Frequency To Voltage Conversion Problems for Guitar

hey guys,

I am currently working on a final project for my university studies and I am running into a bit of trouble. I am using a 741 Op-Amp to amplify the signal from an electric guitar, which in return is fed into a LM2917 Frequency to Voltage converter (necessary for PIC 16F microprocessor). Now here is the problem, according to the LM2917 datasheet, the accuracy of the voltage output depends on the values of capacitor 1 and resistor 1. I have spent hours trying to figure it out but either I get a very inaccurate output with the voltage constantly changing or I get a precise output but which is capped at 8V with the components used on the schematic. I would greatly appreciate the help as I am getting close to complete frustration :S

The circuit is powered with 9V

The description for finding R1 and C1 is found on page 7. I understand the calculations and the process but I think I might do something else wrong that I am not aware of.

LM2917 Datasheet:https://www.electro-tech-online.com/custompdfs/2011/03/LM2907.pdf

Thank you very much for any advice in advance

Phil

What EXACTLY! are you trying to do?.

Your diagram is also incorrect, and why would you be using a 741 in 2011?.

One of our first projects involved the 741 and therefore thought I might as well stick with it. The PIC 16F628 should determine what note on a guitar is played by converting the frequency produced by each note. The aim of the project itself is to build an instructional guitar displaying the player the notes from a song using shift registers and checking whether the correct one has been played.

What is wrong with the diagram?

Phize said:

One of our first projects involved the 741 and therefore thought I might as well stick with it. The PIC 16F628 should determine what note on a guitar is played by converting the frequency produced by each note. The aim of the project itself is to build an instructional guitar displaying the player the notes from a song using shift registers and checking whether the correct one has been played.

Click to expand...

It's really a VERY, VERY poor way to try and do it, and wouldn't be very accurate.

It's also an extremely poor use of a microcontroller, which could measure the frequency directly far more accurately than the poor frequency/voltage conversion. For a FYP its really pretty poor, have your teachers accepted it? - I'm presuming you're doing some kind of Electronics degree?, but this is more 14-15 year old school stuff.

What is wrong with the diagram?

Click to expand...

At a brief glance here's no connection to the input of the opamp, and there's no power supply decoupling, it's really a pretty poor effort.

I'm an industrial design student and took the electronics course as an optional module. We learn the basic principles of these systems and have used components so far like stepper motors, ds1820, alphanumeric LCDs and iButtons. In terms of coding I have a basic-intermediate understanding of Assembly. It is not my major project but a final module assignment (sorry if that was misunderstood). I am happy to accept any suggestions of possible modifications or improvements.

If you Google "PIC guitar tuners" you will see what others have done.

It's going to be extremely difficult to decode "playing" the guitar. Most guitar tuner devices barely cope with a single long sustained note. Short notes and especially multiple notes, chords and finger noise produces an incredibly complex waveform.

I'm sorry to be the voice of doom as it sounds like a real interesting project, but it may require a DSP micro and someone with high level skills in fourier analysis and weeks to months of sepcialised code and even then will have limitations.

A small microcontroller might be good enough to detect "playing" of slow single notes, provided you keep the strings damped and finger noise to a minimum. Good luck!

Mr RB said:

It's going to be extremely difficult to decode "playing" the guitar. Most guitar tuner devices barely cope with a single long sustained note. Short notes and especially multiple notes, chords and finger noise produces an incredibly complex waveform.

I'm sorry to be the voice of doom as it sounds like a real interesting project, but it may require a DSP micro and someone with high level skills in fourier analysis and weeks to months of sepcialised code and even then will have limitations.

A small microcontroller might be good enough to detect "playing" of slow single notes, provided you keep the strings damped and finger noise to a minimum. Good luck!

Click to expand...

Practical Electronics magazine did a PIC based guitar tuner a number of years ago, it used switchable filters (controlled by the PIC) to reduce harmonics and leave purer notes.

But as you say, a DSP is really the way to go.

The answer you are looking for is that you are using the wrong IC... the LM2917 has a protection Zener diode that limits its operation according to the datasheet to 7.56V(page 3) .

If you do not want this protection you should opt for the LM2907 which does not have this added "protection" and will not limit you to 8V out.

all the best,
StudentSA

Hello,

I too am struggling with the LM2917 IC. I want to set it up for a 0 - 4Khz input with an output of 0 - 5V.
My VCC is 12V. Can someone go through a quick explanation on how to calculate R1, C1 and C2. based
on the information from the datasheet? Any help is greatly appreciated.

testleads... Please see if this previous post helps otherwise I suggest starting your own thread with your problem.

https://www.electro-tech-online.com...y-voltage-converter-non-linear-output.113363/

Regards,
StudentSA