Possible solution to reduce the noise of the terrarium build

[purpleasteroid]

Hello,

Few weeks ago, I was able to put together a working terrarium build. I ordered two boards, the first one was a total failure, but the second one is working fine.

edit (Sorry I have hit ctrl-enter by mistake... finishing the post in a reply) edit

 

[purpleasteroid]

Once, I had a working build I began to setup my workspace for development. I created this template project for creating new FX:

GitHub - soundg33k/DaisyFxProject: Basic Daisy Seed Fx Project Project with all required libraries.

Basic Daisy Seed Fx Project Project with all required libraries. - soundg33k/DaisyFxProject

github.com

Basically, this is simple project that can be used as a template to create new FXs. It include a function to test the leds, switches, and potentiometer. After writing this I realized that another user of this group created something similar.

Once I confirmed that all of my hardware was working and that I was getting signal through the unit, I began experimenting with a simple distortion clipping algorithm. I realized that the sample values was getting were in the range in between 0.001 and -0.001 (from memory, it might not be exact, but close). I was able to have fun with it.

Then, I started experimenting with reverb. I got the code from:

GitHub - fxwiegand/terrarium-reverb: A reverb example for the terrarium from PedalPCB

A reverb example for the terrarium from PedalPCB. Contribute to fxwiegand/terrarium-reverb development by creating an account on GitHub.

github.com

I compiled it, flashed it on device and began to play with it using my guitar and amp.

Once I plugged it into my amp, I noticed that the pedal was very noisy. I was hearing a "hiss" and it sounded like if this was boss metal zone at full gain. I removed the pedal from the signal chain and the noise was gone. For sure, when I was playing, it was hard to notice the noise, so it could be used if you used with a noise gate... maybe...

However, I was not really happy with the idea of using a noise gate, so I began to see how this could be fixed.

First, I thought that there was something wrong with my build. So, I double check everything. I was not able to find anything wrong with my build. Then, I looked on this forum and I saw that some people complained about 1kHz noise. They said that you could reduce the noise by incrementing the block size and changing the sample rate from 48kHz to 96 kHz. I tested that and I confirm that it reduces the noise. However, the noise it still present and it my opinion the noise is to high for being used as reverb or delay pedal for a clean guitar sound.

So I remembered about my distorting clipping algorithm. The signal was super weak. So, I came up with the following hypothesis: "the signal going through the daisy seed might be too weak. If this is the case, the output signal is operating at a level that is close to the noise floor of the unit".

 

[purpleasteroid]

I began to look at solutions to amplify the signal from my guitar to the daisy seed and then a way to attenuate the signal from the daisy seed.

There is currently an opamp in the circuit but it is setup as a voltage follower. I began to experiment with modifying the circuit to create an non-inverting amplifier for the input. For the output, I am using an inverting amplifier to attenuate the gain(I can invert the signal in code, I don't think it is a big deal to use an inverting op amp).

With this setup, I am able to greatly reduce the noise of the unit to a level that is hard to notice. However, I am still trying to figure out some values of capacitors and resistors... I am a total noob in when it comes to designing circuits. This is why I am writing this post.

Could someone help to design a circuit to amplify the guitar signal by around 10-15 dB and then attenuate it back by 10-15 dB?

What do people think about my theory... is the daisy seed operating at near noise floor level in the terrarium build?

 

[purpleasteroid]

So it has been an interesting journey. I learned few things about opamps. I have tried to boost the signal, but it was not transparent.

I found this post on electro-smith forum and it seems to confirm my theory about the idea that the output signal is operating at a level too close to the noise floor of the unit:

Rev5 seed producing ~13dB higher noise floor

To address your question, I feel some Daisy-History is necessary, but the short-version is that the difference in noise floor is related to the codec installed on the board. The original release of Daisy Seed (rev4) had the AK4556 codec installed. The AKM factory that manufactured that codec...

forum.electro-smith.com

We are aware of some of its deficiencies, especially when boosting guitar level signals to the ranges needed. The last of the original “Petal” breakout boards shipped with the last of the Rev4 Seeds, and we decided to discontinue it at that point because we felt it didn’t live up to the quality we like to ensure. We are currently working on a few newer pieces of hardware designed specifically for effects pedal development. At least one of these should be available in the next few months.

It seems like the last revision of the daisy seed is not good for pedal development (it has a higher noise floor), this is why they discontinued the Electrosmith Petal hardware!

I have a rev 7 daisy seed (revision number is printed on the center of the seed, just found out).

The terrarium PCB was probably not noisy when it was released back when the rev 4 daisy seed was being sold.

Hopefully, this post will shed light on why some people were having noise issue with their build: people have different revision of the daisy seed.

Oh and one thing, I do confirm that there are issues with the block size. I was getting some noise at 4kHz, 8 kHz and 12 kHz. I was running at 48 kHz with a block size of 12... 48000 / 12 = 4. Seems like I was hearing/seeing this (with the harmonics).

Setting a blocksize of 2 fixed the issue. This gives a frequency exactly at Nyquist frequency (given a sampling rate of 48000).

However, even with a block size of 2, the 4k-8k-12k noise is gone but I still have noise across all frequencies. When the pedal is in between my guitar and my amp, I can hear the noise. It is not huge but it is there and it can't be used on a pedal board (unless you don't care about degrading the signal). I did compare with other pedals I have (not analog one, but one that is digital like a Strymon).

I also tried to reproduce the terrarium schematic on my breadboard and I have the same noise problem. There is one thing that is interesting. Since it was on a breadboard, it was easy for me to connect the output of the input opamp to feed it to the input of the output opamp... In other words, I took the wire that was feeding the pin 16 (input 1) and connect it directly the wire that normally comes out of pin 18 (output 1). This way I was bypassing the daisy seed. When I bypassed the daisy seed, the sound was crystal clear, no noise!

Also I have 3 different daisy seed units and they all have the same noise issue (they are all rev7 :-( )

At this point, if Electrosmith did not release a petal v2... I will forget about my idea of amplifying the signal and attenuating it (unless someone has suggestions!) At this point, I will just give up on the terrarium and wait for a better platform (either something new from Electrosmith or something else).

TLDR, if you want to do a terrarium build and you don't have access to a rev4 Daisy Seed, you will probably have some noise issues. The build can be used for a school project or for someone that don't intend to use this on his pedal board. Maybe one thing to consider would be to add an 3PDT switch to at least be able to bypass the noise when the effect is off, maybe for some effects this could be ok (like a distortion). Currently the bypass is done in software and you can't get rid of the noise when the effect is "bypassed".

 

[purpleasteroid]

I actually found a way to save my terrarium build! One of my guitar amp has an effect loop and the effect loop is line level! I decided to bypass the opamp section (input and output) and plug the input/output directly to pin 17 and 19 of the daisy seed (input 2 and output 2).

I made some test with it and it works well since the daisy seed is operating at line level (the level it was designed for). Using the FX loop, I have no noise, the sound is crystal clear and transparent. I experimented with a reverb algorithm and clean guitar sound. Beside the noise of my single coil pickup... I could not notice anything coming from the device.

Although this might not work with every guitar amp. I have another amp and it has an FX loop as well. However, the fx loop is not at line level, so not usable with the amp.

This solution might not work for everyone. However, it will save my terrarium build from becoming a paperweight.

 

[keyth72]

I actually found a way to save my terrarium build! One of my guitar amp has an effect loop and the effect loop is line level! I decided to bypass the opamp section (input and output) and plug the input/output directly to pin 17 and 19 of the daisy seed (input 2 and output 2).

I made some test with it and it works well since the daisy seed is operating at line level (the level it was designed for). Using the FX loop, I have no noise, the sound is crystal clear and transparent. I experimented with a reverb algorithm and clean guitar sound. Beside the noise of my single coil pickup... I could not notice anything coming from the device.

Although this might not work with every guitar amp. I have another amp and it has an FX loop as well. However, the fx loop is not at line level, so not usable with the amp.

This solution might not work for everyone. However, it will save my terrarium build from becoming a paperweight.

Just to add to this, I bet a re-amp box would do the trick if you put it between the line output of the daisy and an amp.

 

[Gahoo]

Just chiming in here to say that I was able to fix the noise issue with the latest revision of the Seed by adding an RC low pass filter between the Seed's output and the input of my output buffer. I don't have a Terrarium PCB but the buffer arrangements are identical.

 

[Gordo]

Gonna check this out

 

[Aentons]

Thanks for the tip!

 

[RobinMallard]

Very interesting, so in theory this could potentially grafted in the vicinity of C2 and R9? I’m thinking to add the 10k resistor in series with the 1u cap and the 1nF cap to ground at R9 maybe?

 

[Gahoo]

Based on the Terrarium schematic, I would lift the leg of C2 that's connected directly to pin 18 of the Seed, and add your RC network there. You could probably tack the ground leg of C* (your RC network cap) to whichever end of R9 is grounded on the Terrarium PCB.

I don't have a Terrarium PCB in front of me to send photos of how to add these components but I'm sure someone here can figure out an elegant solution!

 

[RobinMallard]

Yup, that’s basically what I was thinking, I’ll have to see if it makes more sense to then put C2 and the RC filter on the pot side of the PCB since space may be tight to also fit the seed above everything.

 

[purpleasteroid]

Sadly the solution does not work (at least for me). I just tried it on my breadboard. I don't hear a difference. I am still hearing a hiss. I analyzed the signal with a spectrum analyser and the noise is there all over the place (from 20hz to 20khz). There is no obvious peak or anything. I would have been surprised to see this working since a low pass can't really help if the noise is present all over the frequency range.

This is definitely not transparent to the digital pedal I own. I compared it with a Strymon and a TC electronic pedal. The Strymon makes zero noise. The TC electronic one makes a tiny bit of noise... but nothing compared to my terrarium breadboard version.

What is easy to do with a bread board is to connect the input section directly to the output section (bypassing the daisy seed). When I bypass the daisy seed then the circuit is very transparent. It does not color to sound or add noise. I would expect that when the daisy seed is in the signal chain then it does not add noise.

Thanks for posting you suggestion and idea to this thread. It would be great if we could find a solution.

 

[Aentons]

Sadly the solution does not work (at least for me). I just tried it on my breadboard. I don't hear a difference. I am still hearing a hiss. I analyzed the signal with a spectrum analyser and the noise is there all over the place (from 20hz to 20khz). There is no obvious peak or anything. I would have been surprised to see this working since a low pass can't really help if the noise is present all over the frequency range.

This is definitely not transparent to the digital pedal I own. I compared it with a Strymon and a TC electronic pedal. The Strymon makes zero noise. The TC electronic one makes a tiny bit of noise... but nothing compared to my terrarium breadboard version.

What is easy to do with a bread board is to connect the input section directly to the output section (bypassing the daisy seed). When I bypass the daisy seed then the circuit is very transparent. It does not color to sound or add noise. I would expect that when the daisy seed is in the signal chain then it does not add noise.

Thanks for posting you suggestion and idea to this thread. It would be great if we could find a solution.

I had lots of noise on a breadboard too. I suspect the breadboard is not great at letting all the noise go to ground as efficiently as it needs to. I'm currently building the pcb with the same components to see if I get a different result. I'll try the low pass too.

 

[Gahoo]

Sadly the solution does not work (at least for me). I just tried it on my breadboard. I don't hear a difference. I am still hearing a hiss. I analyzed the signal with a spectrum analyser and the noise is there all over the place (from 20hz to 20khz). There is no obvious peak or anything. I would have been surprised to see this working since a low pass can't really help if the noise is present all over the frequency range.

This is definitely not transparent to the digital pedal I own. I compared it with a Strymon and a TC electronic pedal. The Strymon makes zero noise. The TC electronic one makes a tiny bit of noise... but nothing compared to my terrarium breadboard version.

What is easy to do with a bread board is to connect the input section directly to the output section (bypassing the daisy seed). When I bypass the daisy seed then the circuit is very transparent. It does not color to sound or add noise. I would expect that when the daisy seed is in the signal chain then it does not add noise.

Thanks for posting you suggestion and idea to this thread. It would be great if we could find a solution.

Interesting... adding the low pass filter has helped clear up the noise on two separate projects of mine that use the Seed and has worked for a couple other people's projects as well. I would suggest trying the fix on a PCB (like the Terrarium) just to rule out your breadboard being the issue.