1V audio burst every OTHER pushbutton

[Les Jones]

The use of the phantom power supply greatly restricts you options. The hall effect device and the photo interrupter are out as they take more current than the phantom supply is capable of. A magnet operated reed switch seems to be the only solution in combination with dougy83's CMOS tone generator. I don't think Tony Stuart's idea of using an electret microphone to detect the thump when the pedal is released would work as it is this thump we are trying to suppress. (Although I think it could be made to work within the current limitation of the phantom power supply.) Just as an aside I knocked up a tone burst generator (Without the 0 - 500 mS delay) using an ATtiny13 and it works. It times the 50 mS burst by counting 500 cycles of the 10 Khz tone. The use of this idea would probably be excluded due to the limitations of the phantom supply.

Les.

 

[devin sheets]

What if it was a battery? Can you draw up a circuit that runs on AA? Or would it need to be bigger, like a 9V?

 

[Les Jones]

The CD4093 that dougy83 uses needs 5 volts. An opto interrupter would need at least 2.5 volts to drive the LED and allow for a resistor to control the LED current. The current taken by the LED would be about 20 mA The ATtiny13 tone burst generator I built takes about 2 mA at 5 volts but this device should work on about 3 volts and take a bit less current. (I have only tried it on 5 volts.) Could you not mount all the tone burst generators in one box using a mains power supply ?

Les.

 

[devin sheets]

They can't all be in the same location unfortunately. Each piano needs its own unit... I figured that the 48V Phantom Power would act like a central power supply... but if it won't work, then batteries are fine, just less convenient.

 

[Tony Stewart]

Have you discussed, if this design is to be a portable design, or permament?
Is there any room for a silent Beryllium Copper spring contact switch?
How do you intended to mount it , or what room is there for mounting?
Below the pedal only? or inside the piano?
Is there any possibility that a soft release is to be ignored, for soft music so you doesn't modulate the amplitude incorrectly?

 

[Les Jones]

You only need one photo interrupter. The light path is only interrupted when the pedal is in the rest position and the first mm of it's travel when pressed. When the pedal is pressed the output of the opto interrupter goes from low to high. It will stay high for all of the travel down to the limit of travel and while it is travelling back to the rest position. When it is a few mm from the rest the light will be again interrupted causing the output to go back to a low level. It is this transition from high to low that triggers the tone burst.
**broken link removed**
The top picture is how the opto interrupter is connected to the tone burst generator. If used with dougy83's circuit both the interrupter and tone burst generator can can be powered from a 9 volt battery. The lower picture is the tone burst generator using an ATtiny13 micro controller. The anti phase square wave outputs come from pins 2 and 3 of the ATtiny13.

Les.

 

[cowboybob]

Also, there is only one oscillator (maybe two now) but I'd need 10 of them, or a way to change the patching accordingly every time someone plays a pedal...

Are we to understand up to 10 pianos will be used?

This level of complication cries out for a uC solution, which would also eliminate the problems of a single, non-latching Hall Effect switch.

 

[devin sheets]

You guys are so far above my head right now haha, I'll have to wait until the consensus is in, then maybe someone will be kind enough to scribble out the entire circuit for me

 

[Tony Stewart]

no devin, you need to be more explicit about your application and interface.
the best solutions depend greatly on the fine details and your answers to questions I suggested in #65

the lack of interface requirements detail is why it is taking so long to solve.
Please expand interface restrictions, possibilities with details and qty needed.

The better the interface specs, and your capabilities, the easier it is to resolve. It you need every last detail on how to build it, you could be in trouble. All of these ideas can work, but you dont see it that way yet nor do we for different reasons.

 

[devin sheets]

what can I answer for you?

 

[devin sheets]

I don't know what I don't know here, so it is hard for me to give details unless I'm answering a specific question...

 

[Les Jones]

I will describe my understanding of the problem/solution. If my understanding is wrong devin can correct me.

There will be one of these units installed in each piano It is required to generate a 50 mS long burst of 10 Khz just before the pedal reaches its rest position when it is released. The foot taken off the pedal. The purpose of the unit is to mute the microphone input just before the thump which occurs when the pedal is fully released

The 10 Khz signal is to be injected into the microphone cable at the piano end of the microphone cable. (Dynamic microphone with balanced output. Fairly low output impedance.) The 10 Khz signal will be a high level which will cause the compressor in the mixing desk to reduce the gain on that channel thus reducing the amplitude of the thump. The 10 Khz will be filtered out by a low pass filter at the mixing desk. (I assume there must be something like a graphic equalizer on each input channel.) It has just occurred to me that the compressor must come before the graphic equalizer for this scheme to work.

The OP has no electronic knowledge at component level but does understand at the block diagram level.

I think the problem needs to be looked at in three parts. Trigger mechanism. Tone burst generation. Powering the unit.

Trigger mechanism.
Possible solutions.
1 Micro switch. Problems. Too noisy. contact bounce. Advantages. Does not consume any power.

2 Reed switch and magnet (Alarm contact set.) Problems position of operation not very precise, contact bounce. Advantages. Does not consume any power. Tolerant of contamination.

3 Home made contacts made from beryllium copper finger stock. Problems may become oxidized or contaminated with dirt. Possible contact bounce. Advantages. Does not consume any power.

4 Hall effect sensor and magnet. Problems. Consumes more power than available from phantom supply. Possibly position of operation not very precise. Advantages Tolerant of contamination.

5 Opto interrupter. Problems. Consumes more power than available from phantom supply. Could be contaminated with dirt. Advantages. Fairly precise operating point.

Tone burst generation.
Possible solutions.
1 CMOS gates Problems. Each unit would have to be individually adjusted after building. Advantages. Cheap. Very low power consumption. Fairly wide operating voltage range. (5 - 15 volts.)

2 NE555 / NE556 Problems. Each unit would have to be individually adjusted after building. High power consumption (Unless CMOS version used.) Advantages. Cheap. Fairly wide operating voltage range. (5 - 15 volts.)

3 Micro controller using internal oscillator. (EG ATtiny13, PIC12F629) Problems May need Oscillator calibration adjusting after building. OP would need to buy programming hardware.
Fairly limited operating voltage range. May need a voltage regulator. Advantages. Very low component count.

4 Micro controller using external crystal . (EG ATtiny2313, PIC16F628) Problems. OP would need to buy programming hardware. Fairly limited operating voltage range. May need a voltage regulator. Advantages. Fairly low component count. No adjustment needed after building.

Power supply.
1 Batteries Problems. Need to be checked/replaced before use.

2 Phantom power supply. Problems Very limited current available which would greatly limit which trigger and tone generation solutions could be used.

Les.

 

[cowboybob]

Just to muddy the waters:

devin, are these sustain/mute "thumps" at all consistent, one piano to another? Whatever the sound signature, I wouldn't think that this particular "noise" is anything like the sounds a piano can make or might be expected to make. And are the microphones picking up the thumps sonically or mechanically connected to the pianos?

I know nothing of the capabilities of a profession mixer. But if these thumps could be characterized (see 1. below), perhaps they could be specifically filtered out, either by the mixer or custom, passive, adjustable filters (as opposed to the momentary, general reduction/muting of ALL sounds from the piano) .

1. For instance, this paper: https://digital.library.unt.edu/ark:/67531/metadc30427/. Excerpt: "... The first portion of this paper describes methods for quantifying noise content using FFT analysis procedures. ..."

 

[devin sheets]

I've tried to think up a way to characterize the sound of the thump at the waveform level, but it is too similar to various "musical" waveforms the piano makes in its lower register. The "pickup" style mic I'll be using works like a guitar pickup, and it only picking up the vibration of the metal piano string within the magnetic field, and by extension it also picks up any macro movements or jerks within the piano because the pickup is attached to the internal brace bars and its own movement would be interpreted in the inverse as the movement of the strings. So when the dampers come slamming down on all 88 strings, there is both the immediate movement of all 88 strings, and there is the overall jerk of the piano body. And, each piano's thumping sound is different, and in fact each thump on the same piano is slightly different if you really look closely at the waveforms. I could get creative maybe if I was in a studio environment with lots of time and precision, but I need a hard and fast solution for live sound where I've got 10 of these to slap on and calibrate with almost no sound check time and a thousand other things things to also worry about before the show starts.

 

[devin sheets]

Les, that's a good description, just a few clarifications:

5ms burst, and also the ability to select the timeframe from o-500ms within which the release must happen in order to trigger the burst. Also, the ability to select the overall delay of the trigger from 0-500ms and the volume output with a max of 1.4V rms. The 10kHz burst need not be precisely 10kHz; anything between 5000-15000 will work. If the summing of the two balanced signals into one is too much of a hassle, I can just leave them as two separate signals and use two inputs on the mixer.

Are you sure the Phantom Power isn't enough to run the optical option?

 

[Les Jones]

I don't know what you mean by clarification of the 5 mS burst. dougy83's design already has the 0 to 500 mS delay feature. This could easily be added to the ATtiny13 design by connecting a potentiometer to the ADC input. the 10 bit value read by the ADC would set the delay. I am not clear why you would need a delay as much as 500 mS. Does this mean that the damper bar may not come down for up to 500 mS after the pedal has reached the fully up position. I would have thought that the bar came down within a few mS of the pedal reaching the fully up position I found the information on the phantom while trying to find out which pin on the microphone cable carried the 48 volt phantom supply. I found that there was no wire with just 48 volts on it. The 48 volts is injected into each wire of the balance audio signal via a 6.8 K resistor. To use this supply you would have to tap it off the audio pair with a pair of equal value resistors of a value that would not load the audio signal excessively. See the link to information on phantom power supplies I put in post #52 Also look at this link **broken link removed**
If you could find a photo interrupter that would work with very low LED current (Less than 5 mA) it might just be possible to get it to work.

EDIT. Was the confusion about the 5 mS burst because I was talking about a 50 mS burst in message #61 ? I had remembered the time incorrectly and wrote the code for 50 mS To chnge it to 5 mS all I have to do is count 50 cycles of the 10 Khz waveform instead of 500. This just means changing a number in one line of code.

Les.

 

[devin sheets]

I guess the reason I had 500ms of overall possible delay was because I really don't know how much latency will be introduced between the pickup itself and the processing on the mixer and also the bit of time between when the burst is triggered and when the dampers actually create the thump. The timeframe delay of up to 500ms is because I don't know yet how slow one must play the pedal on each piano in order to avoid a thump. I figured that would be more than enough of a max delay for both variables? Is it problematic to have them this large, or is it somehow easier to set them lower, such as 100ms?

 

[devin sheets]

It really is only a convenience factor to run the thing on the phantom power... I can use batteries if necessary, but that's just a hassle and a lot of extra expense.

Also, it really isn't necessary to have a summing of the two signals before the input at the mixer, I can live with two separate signals.

 

[devin sheets]

Whatever design we come up with, it will have to be simple enough that my electronics friends can understand it. I'm counting on them to guide me in buying parts and then helping me assemble ten of these things... my guess is that they aren't quite as knowledgeable as you all are, but they can certainly follow diagrams and instructions.

 

[Les Jones]

You are still not giving much information. For example you have not said exactly what you want clarifying about the 5 mS burst. You need to give some details about the capability of the person that is going to build this unit. (Such as is he familiar with micro controllers and if so what family. Also does he have the hardware and software to program them.) We need to give you a design of tone burst generator that you can get built. Also which trigger mechanism you plan to use for the first test. (If you plan to use a mechanical switch the we need to consider filtering to deal with any contact bounce.) If I was doing this I would try the reed switch (Alarm contacts) approach first. You also need to tell us how much trouble we need to go to to convert the square wave to something like a sine wave. Bear in mind that a square wave is made up of the fundamental plus all the odd harmonics. the amplitude of the harmonics decrease as the order increases. So the harmonics that MAY cause you a problem would be 30 Khz, 50 Khz, 70 Khz, 90 Khz. We also need to know the impedance of the microphone signals. (This will probably be mostly influenced mostly by the microphone rather than the mixer input impedance.) If I was doing this I would make decisions when testing. I would start just inserting the square wave via a resistor. If that gave problems then add a low pass filter. If it was still not good enough I would add a 10 Khz tuned circuit. (Capacitor and inductor) The 0 - 500 mS delay is not a problem but I don't think it will be needed. You will probably need the person that is building the units to go to site with you to make changes that may be required while testing.
Another approach to this problem would be to gate or attenuate the microphone signal in the unit even before it reaches the mixer. If gating would be acceptable the a quad bilateral switch such as a CD4066 would probably do. If an electronic attenuator was required (To give controlled attack and decay.) then I would suggest copying the compressor circuit in the mixer.

Les.