Teeny Guitar Audio Board

UPDATE!

Sorry for the lack of updates guys. Last month has been crazy as I nearly changed jobs (and countries!). In the end, I'm staying where I am and it's back to work on the Teensy Guitar board. I've spent the last week assembling and testing a proto. Here's a summary
- power supply circuits all appear to be functional
- schematic was missing DC blocking capacitor on preamp feedback paths (oops!)
- analog AUDIO IN is passing to DRY OUT, and to AUDIO OUT via Teensy Audio Library (TAL)
- USB audio from PC passes to AUDIO OUT via TAL
- USB audio from AUDIO IN to PC via TAL

What's next?
- populate and test the MIDI circuits
- populate and test the SPI memories
- populate and test the external power supply circuit
- evaluate noise performance

proto_board.jpg
 
Last edited:
Hi,

I am working on some project, with quiet the same principles.
my pickups are piezo sensors (I don't know their exact impedance but It should be is very high)
I am working with the teensy audio board (the SGTL5000),and Teensy 3.2, My knowledge in the analog field is really weak.

I looked in your schemes and specifically in the buffer section,
I wonder why you chose these values for RA1,CA1,RA3,RA4?
And what is the idea behind the RA7,RA9,CA3?

For testing the circuits I connected 2 PIEZO pickups.
One PIEZO is connected straight to the LEFT LINE-IN.
the second PIEZO connected to a buffer same as you designed, but no RA1, and RA3=RA4 = 1M and CA1 = 3.9nanoF (I would like to filter the low frequencies [50-60Hz from the 220V..])
I see that I've got constant noises(bzzzss...) in both of the channels. the one with the Buffer has much low-frequencies noise (hmm..)
How can I reduce the noise to get more studio-like signal?
BTW my OP-AMP is feeded by the 3.3V of the Teensy...


Thanks in advance!
 
CA1 is used to AC-couple the audio signal to the op-amp. Audio signals vary around some offset other than 0V in order to avoid using negative voltages in the circuits. RA3 and RA4 form a voltage divider to bias the AC signal to a suitable offset for use with the 9V power-op amp. They also set the input impedance, which needs to be high for things like passive guitar pickups. RA1 is used to ensure voltages cannot build up on CA1 when nothing is plugged in, it acts like a drain. If voltages build up on the cap you can get a very loud pop when switching things like TRUE BYPASS pedals.

The components at the output of the preamp (RA7, RA9, CA3 etc) form a high pass filter (once again, to AC couple the audio signal from the op-amp) as well as a low-pass anti-aliasing filter for the ADC.
 
Blackaddr said:
The components at the output of the preamp (RA7, RA9, CA3 etc) form a high pass filter (once again, to AC couple the audio signal from the op-amp) as well as a low-pass anti-aliasing filter for the ADC.
Click to expand...

Thanks!
I am working with your design(except that RA3=RA4=1M), but I've got much noise from that channel, Do you have an idea why?
This is the Piezo connected straight to the LINE-IN
str.JPG
This is the Piezo connected to the buffer you designed (maybe because I built it on a matrix?)
buf.JPG

g.JPG
 
@alonper - at this point you should start a new thread to debug your project rather than hijack this one.

If you use my preamp, note also that my updated schematic in post #46 was missing a capacitor between RA16=1K and GND. The cutoff frequency for the feedback path is 1/2*pi*R*C where R = RA16 and for audio applications should be less than 20 Hz.
 
Blackaddr said:
@alonper - at this point you should start a new thread to debug your project rather than hijack this one.

If you use my preamp, note also that my updated schematic in post #46 was missing a capacitor between RA16=1K and GND. The cutoff frequency for the feedback path is 1/2*pi*R*C where R = RA16 and for audio applications should be less than 20 Hz.
Click to expand...

Alright!
Thanks a lot! sorry for hijacking..
 
Hi Blackaddr !

Very interesting project. We are considering starting a project of effect/synth pedal, with requirements quite close to what you mention in your initial post.
Would you mind if we start from your schematic ? Do you have a github project linked to it ?
 
Shogun Web said:
Hi Blackaddr !

Very interesting project. We are considering starting a project of effect/synth pedal, with requirements quite close to what you mention in your initial post.
Would you mind if we start from your schematic ? Do you have a github project linked to it ?
Click to expand...

Feel free to use the schematic, though be warned I'm currently in the process of testing a prototype and as usual there changes required! I will post an updated schematic once I have revision 2 ready. In the mean time I'm happy to discuss design details of my board in the thread.
 
Blackaddr said:
The components at the output of the preamp (RA7, RA9, CA3 etc) form a high pass filter (once again, to AC couple the audio signal from the op-amp) as well as a low-pass anti-aliasing filter for the ADC.
Click to expand...

Please, Can you explain how RA7, RA9, CA3,CA5 form a HPF and LPF? (I tried to play with the values in the 1/2piRC and I cant understand how it fits with 20-20Khz)
and why is it necessary to form a HPF if you already got one with the RA16 and the series cap[8uF for (1/2pi*1K*8u)~20hz?

and RA17/RA16 make the gain?

Thanks a lot and sorry for my newbie's questions...
 
alonper said:
Please, Can you explain how RA7, RA9, CA3,CA5 form a HPF and LPF? (I tried to play with the values in the 1/2piRC and I cant understand how it fits with 20-20Khz)
and why is it necessary to form a HPF if you already got one with the RA16 and the series cap[8uF for (1/2pi*1K*8u)~20hz?

and RA17/RA16 make the gain?

Thanks a lot and sorry for my newbie's questions...
Click to expand...

The structure for the circuit actually comes from the datasheet of the WM8731 and is designed to complement the filtering inside the codec. I've modified the values because the the 12.5K input resistance of the WM8731 could form a voltage divider with RA7 and reduce the input amplitude on the assuming the incoming signal was line-level. I have not simulated the circuit and it is quite possible it is not behaving as I expect since I haven't characterized it yet either.

The intent was RA7/CA3 form a LPF with cutoff frequency of approx 159 Khz (WM8731 datahseet originally had 130 Khz). This helps with some pre-ADC anti-aliasing. Further anti-aliasing is then applied inside the CODEC. If aliasing is still prsent, the cutoff for this filter could be reduced further.

RA9/CA5 form a high pass filter to AC couple the signal from the 4.5V bias out of the OP-AMP to the VMID bias required by the CODEC. The cutoff of the HPF is approximately 1.6HZ. This removes the DC and ensures the low bass signals around 20 Hz aren't attenuated.
 
@Blackaddr
Thank you so much for the detailed explanation.
Do you think that I must re-design the circuit to fit the Teensy Audio adapter(SGTL5000)?
(I am asking because my knowledge in analog is very basic..)
 
alonper said:
@Blackaddr
Thank you so much for the detailed explanation.
Do you think that I must re-design the circuit to fit the Teensy Audio adapter(SGTL5000)?
(I am asking because my knowledge in analog is very basic..)
Click to expand...

I'm going to say no, it should work, it's just a passive filter and coupling circuit.
 
Blackaddr said:
UPDATE!
What's next?
- populate and test the MIDI circuits
- populate and test the SPI memories
- populate and test the external power supply circuit
- evaluate noise performance
Click to expand...

MIDI Circuit - confirmed working!
SPI Memories - confirmed working!
External power - confirmed working!
Noise Performance - in progress!

After disabling the HP filter in the codec (similar problem to SGTL5000 board) the self-noise performance is very good ( > than 110db at low frequencies and > 125db at the higher frequencies).

I'm going make a convenience library for the board utilizizing components of the Audio library and put some demo code on github.

I will also start working on Rev 2 of the PCB to address some minor changes I'd like to make after characterizing.
 
*** UPDATE ***

Schematic REV 2 is complete.

Changes include:

- redesigned analog supply - Both 9V analog for guitar preamp and 3.3V analog for ADC are now both generated off of dedicated LDOs. Previously 3.3V analog was created from 3.3V digital with an inductor filter circuit.
- External supply circuit redesigned to optionally tie in with VIN pin on Teensy.
- added two GPIO headers (containing a total of 8 GPIO), one of which contains SPI0. This will permit adding stuff to your pedal like a SPI display**, stomp switch or rotary encoder circuits.
- added missing feedback filter capacitors to preaamps to ensure gain is only applied in the audio domain (DC and high frequencies beyond 200 khz are not amplified)
- added power LED and one user LED.

** Using the EXT/SPI pins for SPI operation means loosing the ability to use one of the two SPI memories.

I've updated the OP with the new schematic, but here's a link as well:

Teensy Guitar Rev2
 

Attachments

  • teensy_guitar_rev2.pdf
    247 KB · Views: 648
Rev2 PCBs have shipped from the factory. Should be building another round of protos in a couple weeks once they arrive.
 
Hi Blackaddr

Actually I would be very interested to buy one of your board (assembled would be even better).
Let me know if and how we could do it...

Ps: sorry if message was double posted due to browser problems.
 
jcugnoni said:
Hi Blackaddr

Actually I would be very interested to buy one of your board (assembled would be even better).
Let me know if and how we could do it...

Ps: sorry if message was double posted due to browser problems.
Click to expand...

I should have it up on Tindie in a week or so to ensure safe transactions. What configuration do you need? I plan to have a lower-cost base solution, then you can add more components as optional add-ons.

Base configuration would be:
- single channel in, single channel out
- no on-board connectors (you provide your own, ideal for custom enclosures)
- no SPI memory or MIDI circuits

Optional Add-ons:
- two channel input, two channel output
- MIDI circuits
- 2 MBit SPI additional RAM
- connectors (3x 1/4" stereo, 2x MIDI 5-pin DIN, 2x 40pin female header for Teensy)
 
Last edited:
Cool.

The base option is perfect for me.

I am not so much at ease with soldering SMT and multipin chips but adding bigger components if needed later will be just fine.
Let me know when it is up on Tindie.
 
I am interested in an assembled board with add ons. Looks very capable. I'll keep an eye out for when you are ready. Cheers Kb
 
Hey guys, sorry for the delay in getting the boards up on Tindie. I've had to work insane extra hours at my day job the past week and didn't manage to get the listing setup before leaving on vacation. I'll be back in a week and will be a top priority when I'm home.

Paul, I'll be sure to post on the blog projects forum! The Teensy 3.5 and 3.6 are really the perfect match for DIY guitarists looking to start experimenting with their own audio and MIDI projects. Can't say thanks enough for bringing such capable hardware to the Arduino community.
 
You must log in or register to reply here.
Back
Top