Non working Pedalshield

6 years 4 months ago - 6 years 4 months ago #1241 by Charlez99
This forum topic is the continuity of www.electrosmash.com/forum/hardware-peda...-u1-pin-3-1v?lang=en
but under a different name as previous topic name was wrong (the problem isn't related to U1 pin3 voltage), so more people may benefit or contribute.
I suggest you read it before continuing here.

INTRODUCTION

A bit of background : being an electrical engineer (but I know little about small electronics) and a weekend guitarist I always wanted to build my own pedals, but was always too busy to take the time to do it.
So all this started off after listening to an interview with Ray about the Pedalshield on the Programming Academy podcast. Also I decided to order one, after all that would save me time, and I had an Arduino Due doing nothing in my stuff, so why not...

My objectives were mainly to learn about digital guitar sound effect, then to add an LCD on the Pedalshield to show adjustment and pot functions, then to add simultaneous multi effect (software improvement), and finally to design an Android interface (let's say a bit like the Amplitude app) using App Inventor 2 running over Bluetooth to control the Pedalshield by adding a HC-05 module to the Due.

A quick note about my Due board; I bought it on AliExpress in November 2014, not the latest PCB version, and obviously a Chinese clone (for about 20$ instead of 40$). Also it suffered from the well documented problem that it would not restart after upload 1 time out of 4 (over 39K viewers on Arduino forum forum.arduino.cc/index.php?topic=256771.msg2512504#msg2512504) that I fixed by soldering a 10K resistor R99 as shown below.



Before I hit the meat on the subject, I want to highlight that the LED that was acting strange (described in previous tread)). As I was suspicious about DO3 pin to provide enough voltage for the LED, I change R21 from 1K to 470ohm (for less voltage drop and higher current) and install a 2 pin female header socket so I could easily replace the LED for red vs blue.

MY PROBLEM: BOARD IS VERY NOISY (COMPLETLY UNUSABLE)

As board was very noisy, I started by troubleshooting the power supply. I made trial & error tests using various capacitors while looking at the power supplies voltages on my scope. It simply end up that higher values of electrolytic capacitors where the best at reducing noise. So I added a set of 220uF 6.3V cap on +5V and -5V (C17 & C18 beside TC1044S), and also on the 3.3V. I soldered them on the back side of the PCB as space left for capacitors was only for ceramic capacitors (see picture below). But still the pedal is incredibly noisy.



Based on wrong information (as described in previous topic) my early feeling was the input op-amp stage on the pedal was defective. When looking at waveform captured by the Due transmitted over serial to the Arduino plotter, I realized the were very similar as those I was looking at on my scope (images shows a "E" string pluck after 1 second, Yellow = Input, Blue = Output). So that proves the problem wasn't on the input stage (wrong early feelings) of the pedal because the sound capture is OK (no clipping and no noise).





So the problem must be at a later stage, the noise is produced either at the DAC level or at the op-amp output stage of the pedal.
To investigate more I used the sinewave sketch (clean signal in software, noisy signal in hardware) and I compare the DAC0 & DAC1 signals of the pedal output on scope to realize the output stage is only doing what's it's suppose to do; amplifying and buffering the signal. So he problems lies in the Due board DAC output.



I made some research on this and found I wasn't alone with this problem, this is common. Lots of people are having similar issues.
See dirty_DAC0.jpg and dirty_DAC1.jpg on forum.arduino.cc/index.php?topic=194694.0

Now my latest feeling and conclusion is that my Pedalshield may never work properly because my Due board may be the culprit.
It is because it's an old revision of the board ? because it's a Chinese clone that has an ineffective noise reducing ground plane design inside the multilayer PCB ?

Just to make my mind clear, I would appreciate if someone that has a good working Pedalshield Due could post a scope screenshot of the pedal in & out signal running the clean sketch (pedal engaged).

Did I miss something? Is my conclusion OK? Did anybody encountered the same thing? Please comment.
After all that time lost, money lost, and frustration, I wonder what are my options beside to procure a new Due?

I'll most likely take a break on that project as I have a dozen other project to spend my time on, but I may consider designing a simple adapter board to plug the PedalShield on a Teensy 3.5 (recall ; Teensy 3.5 is a high end M4 Cortex board (with FPU), so it's 4 to 20X faster than Arduino Due M3 Cortex depending on what you do with it) with 22 ADC inputs (13 bits) + 2 DAC output (12bits), and much cheaper and smaller than a Due (good for smaller pedal packaging). That extra horsepower would be used to run many more simultaneous effects than the Due can ever do...

Thanks

PS
I want to notice that during my investigation I found this incredible tool to drive the Due DAC output :
www.hackster.io/BruceEvans/arduino-due-a...orm-generator-a9d180
github.com/Bruce-Evans/ArduinoDueArbitra...dController/releases
It works marvellously. A masterpiece. Hats off to Bruce Evans.

Please Log in to join the conversation.

6 years 1 month ago - 6 years 1 month ago #1385 by Ray
Replied by Ray on topic Non working Pedalshield
Thanks for your contribution and documenting so well your progress.

So I added a set of 220uF 6.3V cap on +5V and -5V (C17 & C18 beside TC1044S), and also on the 3.3V. I soldered them on the back side of the PCB as space left for capacitors was only for ceramic capacitors (see picture below). But still the pedal is incredibly noisy.

We added C17 and C18 caps to the board "just in case", but after placing many caps, sizes and types we did not see any improvement on the audio quality, this is why they are currently DNF (do not fit) parts. But well played, I would probably do the same if I get a noisy system.

Just to make my mind clear, I would appreciate if someone that has a good working Pedalshield Due could post a scope screenshot of the pedal in & out signal running the clean sketch (pedal engaged).

I am taking a plane in 3 hours, but next week I will post screenshots of what do you need so we can compare (I also have a Hantek around).

Did I miss something? Is my conclusion OK? Did anybody encountered the same thing? Please comment.
After all that time lost, money lost, and frustration, I wonder what are my options beside to procure a new Due?

I think that you pretty nailed it, I had people before with crazy problems and at the end, the root cause was a bad Arduino board. Let's wait until next week so I can post my screenshots and make final conclusions.

Please Log in to join the conversation.

6 years 4 weeks ago #1386 by Ray
Replied by Ray on topic Non working Pedalshield

Rendering Error in layout Message/Item: array_keys(): Argument #1 ($array) must be of type array, null given. Please enable debug mode for more information.

Please Log in to join the conversation.

Time to create page: 0.099 seconds