This week the team continued working on the power supply, Arduino/touchscreen UI, and microphone preamp.
Team
Tuesday afternoon the team met at the open lab on West Campus to conduct max-load bench testing of the power supply.
Power resistors were attached to each voltage rail in order to give each it's own current draw. The output voltage was then measured across the load resistances in order to assess the power supply's performance.
The 3.3 and -5V supplies had no issue supplying their respective currents (~50mA, and -35mA). The 12 and 5V regulators did not perform optimally. Both regulators became too hot and entered thermal shutdown.
Thermal shutdown shown below of the 5V regulator:
The team decided to redesign the circuit using two 12V regulators in parallel. These will power the 5 and 3.3V together and spread the load current. Repurposed heatsinks (pulled from a junk TV power supply) will also be added to aid in heat dissipation.
Luckily, the team had also purchased some switching 12 and 5V regulators that were initially to be used with the battery supply. These may also offer a potential solution if the linear regulators prove troublesome.
John
On Monday, John visited Skycraft to purchase some power resistors for the power supply bench test. Using the proposed power budget and Ohm's law, John was able to find the equivalent resistance that would draw the specified current from each voltage rail.
For instance:
12V supply specified current is 320 mA
12/0.320 = 37.5 Ohms
12*0.320 = 3.84 Watts
So, the dummy load for the 12V supply will need to meet these specs.
2 75 ohm 10 Watt power resistors wired in parallel gives a resistance of:
75/2 = 37.5 Ohms
This same procedure was followed for all other resistors. The final list of specs
48V(+-24V) - 137 Ohm - 16.8W
12V - 37.5 Ohm - 3.84 Watts
5V - 7 Ohm - 3.55 Watts
3.3V - 66 Ohm - 0.165 Watts
-5V - 143 Ohm - 0.175 Watts
The closest available resistors where chosen. The 1/2W metal film resistors purchased from amazon can be used for the 3.3 and -5V rails
After the on-campus testing revealed some inadequacies in the power supply implementation. John worked to construct a prototype board of the revisions he and Hunter thought appropriate.
By adding a second regulator in parallel and some heat sinks, the current should be spread between the two 12v regulators and the heat should dissipate more quickly.
After completing the new power supply, John revised the simulation of the preamp circuit to prepare for it's construction. The cap values were changed to reflect the parts the team has on hand and a 2nd order LPF was swapped in to see how it would affect the frequency response and group delay. All component values were found using the Sallen-Key filter designer tool in Octave.
Final output is shown in blue. HPF and LPF -3dB frequencies are 0.85 and 5kHz respectively.
A group delay of 140uS at 300Hz is likely too high. 1/10th of the period of 300Hz is ~333uS
Modifying again for a first order LPF with cutoff 2Khz
80uS of group delay is a good improvement but more simulation work will be done before breadboarding.
Hunter
This week, Hunter updated the “weekly minutes with the professor” page.
Screen calibration was required before programing buttons and shapes to the Arduino's TFT display. The blue box in the picture below shows the original, and new, pin numbers after successfully completing the calibration.
This method is used for different types of shield/TFT screens using different pins:
The calibration method, can be seen in the picture below, and is done by pressing down on each square as they are highlighted, one at a time.
The results are displayed in the program and on the TFT screen once complete:
The X and Y axis values can be seen below green, after calibrating:
This video below shows the X and Y values changing, and functioning properly, as the TFT LCD is touched and finger slides on the screen at various locations.
Testing the programmed buttons, shapes, text, and color output, which can be applied for designing the graphical user interface (GUI):
Comentarios