Preliminary Detailed Design
Team Vision for Preliminary Detailed Design Phase
- We planned to do preliminary testing of a possible control system on a scaled-down version of the existing control system, as well as, performing a feasibility analysis for the latch mechanism and process collected sensor data for validation.
- We developed the scaled-down version of the control system, validated operation and successfully was able to control it with Digital I/Os. Sensor data was processed with some unusual but explainable results and models were generated for the latch mechanism for analysis.
Prototyping, Engineering Analysis, Simulation
Electrical Test Bench
Existing SystemThe existing system utilizes the following components:
- 1 Four Pull Double Throw Relay
- 4 Double Pull Double Throw Relays
- 1 Triple Pull Double Throw Relay
- 1 Four Pull Double Throw (on-on) Switch
- 1 Four Pull Double Throw (on-off-on) Switch
- 1 Double Pull Double Throw (on-off-on) Switch
- 4 Standard Four Pin Male Mill Connectors
- 2 Standard Nine Pin Male Mill Connectors
Relays are categorized as electromagnetic switches. This switch is operated by a small electric current induced by an applied bias. In the state when no bias is applied the common port is connected to the NC (normally closed) port. When a bias is applied the switch activates and connects the common port to the NO (normally open) port.
Control System Schematic
Scaled Down Control SystemTo properly conduct tests, a scaled down model of the control system was built. The scaled down model was wired in the 'automatic mode' state which allowed for the 4PDT (on-on) switch as well as the 4PDT relay to be excluded. The following system parameters were scaled:
- Coil Voltage: 24V => 5V
- Input Signal: 120V AC Hot => 5V, 120V AC Neutral=> GND
- Output: Motors => LEDs
The first test was conducted to see if a relay could be tripped via a digital signal. An Arduino Uno was used to interface with the scaled control system. A code was written which contained the following functionality loop:
- Power Dome Motor
- Induce CCW Rotation
- Induce CW Rotation
- Induce CCW Rotation
- Power Down Dome Motor
- Power Shutter Motor
- Open Shutter
- Close Shutter
- Open Shutter
- Power Down Shutter Motor
The red LEDs represent power while the green and blue LEDs represent motion. The top row is dome control while the bottom row is the shutter control. The link below showcases the test results.
24V Relay Digital Operation FeasibilityThe control system in place uses 24V DC relays for automatic operation. The next test was conducted to see if it was possible to trip a 24V DC relay with a 5V digital signal supplied by the Arduino. The proposal, shown below, was to utilize a 5V SPST relay to act as an activator for the bias across the 24V DC relay coil.
The link below shows the 24V DC Relay being tripped by a digital signal from an Arduino.
The following schematic is based on the success of the previous test. By placing a SPST relay between the pin outs of the control system in place and the digital signal provided by the Arduino, the control system can be control automatically.
EncoderThe orginal 1000PPR encoder had stopped working while trying to test it with a the National Insturments myDAQ device. We have replaced it with a 128PPR encoder since then. Based on our previous feasability for the 1000PPR encoder, we believe that the 128PPR encoder will be suffcient for our needs.
The image below shows the processed data for the 128PPR encoder collected from the myDAQ and 1000PPR collected from the observatory. The 128PPR encoder was turned by hand just for feasabilitly of data collection while the 1000PPR sensor was turning from the dome motor. These results show that the 128PPR encoder, while turning at a nominal speed should be able to be processed by a NI DAQ device though further testing will be required for more accurate results.
Temperature, Humidity, & Pressure SensorThe following plots show that the Sensor operates well. Values seem reasonable for the cold rainy night the data was taken on. The random drops in points are due to nature of how we took data. This should not be an issue for the final system.
Gyroscopic SensorNo data was collected for the Gyroscopic Sensor because we incorrectly assumed it measured angular position when in actually measured angular rate. When attempting to mount the sensor the telescope for data collection, we realized the telescope was moving too slowly for the sensor to read any data that was not noise from just trying to hold the sensor in place. For this reason, the Gyroscopic sensor will be removed from the planned system. Alternately, we will try to either read the position of the telescope from the existing serial line to the computer or require a user input in a user-friendly front panel in the controller software where the RA/DEC of the target will be inputed.
- Available space/real-estate
- Motor torque sufficient to support moment arm and pull chain
- A method of clearing the moment arm from the chain's path of motion when latch is engaged
- Light, strong, and inexpensive materials
- Design structure for weight, strength, and cost
- Sensing capability to detect latch engagement
- Be able to connect to some sort of integrated control system, whether its own or connect with the dome controller
- Requires a safe and easily available power supply without cluttering the dome with extra cables
Design 1The first design incorporated the use of a hollow aluminum (Aluminum 6061) moment arm to reduce weight with connectors for easy assembly. This arm would provide the main means of transferring torque from the main motor to the chain. The necessary translational motion of the hook would be supplied using a rack and pinion concept, where a modified rack was the hook arm and the pinion(s) would be controlled by at least one smaller supplementary motor attached directly to the moment arm, unlike the main motor which is attached to the dome wall itself.
A standard rack-and-pinion model requires the motor to be located somewhere to the side of the moment arm so the pinion could rest on the rack arm itself. However, this provides an unstable structure with respect to too many degrees of freedom (DOFs), and so either two motors would be required, or another means of using a rack-and-pinion setup would need to be employed.
A secondary idea of centering a modified rack on top of a motorized adjustable gearbox was considered. An example of an inexpensive potential gearbox is as shown:
After incorporating this new idea into the system, it became clear that the design would pose many issues ragarding design, manufacturing, and utility, condensed into the following list, and would require a redesign:
- Generally overly complicated design
- Additional components add unnecessary mass and volume to the system
- DOF considerations were too counterproductive; translation rack required too many additional attachments to prevent translation in the X-direction or induced moments about all three axes
- Shaft design could be difficult to manufacture without more specialized tooling
- Complex geometry would make FEA analysis quite difficult, which is a key component of a feasibility analysis before money is spent on materials or components
Design 2The alternate design drastically simplifies most elements of the mechanism as a whole. It incorporates a similarly shaped moment arm to the first design, but has the option of being solid or tubed. The new design reduces component requirements and replaces translational motion with a hinge system designed to use two small electric motors in tandem to pull the hook arm vertically or past vertical for chain clearance. Strong, but flexible wire would be wound using these motors to raise and lower the hook arm as necessary. Initial base models of this design are shown below:
Latch Moment Arm Assembly Model - Extended Arm
Latch Moment Arm Assembly Realistic Render - Extended Arm
Latch Moment Arm Assembly Model - Retracted Arm
Latch Moment Arm Assembly Model - Retracted Arm (Alternate View)
Latch Moment Arm Assembly Realistic Render - Retracted Arm (Alternate View)
Note: Designs are works in progress and are subject to change as further analysis and/or testing is performed.
Engagement SensingTo satisfy the design requirement of the mechanism having a way to detect whether the latch is engaged with the slit panel or not, a number of different options exist for possible sensing methods ranging from strain gauges to piezoelectric pressure sensors. Based on cost and ease of use, the latter appears to be a better option, an example of which is shown below:
The piezoelectric pressure sensor pictured above is relatively inexpensive and can withstand forces up to 100 lbs, more than strong enough to avoid being crushed by the latch hook.
Cost AnalysisThe following list contains an approximate list of costs related to Design 2 shown previously:
- Main Motor: ~$100 - $500 depending on required size/torque
- Mini Electric Motors: $15 - $20 each (Need 2)
- Material Cost (Al 6061): ~$0.80/lb on average, total for 1.181 lb = $0.95
- 100 lb Piezoelectric Pressure Sensor: $19.95
Remaining Analysis Tasks and GoalsThe following incomplete tasks will be necessary to complete a full feasibility analysis for the Latch Automation. These tasks are expected to be completed in time for the Final Detailed Design Review:
- Model refinement
- At least three alternate design options established
- Prototype construction to verify dimension accuracy during the next dome visit
- FEA analysis to determine structural integrity and
risk of mechanical failure
- Includes thermal analysis to take extreme temperatures into account
- Continued cost analysis
- Continued sourcing of mechanical components and material suppliers
- Determine data and power supply methods/sources with rest of team
- Create drawing packages for finalized design
- Develop manufacturing and testing plans for MSD II
Feasibility: Prototyping, Analysis, Simulation
MaxIm DL supports the motion of the telescope and the result of the research is to use Meade software because it allows to connect the telescope to the PC and I assume that the connection is not setup properly. That software has the capability of activating the port in order to connect the telescope.The problem with the serial connection on computers is that some computers do not have serial connections on their motherboards. The old computers used to be have a serial port. So, one of the possible reasons of the disconnection is the port, and using this Meade software might solve this problem, which works as a driver.
- Gathering the information about the software to have a test plan for the next visit.
- First, use the cable and setup the serial connection defaulted to right COM port in the PC and the software’s properties. Then, surely confirm the three ports of the PC MaxIm DL, and Meade software are same to let the telescope talk to the PC.
Bill of Material (BOM)
EncoderInitial Encoder data was taken by rotating the dome in 90-degree increments to be consistency in the data. This was to allow use to make sure the expected value from processing data will be seen. From the plots above in our prototyping, this was not seen due to the time step used in the Arduino not being fine enough to pick up each pulse of the sensor.
A similar test will be performed with the new replacement encoder for feasibility. This time, instead of an Arduino, we will try to use a DAQ device from National Instruments to have a finer time step and hopefully alleviate this issue.
Doing a similar analysis as was done with the 1000PPR encoder, we should expect between about 10 to 32 pulses per degree of dome rotation with this encoder
Temperature,Humidity,PressureTemperature, Humidity, and Pressure was collected near the center of the dome just as a static value. The night we went out to collect data, was cold and raining so the lower temperature and high humidity is expected.
Future testing will be needed to generate a better baseline and see differences in values under different weather conditions to better be able to predict inclement weather approaching. This will likely include running the sensor inside the dome for several days and collect data over that time to see how sensitive the sensor is.
DebuggingThe scaled control system was physically built on a breadboard. The circuit did not function at first and was debugged utilizing an oscilloscope in conjunction with a multi-meter. The schematic was updated with the necessary changes.
VerificationA DC power supply set for 5V was used to power the system as well as supply input signals to the relays. The following tests were conducted:
- Physical Relay Engagement Test-The system was powered and the relays were physically tripped via a wire connected back and forth to ground. This verified the circuit was functioning properly.
- Digital I/O Relay (5V) Engagement-To ensure we are able to engage the relays automatically, a set of digital I/O pins from an Arduino were connected to the coils of the relays. All relays engaged properly and the test was a success.
- Digital I/O Relay (24V) Engagement-The relays in the dome control system use a 24V coil. This test was done to ensure a low voltage digital I/O signal could engage a 24V coil. The first setup involved an op amp circuit to amplify the 5V to a higher bias. This did not work due to a common ground issue. The second attempt utilized a 5V spst relay to act as an intermediate between the digital signal and the 24V relay. The 24V relay engaged and the test was a success.
Design and Flowcharts
For a master list of project risks, click here.
Plans for next phase
Ahmed Alhurubi Three Week Plan: Ahmed's Goals
Joseph Brescia's Three Week Plan: Joe's Goals
Raymond Castro's Three Week Plan: Ray's Goals
Wilson Quizhpi Three Week Plan: Wilson's Goals
Sarah Williams Three Week Plan: Sarah's Goals