Gallery, Projects and General > Project Logs

Diy optical pickup for guitar -- is it possible?

<< < (7/26) > >>

PekkaNF:
There is always a little buffereing and level adjustment on analog/digital signals. You may want to search logic level on TTL or mathever logic circuits you have. There are some differerences wetween different logic levels voltages and sourcing or output currents. Usually schmit

I have this very generic function generator and it does pretty much what it says. Cheap, not extremely accurate and not the best quality signal (some noise and such) and pretty close to 100% that I need on general simple signals. Only thing is that the buffer does really output great signal to 50 input at more than 5 MHZ or such. Should not be problem for your use.

https://www.ebay.com/itm/24Mhz-Dual-channel-DDS-Function-Arbitrary-Waveform-Signal-Generator-Sine-Square-/181775301443?hash=item2a52a6fb43:g:RxEAAOSw~ZdVfuQg

It does not have the skleekest user interface, but it is easy to use on all normal waveforms, dutu cycle, offset voltages and such basic functions you need. It is not that great on advanced waveforms, but that a little more advertising gimmic, if I ever need that I would buy completely different "box".

There are cheaper ones for square vave dutu cycle, but they are even harder to use effectively. Least the ones I have tried.

sorveltaja:
Pekka yes I've been looking for different ways for analog to digital conversion. Well not exactly that, but more like simply chopping an analog signal wave to 0 to 5 volt square wave.

Thanks for suggesting the function generator. I checked the link, and that one, or similar should fit for the purpose.

Little bit of hands-on testing: I took one pair of sender/receiver out of the guitar, to see, how the signal(~50khz) transfers through them.

Pulsed output from the 555-based square wave generator:
 

Then it goes from sender to receiver, whose output is fed to an op-amp like this:
 

Resulting output, although not symmetrical, sits finally around the zero volt point:
 

Now, that the signal crosses zero point, it is sent to a very simple zero cross detector, which requires only one resistor to be added, that I found from the LM393's(comparator IC) datasheet.
This version didn't work at all:
 

This one gave at least some kind of output. Perhaps the right direction, but not quite there yet:
 
 

I always like to start with minimum part count, as on project like this, the number of parts are eventually multiplied by six(except the possible dual/quad IC's). But it is to be expected, that the zero crossing detector needs more parts to make it more useable.

I'm sure, that using the zero crossing detector is just one option, but as I have comparator IC's at hand, I'll give it a shot to see where it leads.

Plenty of testing ahead, as the visual results so far exclude the effects of actual, vibrating guitar string in the infrared beam.

 

sorveltaja:
At first, I'll have to admit, that there was a gross mis-reading by me, of the values showed on scope. That dawned to me, when tracking down the reason, why I can't get that zero cross detector to work, as expected.

After digging on the net for alternative ways to amplify the signal from the receiver, I found out, that photodiodes and 'transimpedance' amplifiers go rather often hand-in-hand.

Without knowing much about it, or even having an actual transimpedance op-amp chip, I looked for the configuration.

Bit of testing with one half of the very basic TL062 op-amp:
 

Output was fed to one of 'schmitt triggers', which is 1/4 of the 4093 IC, that I have at hand:
 

The inputs were simply tied together:
 

Yellow curve is from the op-amp output, while blue one is from the 4093 chip:
 

If all goes well, there could be a possibility to move on with the project.

sorveltaja:
Not much to report, just some testings to confirm the previously noticed thing, that the photodiodes I have, seem to have a very narrow operating range, what comes to the light intensity of the sender(IR-led).

Applying more and more current for the sending ir-led doesn't help, although one could expect, that the receiving photodiode would take all the 'juice' gratefully, burp, and demand more.

But after all, there is at least one solution, that I noticed after testing: by using the transimpedance(ish) configuration:
 

Replace the 1M resistor with 1M trimmer potentiometer. That way the narrow operating window of the photodiode can be altered to suit the following stage, which is schmitt trigger.


   

sorveltaja:
I did some testings to see, how the vibrating string affects the pulsed infrared beam between sender and receiver. Had to use the thickest string on the guitar for that, as the effect is quite subtle, when looked at the scope.

Here the string is plucked, and its vibration appears only at the trailing edge of the pulsed square wave:
 

The string signal seems to be more audible, than visible. It has some noise also, but not as much, as the earlier used phototransistors had.

I used again laptops mic input as an audio probe. It adds some 'weight' for the signal seen on the scope by altering it. Perhaps it could be better to use them separately, one at a time, to find out the best output ranges, either visibly or audibly.

The ~100kHz 'carrier' frequency should be sifted out at some stage, I guess. Though I'm not sure, what sort of modulation it has with plucked string signal.

After all, this projects' goal is not to produce a sweet, clean guitar tone. More like to squeeze the outputs to a form of a hex fuzz, that could possibly be used for a pitch tracking purposes also.

Navigation

[0] Message Index

[#] Next page

[*] Previous page