE quad Schmitt trigger. The first timer generates the speed limit pulse, second timer generates the adjustable delay, third timer generates the 5 msec pulse, and the fourth timer generates the 10 KHz gated burst. Two of the Schmitt triggers interface the phototransistors while the other two provide the balanced audio driver.

An alternate version using a backend from the previously posted circuit is also shown. Note the use of the CMOS version of the dual 555 timer: Texas Instruments TLC556N. The CMOS timer should eliminate the need for massive decoupling to deal with switching spikes from the standard 555 (which should not be in audio circuits anyway).

EP69_2448.jpg
 

Attachments

  • EP-69_1.jpg
    EP-69_1.jpg
    76.6 KB · Views: 77
  • 100_2452.jpg
    100_2452.jpg
    13 KB · Views: 79

devin

Jul 28, 2015
21
Joined
Jul 28, 2015
Messages
21
Wow this is incredible! So far above and beyond what I expected.

I did a bit more experimenting with my dinky mechanical version, and it seems that the greatest discrepancies in time alignment are no more than about 100ms, and with that spacing shown in your pictures, the greatest span of time between triggers top and bottom would also be about 100ms... so while it might not hurt to have .5 seconds on both, would it be more convenient to have a pot with its max at 100ms instead of 500ms?
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
Two time periods need adjustment, both relating to the velocity of descent of the damper. There is the speed limit pulse width which determines the slowest velocity to trigger the 10KHz burst, and the delay pulse width for the time between triggering the bottom sensor and the damper slamming into the strings. In the absence of performing instrumented testing with a digital storage oscilloscope, adjustment must be accomplished by what 'sounds right' over the range of how the sustain control is employed.

All I can do from here is to estimate the shortest time for the speed limit pulse. From the equations of motion for falling bodies the terminal velocity for free fall from a height is given as v(h)= √(2∙g∙h), where g= 9.8 m/sec² or g= 386 in/sec². What height does the damper fall from? Assuming it's 1½" the terminal velocity would be 34 in/sec. In order to sense that velocity with sensor separation of 3/8" the speed limit pulse would need to be (3/8)/34 = 11 msec. So the shortest speed limit pulse would be 0.01 sec, but you need to decide what the longest speed limit pulse might be necessary for a lower descent velocity controlled by the sustain pedal rather than gravity.

Similar considerations apply to the pulse width for physical impact delay following activation of the bottom sensor.
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
This is a photo of the finished sensor head. Doing it again I would avoid using cross tubes for holding the LED and phototransistor since it is difficult to accurately drill large holes through tube. Would prefer to drill into a block of aluminum (½"x1"x1") or even dense hardwood. The slider tube with the optical slits needs to be thinwall brass tubing.

Phototransistor is BPW77NA with the base lead cut off. LED is super-bright red, 12,000 mcd, 8° dispersion angle, 627 nm dominant emission wavelength. The phototransistor is most sensitive to near-infrared at 800 nm but is still 50% sensitive to the LED visible red.

Measured performance through the optical slit gives a transistor current of at least 2mA from an LED current of 10mA. Transistor current begins to fall off for Vce below 0.5V. When the slider blocks the optical slit there is no measurable transistor current, i.e. less than 0.1µA.

Wired a sensor mock-up with 10mA LED current drive, 8200Ω collector load resistor and 9V supply voltage. With a slider blocking the LED, sensor output was 9V. With the slider removed, sensor output measured 143mV - more than sufficient to drive a Schmitt trigger gate input.

Sensor-Head.jpg
 
Top