Build & Test Prep
Table of Contents
Team Vision for Build & Test Prep PhaseThe team's plan for this phase is to review our MSD I progress and prepare for MSD II. With preparing for MSD II, the test plans need to be completed, along with the software design. More detail needs to be added to the software flowchart. After meeting with David on 1/19/19, the team will program the board with David's pass through software that will allow a signal to be passed to an amplifier. David's Bill of Materials (BOM) will also be completed.
The team has reviewed the MSD I work, and has completed the test plans. The team attempted to program the board, which was successful. However, the diagnostics test that David provided failed, likely due to not being able to connect to the external clock provided on the board. Debugging this issue will continue through the first half of next phase. The pass through software will be programmed next after the board is programmable and working. The Bill of Materials is completed as well.
Test Plan Summary
The test plans can be found in the attached link. Each test plan in a separate sheet, labelled by Sx, where x is a number 1 thru 9.
Most of the equipment necessary for testing can be found in any of the EE labs on 3rd floor of Gleason, or owned by a team member. No equipment will need to be ordered. Access to a Spectrum Analyzer will be needed to complete the testing. An alternative to this would be to use the Analog Discovery Module if access to a Spectrum Analyzer is denied.
To start with the software design, we will briefly review the board hardware. Below is a detailed block diagram of the pcb hardware:
The programmable parts of the board are the 3 ADCs and the PIC microcontroller. Each ADC used 2 strings and multiplexes them together, yielding 3 I2S signals that can be processed by the PIC.
We will begin our software with a simple overview flowchart:
This is a bit different from the flowchart last semester. At the beginning of each while loop, a number of samples are taken and added together (multiplexed). The digital encoder is then used to find the clipping cutoff level, and the block of samples is compared to the cutoff level. Any level below the cutoff is passed through, and any level above the cutoff is reduced to the cutoff and then passed through.
The selection process for clipping can occur similar to the image shown below.
The important takeaway from this image is that the selection for clipping is quick, which will limit the latency. The output in this image refers to the output to the block in the Basic Software Block Diagram directly above this image named "Set Clipping Level". The mathematical equation that will set the clipping level will change depending on where the clipping will be, which will be determined by testing various levels and choosing the best levels that represent the distortions we want.
One interesting thing with this design, is that at the sensors, there are six distinct electrical signals, one for each string. Traditional electric guitars have one electrical signal that contains sound from all six strings, and to get our device to work with traditional guitar setups, these 6 signals need to be multiplexed together. Below is a block diagram that summarizes the hardware and software multiplexing that occurs:
The signals are combined once in hardware and once in software. This shows how the signals are multiplexed in hardware by storing arrays and adding the arrays together elementwise, before any processing happens and before the signal is pushed to the output of the board.
Risk and Problem TrackingRisk Management can be found here. Below is an image of our updated risk assessment.
This image shows a new risk: debugging the board. This has manifested itself and requires the team to take time to debug the board. With David's help, this risk will be mitigated, as he has given us some new updates (as of January 2019) for the board that should fix this issue. This will be our first steps in debugging before tracking down more problems.
Most of the mechanical adjustment mechanisms are in the toolroom to be made. Getting this in early is very important, as we are currently first in line.
The guitar will now be routed on a manual upright mill, since the dust created from milling wood may gunk up the CNC mills. The guitar will be completely disassembled, and then clamped to one of the Bridgeport mills in the ME shop. One last check will be done to make sure everything is ready to go before we rout the guitar. We will start routing by next week the latest.
Due to an error, a new pickup plate has been designed. The current rout in the guitar is larger than expected, so a larger plate will be made.
Finalized Bill of Material
Below is a list of all of the components used to populate the PCB of the pickup. This is mostly completed, as only the connectors need to be added to bill of materials. Furthermore, the cost of these components also remains to be determined. Some of the ceramic capacitors used in the schematics are out of stock for many retailers, which is what is causing this delay. The anticipated cost for all of these parts is between $100-$200.
Next is a list of all the materials used outside of populating the PCB. This includes any aluminum needed for the pickup fabrication, through hole optical sensors, etc. The cost of these parts not including shipping or price breaks is $120.82, with roughly two-thirds of that being due to the cost of the aluminum bars, springs, and screws.
Design Review MaterialsOur materials for the review are the following:
- Pre-Read: This page
Plans for next phaseAs a team, next phase the team will be putting together their individual components to prepare for integration. The machine shop will begin working on our mechanical side of the pickup, and the electrical team will work on debugging the board and getting a pass-through software working on it. Along with that, the distortion algorithm will be implemented on the board after the pass through is running.
Each of the individual phase goals and tasks to be completed can be found in this document.