Convert an audio signal into digital data

[RohanDey]

Hi,

I am working on a digital guitar wireless unit project.

So in the first phase I will need to convert my guitar signal into digital data which can be then transmitted over any wireless system, i.e. wifi, bluetooth etc.

Later on I will make the receiver part where I will need to convert the digital data into an analog signal which can be fed to an amp.

I have very limited knowledge on electronics which deals with audio signals. So I was looking for some help regarding what all should I learn before starting to work. For eg, how do I choose the optimum ADC and DAC, do I need any pre-amp before my ADC, what are the major issues which I should take into account from the very beginning etc.

Thanks for reading.
Any suggestion is greatly appreciated.

thanks..

 

[chuckey]

You can get free downloads of software to convert the audio going into a PC to some recogniseable format which then can be sent out via wifi.
To make a digital wireless link between your guitar and a receiver can be done but if you want a good signal to noise ratio is complicated. I think you need at least a 10 bit ADC. You need an upper frequency response of 20 KHZ. You need a pre-amp to get from 10mV (magnetic pick up) to 5V input range of ADC. The problem here is that you have exceeded the bandwidth allowed on the most popular unlicensed frequency (433.92 MHZ).
I think you will have to do some research into methods used by professionals.
Frank

 

[RohanDey]

Thanks for your reply.

I am not not looking for a solution using a PC.

I will use a PIC micro-controller to handle the data processing. For now I am looking for some expert advise on ADC + DAC selection which can be used with audio signals.

thanks..

 

[darkgnat]

You need to figure out how much resolution you want in your application. If you just want a rough (10-12bit) approximation of the input signal you could look at the pic. You did really give us much of a description of what you're trying to do. You have described a system level, but we need to know what quality do you want. Do you want 22kHz or 44 kHz sampling? how much noise are you trying to limit to in the system, etc. Is this just a hobby project and you don't care that your first pass isn't optimal? You need to pic out a mic, a pre-amp possibly, etc. That's a lot of work you're asking us to do, when there are little projects to be found on google that will get you the analog portion. Then when you have a decent analog signal you can feed it into a pic. If you aren't really fixated on the pic, you might want to look at a raspberry pi, as it has an analog feed in, and you can use higher level programming resources. Again we don't know what you are trying to do here? Trying to learn microcontrollers, or just trying to get your analog signal into wifi/Bluetooth/whatever.

 

[ads-ee]

As a rule of thumb, make sure you keep the latency from the mic to the amp/speaker under 10 ms total, otherwise you'll start to notice the delay.

 

[D.A.(Tony)Stewart]

A very ambitious project you chose for a person self-confessed to be very limited knowledge of electronics.

First you must compare all existing available technologies for wireless. BLuetooth is usually limited to poor audio quality telephony and higher quality vs bandwidth and range are all tradeoffs.


I suggest you read up on topics like codecs cordless phones and before you do anything, define your project scope in terms of time and mat'l budget, learning curve, quality and overall objective.

If serious, then start with a solid spec. SNR, Range, Distortion, Bandwidth, Latency etc. Loss of SIgnal behavior. Dead spot plan ( Dual diversity antenna etc )

Here is one of millions of resources. https://en.wikipedia.org/wiki/User:Vinayr_rao/Wideband_Codec#Wireless

Don't make the mistake most do and try to design something without a spec. This is the best way to learn how to design, build, succeed.

 

[RohanDey]

If serious, then start with a solid spec. SNR, Range, Distortion, Bandwidth, Latency etc. Loss of SIgnal behavior. Dead spot plan ( Dual diversity antenna etc )

Hi,

I was stuck with other work for the last 2 months so could not proceed with my project. I am still at level zero to be honest.

Meanwhile I decided on my spec somewhat, don't know how much can be achievable:
1. Wireless Range - up to 100 meters
2. Frequency - 2.4 GHz
3. System needs to operate in real time.
4. SNR > 100
5. Sampling Rate > 40kHz
6. ADC bit resolution - 24 bit
7. Wireless Speed - at least 2 Mbps (802.11b/g)

I am thinking to transmit raw PCM data without any encoding. Is this achievable without any noise?

Depending on these specs, I have chosen the following hardware:
1. ADC - PCM1803A
2. Wireless TxRx Module - MRF24WB0MA/MRF24WG0MA

I could not decide on the DAC. Can you please suggest a good quality DAC which serves my application? If it comes with a breakout board, its even better.

What I am planning is to do a basic prototype with the above hardware and check how much I can achieve. Then I will look into optimizations.

Do you think my approach is correct?

Thanks..

 

[kookooli]

Hi,

I was stuck with other work for the last 2 months so could not proceed with my project. I am still at level zero to be honest.

Meanwhile I decided on my spec somewhat, don't know how much can be achievable:
1. Wireless Range - up to 100 meters
2. Frequency - 2.4 GHz
3. System needs to operate in real time.
4. SNR > 100
5. Sampling Rate > 40kHz
6. ADC bit resolution - 24 bit
7. Wireless Speed - at least 2 Mbps (802.11b/g)

I am thinking to transmit raw PCM data without any encoding. Is this achievable without any noise?

Depending on these specs, I have chosen the following hardware:
1. ADC - PCM1803A
2. Wireless TxRx Module - MRF24WB0MA/MRF24WG0MA

I could not decide on the DAC. Can you please suggest a good quality DAC which serves my application? If it comes with a breakout board, its even better.

What I am planning is to do a basic prototype with the above hardware and check how much I can achieve. Then I will look into optimizations.

Do you think my approach is correct?

Thanks..

Hi,
It sounds to be a difficult project.
Have you decided that where are you going to save your digital data? I think you should use a SD card and then send the data from transmitter to reciever.
but from my experience I suggest using an ARM Micro for your prcessing.

 

[sonic012]

So you're going to interface a pre-built 802.11 transceiver with a PIC? And use it to transmit real time audio data across the distance of a football field? What interesting projects people get into.

I'm not sure but I think that's out of range for 802.11. You would probably need a directional antenna or something.

So you are talking about
=> guitar pickups - preamp/amp and antialias filter - ADC - encoder (if any) - wireless link - decoder - DAC - reconstruction filter- power amp - speaker.

I guess it could work. It sounds like a big project.

Why 24 bits? CD quality is 44.1kHz at 16bit resolution. And why no encoder?

For a DAC, just search on it at analog devices or TI or microsemi..

https://www.analog.com/en/products/audio-video/audio-da-converters/ad1955.html#product-overview
That one has an eval board.

I bet Microchip probably has DACs built in to some of their microcontroller chips that would do the job just fine.

Sonic012

 

[RohanDey]

So you're going to interface a pre-built 802.11 transceiver with a PIC? And use it to transmit real time audio data across the distance of a football field?

The distance is not very important now. Maybe 100 meters is unrealistic. I will be happy with 30 meters for now.

And why no encoder?

Actually I am not sure whether I need and encoder or not.
Can you please explain, what advantage I would get if I use an encoder in this particular application. Also what disadvantages are there if I don't use an encoder.

Thanks for your reply.

 

[sonic012]

The distance is not very important now. Maybe 100 meters is unrealistic. I will be happy with 30 meters for now.



Actually I am not sure whether I need and encoder or not.
Can you please explain, what advantage I would get if I use an encoder in this particular application. Also what disadvantages are there if I don't use an encoder.

Thanks for your reply.

An audio encoder. People usually think of it as data compression - an attempt to faithfully represent the original message with fewer bits per second. Normally this comes at some marginal, possibly insignificant, loss in signal quality. I am guessing from your prior posts that you would not be interested in that.

However, it goes both ways. Audio coding, or source coding in general allows a higher transmission of information per bit. In other words you can also send more information (better quality) at the same bit rate.

So totally hypothetically, lets say you sampled even faster with even more resolution than 24 bit 44KHz, and then you encode down to the same bit rate as 24bit 44KHz, it would be better quality at the same bit rate.

Slightly higher complexity to encode but its not much.

Brings me back to my original point. I think you're pushing the spec too hard on that. If this is outdoor through a loud speaker, I don't think you'll hear the difference between 24bit wave and 16 bit wave. And either way you should encode. Check this article on a comparison between MP3 320KBs and 24bit wave.

http://www.noiseaddicts.com/2010/04/sound-test-difference-between-wav-vs-mp3/

He even posts the wave files so you can test it out. I bet you couldn't tell the difference even wearing those fancy studio recording headphones. And its 25% the data.

So next question you should have is, "yeah but I need a real time codec for this?" I know they're out there. Maybe somebody else on this forum knows about that.

Another thing that's dangling around in my head, is that 802.11 is packet radio. It's latency is time varying. Now it's so fast (especially if you only have one wifi connection running on it) that I assume the latency wouldn't matter. But on that front, I guess you will need to buffer the data at the receiver before you stream it?

My sense is that the latency of all these things will be insignificant. Codec latency.. low frequency filter latency.. Wifi connection latency.. But if I was going to build that project, I would pin all that down first.