The problem is that when a sample is played there is a background noise also. This noise is still heard for about 30 seconds after the sample has finished playing. This is something I had mentioned to Jamie the creator of the unit back a year or so (or two maybe) which I did resolve at the time but clearly the resolution was not completely fixed.
Quick background on the project. This is a large project which is using the Tsunami as a sample player controlled via a MIDI keyboard. The four stereo outputs have their own VCA modules which are controlled via individual stereo LFOs. The signal then gets mixed to another stereo VCA. This is followed by a stereo VCF module, Effects unit, valve distortion etc etc. A big project.
If I have the board simply connected to my audio system there is no noise when a sample is playing. But the instance that I connect it through a VCA or any other audio device on the same power supply the noise appears during playback. This noise comes out of all the Tsunami outputs regardless of where the played sample has been routed. Which makes the noise level quite loud. If I have the Tsunami connected to its own supply there isn’t a problem. The noise is only present via other modules but it originates from the Tsunami ?
Here’s a voice sample played along with the noise until it stops.
It’s an odd one. I think I have a solution which is an isolated DC to DC converter module. I’ve not tried it as yet. It is certainly related to grounding I think. The odd thing though is that the noise disappears a few seconds after all samples have finished playing. I would have thought your “standard” ground loop would have the noise there the whole time.
That’s the giveaway that it’s not ground loops I think because the noise never changes. In fact it will multiply as you use more of the outputs from the Tsunami. No the VCAs were totally open the whole time. Still to try out the DC to DC isolator which I might do today or tomorrow.
This may not be much help, but I can at least confirm that we’ve had this issue with Tsunami as well - noise keeps playing for some time after samples were either stopped or muted (memory fails me as to which one).
Our “solution” was:
Optimize WAV file volume by multiband dynamic range compression
Maximize sample gain while avoiding clipping (we were using Tsunami-Arduino-Serial-Library for control)
Attenuate analog output using a potentiometer
While this masked the noise well enough for our application, it’s still very much there.
It was with great joy that I came across this post today.
I have been tearing my hair out with this problem of a whistle on my project.
I have been making a midi Hurdy gurdy using a Arduino Uno and a Tsunami Wav Trigger ( here is a early prototype using Wav Triggers , no noise problems … https://www.youtube.com/watch?v=Hquj4RQdFmI&t=2s )
I have since converted my prototype to using a Tsunami Wav Trigger, (no whistle, except using an on board amp) .
But my production finished gurdys have the whistle as you have described, namely a whistle at double the frequency of yours (an octave higher) at 1323Hz , this noise also continues for 45seconds after the files stop playing.
My main line of thought regarding this problem has been with ground loops, filtering the mixer (adding extra caps`) that I have on the 4 outputs of the Tsunami I am using, to no avail. What I could not understand is why my prototype gurdy has no whistle when played through a external amp, but my production gurdy had this noise. I had figured that it must be with the way I had wired all the parts together or my mixer layout was at fault, but I could not resolve the problem.
Can you please tell me is it only the Tsunami that you are voltage regulating / isolating? What input and output voltage on the regulator are you using.
Sam
amongstmyselves:
Thanks for the tips DSp. The best solution turned out to be the dc to dc isolator. Not a hint of noise out of it now. Clean as !
I only have the Tsunami on the isolated voltage regulator / converter. The unit that I bought was a Powertech brand and it was the 9 to 36v input and 5v output with a current rating of 600mA. The specific model is the PD4W30S2405D-MAD. Not a cheap unit at around A$29 but worth it for me due to the frustration it caused.
Before this solution I was contemplating putting the Tsunami on battery power as this also solved the problem.
I have solved this problem using a cheap Mosfet (IRF520) as a means to isolate the power supplys negative to the board from the audio out.
This I achieved as follows.
The Tsunami board has two ground planes, one for the digital side and one for the analogue side. The two ground planes are connected together by a GND jumper located on the back of the board above the letter U of the “sparkfun” logo, cut the jumper to separate the two planes. With the Mosfet connect pin 1 gate to the + supply and board positive input, pin 2 drain to the board negative input and pin 3 source to the negative supply.
To clarifiy my post above this solution to the noise problem is for powering a mixer or amplifier and the Tsunami board from the same power supply.
I have solved this problem using a cheap Mosfet (IRF520) as a means to isolate the power supplys negative to the board from the audio out.
This I achieved as follows.
The Tsunami board has two ground planes, one for the digital side and one for the analogue side. The two ground planes are connected together by a GND jumper located on the back of the board above the letter U of the “sparkfun” logo, cut the jumper to separate the two planes. With the Mosfet connect pin 1 gate to the + supply and board positive input, pin 2 drain to the board negative input and pin 3 source to the negative supply.
I don’t completely understand how you hooked this up. I get that you disconnected the analog section’s ground from the digital ground section, but you say “pin 2 drain to the board negative input”, do you mean to the analog section ground?
Perhaps you can provide a simple schematic, or even a picture?
So just to clarify, I am posting a picture of the Tsunami board and basic circuit of the power in and power to the board circuit.
The Audio out is plugged into my mixer / amp with using the same supply that is powering the FET.
It is the MOSFET providing the isolation of the negative supply to the Tsunami that makes this work. (Using the same trick insted on the mixer might also work ,but I have not tried this).
I have now found the easiest and cheapest solution (1 pence/ nickel) to this problem.
All you need to do is, as before:
The Tsunami board has two ground planes, one for the digital side and one for the analogue side. The two ground planes are connected together by a GND jumper located on the back of the board above the letter U of the “sparkfun” logo, cut the jumper to separate the two planes.
Wire a diode (N4007 general-purpose silicon rectifier diode) in the GND power input on the Tsunami board ( bar end of diode connect to negative supply .
The positive supply is connected as usual to the VIN power input pin.
The Audio out can now be plugged into mixer / amp using the same power supply that is powering theTsunami.
Further to my last post, I have found a issue that is not solved by my diode earth isolation fix.
My fix outlined in my last post works great for connecting a Tsunami Wav Trigger with the same supply powering a virtual earth mixer, whose output is used for passive headphones or a separately powered amplifier.
I have since tried powering a small amp and speaker module from the same supply as the mixer and Tsunami board; unfortunately this is now producing the noise problem from this unit.
I think this problem might only be cured by total isolation of the supply to the amp. I have since purchased a “Isolated Converter Power Supply Module ) , this I will try and report back on results.
Thank you for your suggestion of the " audio ground loop isolator" , but I do not think this is a ground loop problem.
(I will give this a try but the small amp module I am using has a common ground / + input)
I am fairly sure this is digital noise being induced from the Tsunami into the DC supply, hence the whistle stops a short time after the wav files stop playing and then some time later starts again.
This is a design problem with the Tsunami board using it in a all in one box project.
