Plans to make a tone generator for 1hz to 24hz

danadak

Joined Mar 10, 2018
4,057
Sorry I missed the offset requirement between channels of .1 Hz.

This again is a single chip solution that will handle this. All I did was add a second channel
DDS and WaveDac8 to derive the offset freq needed.

1581386856748.png



Note just noticed the DDS's should be driven by the same clock, thats a a trivial edit to what I show above.

If you use $ 10 board that has xtal accuracy. You can choose internal only clocks which I think are +/- 2% over T and V.



Regards, Dana.
 
Last edited:

danadak

Joined Mar 10, 2018
4,057
Generating a waveform is the simple part. Amplifying it and converting it to sound is the hard part. You can buy a set of speakers that go down to 25 Hz if you have enough money, the challenge is that lower 1 to 20 Hz range. And the amplifiers that go down to DC are rare and expensive.
Actually he wants two channels, offset by .1 Hz, so not quite so simple. But with DDS
that problem goes away. And of course external xtal needed.


Regards, Dana.
 
Last edited:

LesJones

Joined Jan 8, 2017
4,174
Even though the TS seem to have lost interest in the suggestions it occurred to me that a mechanical solution might turn out to be simpler. The idea started with a large box with a diaphragm over one face. The diaphragm would be moved in and out (Like a loudspeaker cone.) with a crank on a motor. As the maximum required frequency is 24 Hz the maximum motor speed would be 1440 RPM. The motor speed could be controlled to change the frequency. I then thought of using a large capacity single cylinder motorcycle engine or combining the output of two cylinders of a 4 cylinder car engine driven by a motor. The cylinder head would be replaced with a duct (combining 2 of the 4 cylinders if it was a 4 cylinder engine.) This pulsing air could then be matched better to produce "sound" with a large version of a horn loudspeaker.

Les.
 

MisterBill2

Joined Jan 23, 2018
18,167
Even though the TS seem to have lost interest in the suggestions it occurred to me that a mechanical solution might turn out to be simpler. The idea started with a large box with a diaphragm over one face. The diaphragm would be moved in and out (Like a loudspeaker cone.) with a crank on a motor. As the maximum required frequency is 24 Hz the maximum motor speed would be 1440 RPM. The motor speed could be controlled to change the frequency. I then thought of using a large capacity single cylinder motorcycle engine or combining the output of two cylinders of a 4 cylinder car engine driven by a motor. The cylinder head would be replaced with a duct (combining 2 of the 4 cylinders if it was a 4 cylinder engine.) This pulsing air could then be matched better to produce "sound" with a large version of a horn loudspeaker.

Les.
I suggested a mechanical generator system in post #28, and added more details in post #31. and it seems that many have not realized that the real challenge will be the transducers. So it may be that the TS has indeed lost interest, or else seen that producing very low frequency audio is not a trivial thing.
I still wonder about the intended purpose of the project. It occurs to me that there are a few uses that would not benefit anyone at all!
 

danadak

Joined Mar 10, 2018
4,057
Modifying the feedback network (C6,C7), you can likely get this http://leachlegacy.ece.gatech.edu/lowtim/ amp to operate at DC.
I built the amplifier with some minor changes. I used a 10 turn pot to set the bias. DC output with no signal is a few mV.

Apex, https://www.apexanalog.com, makes some incredible and expensive Power OP amps

I don't know what you would use as a transducer?
Sounds like interesting design. Maybe add an integrator in fdbk loop so you can take out T and V effects on DC
bias ? Just a thought.


Regards, Dana.
 
Back in the '80's I thought it was an interesting design. It appeared in Audio Magazine about 1976. It;s built like an RF amplifier.
I used the alternate FB network so the response is 0.5 to 40 kHz, otherwise it goes out to 0.8 Megahertz. The 10T pot and the use of metal film/metal oxide resistors when I can. There is one carbon composition resistor.

I get fast turn off of the inputs and oututs, FET opto and a relay. The FET logarihmetly ramps up after all of the 4 main caps (9,600 uf) are charged to about 2/3 nominal (50V). If any supply drops out ( 4 of them), we go in shutdown mode. where audio in and sppekers disconnect. I switched the output transistors and they were fine. Only damage was the flameproof resistor.
That circuit needs a little more sophistication.

There is no power switch. System operated by a timer that was in radio-electronics. 20m-2 hrs in Raw timer mode. 0-20 min after audio goes away. It contains filters too, so FM hiss is counted as no audio. Sorry, no numbers or bar graph.

The amps power supply WAS a constant voltage transformer with 18A windings. I used a custom torroidal 4x35vac@3A each.

Amp is barely warm after idleing for a couple of hours.
 

sparky 1

Joined Nov 3, 2018
756
I believe you want to create an effect using audio.
Making audio electronics requires instrumentation. Fortunately some of that equipment can be done on a PC.
A few links that might be useful.
https://www.open-daq.com/posts/experimenting-with-virtual-instrumentation
https://www.phon.ucl.ac.uk/resource/sfs/rtspect/
Because the audio desired is very low in frequency band you might need to sync the tone generated to a multiplier.
For example you might generate and play an MP3 of 24 Hertz but it is seen on the audio spec analyzer at 2400 Hertz.
The 2400 allows more samples so the low audio is spread to give a more meaningful display of that range in the frequency domain.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,167
There are many function generator circuits and systems that cover this frequency range. The lower frequency range is often used for testing and adjusting servo systems of different kinds. and the amplifiers would not be so much of a challenge, although they would not be the standard stereo system types. The challenge will be in the transducers, since imparting energy to air at those frequencies is not within the capability of most audio transducers.
 

npublici

Joined Jan 20, 2016
3
I haven't made many things but want to. I want to create a tone generator capable of binaural sound in the
1 Hz - Delta | Lethargic, 2 Hz - Delta | Deep Sleep, 3 Hz - Delta | Dreamless,4 Hz - Theta | Drowsy, 6 Hz - Theta | Fantasy, 8 Hz - Alpha | Relaxed, 12 Hz - Alpha | Conscious, 16 Hz - Beta | Focussed, 24 Hz - Beta | Active range. Here's the harder part for me, the left speaker must have a slight difference. Example: Headphones - left speaker is 8hz, right speaker is 8.1hz.

What am I using this for? My own research to see if what Nikola Tesla and Albert Einstein believed really worked. Basically, this slight shift synchronizes the left and right hemisphere of the brain improving endorphin release, and creating Alpha waves better. Theta waves occur at a lower Hz (see the chart above).

If someone can help me create this I would really appreciate it. I'm strictly a beginner electronics enthusiast. At 60 years old, I thought it was time to make this. Thanks!
Mike
What sort of transducer/speaker do you plant to produce this? Most are rated 20-20k hz.
 

nsaspook

Joined Aug 27, 2009
13,080
Since you hear nothing below 20Hz then you need a transducer to shake your head at lower frequencies.
You could build something like this to move your entire body. A transducer platform for my home theater.
20030304-002.jpg



Then you need a massive amount of amplification to drive it at the infrasonic range for shaker effects.


On a nice comfortable sofa. :)
1581705038899.jpeg
 

pyroartist

Joined Oct 9, 2015
131
>This again is a single chip solution that will handle this. All I did was add a second channel
> DDS and WaveDac8 to derive the offset freq needed.

So what is the name of the "single chip" you keep mentioning? And where is the schematic for hooking it up?
All you are showing is a computer simulation of a system with several functional blocks.
 

MisterBill2

Joined Jan 23, 2018
18,167
Since you hear nothing below 20Hz then you need a transducer to shake your head at lower frequencies.
I have actually seen such a thing. It was intended for use with video games, it had a fairly powerful set of electromagnets that pulled on a heavy block that was hung by springs, and it was worn like a backpack. It used a quite powerful amplifier to drive it. The amplifier used an off-brand module that there was no replacement available for at the time. So I was unable to repair it, at least not for a cheap price.
 

danadak

Joined Mar 10, 2018
4,057
>This again is a single chip solution that will handle this. All I did was add a second channel
> DDS and WaveDac8 to derive the offset freq needed.

So what is the name of the "single chip" you keep mentioning? And where is the schematic for hooking it up?
All you are showing is a computer simulation of a system with several functional blocks.
The chip is Cypress PSOC 5LP family.

The image is of the IDE, not a sim, so what you see is the chips components dragged and dropped out
onto design canvas, wired up internally and to its pins. What you don't see is right click on component
configs its settings, and a small amount of code needed to translate the A/D reading of the pot to freq-
uency word to set the DDS's clock to WaveDAC8. Right click on WaveGDAC8 configs it, things like amplitude,
samples/wave, waveform (sine, tir-ramp, custom.....)

Remember, in PSOC a component is an onchip resource.

The project is attached. Note the code has not been written to enable the pot, if you want I can write
that to help you get started. I already wrote code to start each component (inside the chip, thats one
line of code ) to just check it out.

The setup of the ide a little more complicated because the DDS is a community component that has to be
"added" to IDE library (unlike other components already part of library). Its about 5 minutes of work max.

IDE free to download - https://www.cypress.com/products/psoc-creator-integrated-design-environment-ide

The setup is set at 1Hz out and 1.1 Hz out. Thats fixed config, eg. right clicking the components and setting
the DDS freq desired and the samples/wave for WaveDAC8. All components have API calls you use in code
to change/manage components in real time.


Regards, Dana.
 

Attachments

nsaspook

Joined Aug 27, 2009
13,080
I have actually seen such a thing. It was intended for use with video games, it had a fairly powerful set of electromagnets that pulled on a heavy block that was hung by springs, and it was worn like a backpack. It used a quite powerful amplifier to drive it. The amplifier used an off-brand module that there was no replacement available for at the time. So I was unable to repair it, at least not for a cheap price.
There is actually a large market for these types of tactile transducers.

https://www.parts-express.com/cat/bass-shakers/3524
ButtKicker BK-LFE Low Frequency Effect Transducer for Home Theaters
ButtKicker® brand BK-LFE low frequency effect audio transducers have been received with rave reviews around the world since they were introduced in late 1999, and for good reason! Instead of using the traditional voice coil design of most tactile transducers, ButtKicker transducers are built using a revolutionary, patented magnetically suspended piston design. Not only does this design create more accurate low frequency effects (drummers swear by them for monitoring), it also lasts longer!
 

danadak

Joined Mar 10, 2018
4,057
Here is a scope DSO single shot capture -

1581718130695.png

The yellow and Aqua traces the 1 Hz and 1.1 Hz outputs. The Cyan is math of 1 Hz trace - 1.1 Hz trace, just to see
there is a relationship. I cannot, on my low end DSO, do a FFT of both. But had I been a little more on the ball I could have used
PSOC mixer component,, mixed the two waveforms, then FFT, to see the .1 Hz component....lazy....

Note if you want a phase shift between channels, using same freq, thats easy to do with the DDS component as it has
a phase offset control parameter that can be set for arbitrary phase offsets.


Regards, Dana.
 
@danadak

Dana:

Very nice. I don't like the input method.

Why not 2 rotary encoders: one for 0-24 hx, increment/decrement by1 hZ per pulse and another rotary encode that increments/decrements by 0.1 Hz

It could be smart enough that when at xx.9 Hz, to increment the Hz side to 24.0 e'g 23.9

Is it easy to add a display: Say the 2x24 serial display?

First order: would be. Freq Left 10.3 Hz Freq Right: 10.4 Hz (ON/OFF/Reverse/Level dbA)

I really like your implementation.

There might make some sense to include an Arduino in the equation. It could control a digital display, do weighting.

How might the arduino communicate to the PSOC frequency, Frequency reverse, On/OFF)

Does the PSOC have encoder software that could make it easier?

You might want to drive an on/off indicator too.

So, I see a front panel with:
A 2x20 display (serial, SPI, or I2C)
Rotary encoder: Hz
Rotary encoder: 0.1 Hz That will effect the Hz Rotary encoder.
On/Off displayed on the front panel and a LED
A push button for ON/OFF with the indicator
Power on/off (indicator optional)
Could have another rotary encoder that does functions
On/off/reverse/dbA/separation freq/Timer (just suggesting stuff, not underlying code)

Color could mess up study, so subject should face away.
Lighting could mess up study.

Should there be an option to set separation frequency? is it possible to make that a "hppk", so the design would allow it later?
Is loudness difference required? Does the PSOC measure the dbA level with a transducer.

hey, I like designing with the what if I had all the resources in the world, equipment, time and money, what would I like. What do I absolute have to have? What can I afford. Try to leave the "Hooks" in place to accomplish incremental changes.

You know the Arduino. I gave lots of possibilities. Can an interface be developed to communicate? Don' know if the PSOC could have a serila interface that basically uses SPI t send commands, SUCH as 'set AB 20.1, 20.3'; SET A 20.1: Set OUTPUT AB-20dbAl; SET A Off; SET B Off; A and B should be L and R. Lock at SCPI commands to see if the format works.

I actually meant that the Arduiuo gets rhe frequencys and displays something like":
DISPLAYL1 Left 20.1 Hz -2.1 dbA Right 20.2 Hz -2.1 dBA OFF"

Sorry door the mixed up stuff. I throw out ideas and expand on them even if its frivolis. The method works.

I expanded the capability if you let the Arduino handle the display. Providing there is a way to communicate.
I brought up possible problems.

Fundamental and offset might be a better way to communicate.
 

danadak

Joined Mar 10, 2018
4,057
Why not 2 rotary encoders: one for 0-24 hx, increment/decrement by1 hZ per pulse and another rotary encode that increments/decrements by 0.1 Hz

It could be smart enough that when at xx.9 Hz, to increment the Hz side to 24.0 e'g 23.9

Is it easy to add a display: Say the 2x24 serial display?

Yes PSOC has a encoder component in its library. trivial to add LCD, using LCD component.

1581721988636.png

There might make some sense to include an Arduino in the equation. It could control a digital display, do weighting.

How might the arduino communicate to the PSOC frequency, Frequency reverse, On/OFF)
Not sure Arduino needed ? PSOC can handle the COMM and calculation, its an ARM M3 core.

Should there be an option to set separation frequency? is it possible to make that a "hppk", so the design would allow it later?
Is loudness difference required? Does the PSOC measure the dbA level with a transducer.
Not an issue to control separation, thats just an API call (with calculations to generate the DDS clock freq).
"hppk", not sure what this is ?

Yes you could measure loudness using SAR or DelSig that is onboard..


Regards, Dana.
 
Last edited:

itkruglak

Joined Feb 14, 2020
3
The Monroe institute has been using the Binaural Beat system for years. See https://en.wikipedia.org/wiki/Hemi-Sync and https://hemi-sync.com/research-papers/the-hemi-sync-process/ . They make CDs offering music combined with low-level BB for meditation, pain relief, etc.

The easiest way to generate these frequencies is with an inexpensive dual channel function generator, such as an FY-6800, $101 on Amazon, or a Koolertron for $93: https://smile.amazon.com/Koolertron...800+function+generator&qid=1581720953&sr=8-10

This used to be the only way to generate stable, low distortion low frequency sine waves pre-1950-ish. You simply beat a variable frequency oscillator against a fixed one. Let's use 1000 Hz as the fixed oscillator and tune the variable one to 1001Hz and filter out the sum and carrier frequencies and leave the difference frequency (1Hz). I owned a 60's Bruel & Kjar audio oscillator that worked this way. The carriers were of course higher, perhaps in the 1 MHz range or higher.

You can use the 1000Hz into one ear and the 1001Hz into the other. (with suitable amplification and headphones), and your brain will synthesize the difference frequency. Keep the levels low.

Good Luck!
 
Top