Yoonseo Log
Present
3/10/2012
Installed Hall Effect Sensor Module x2 on CEB Press V4.0 with bent pieces of aluminum flat bar and small C-Clamps.
Resolved a broken inverter issue with Marcin by connecting the workshop grid to one of the remaining inverters. Reliability is definitely a significant design principle.
Discussed short-term implementation of forums/stack overflow-like software to help development process. 2 major factors for the success of such software identified as access and action relevance. Access made easier with invites, action relevance made easier by the right questions to prompt the right answers.
Tested the CEB Press with manual control. Teamed up with Marcin to fix some issues. Troubleshooted total non-movement by increasing up the hydraulic pressure from the powercube. Troubleshooted primary cylinder non-movement by loosening front support plate, but discovered that the real problem was a loose quick connector connection from the primary solenoid valve. Full manual control success. Back to solo work.
Tested the CEB Press with test code. Troubleshooted weird sensor-triggered movement by examining sensor analog values. So! The hall effect sensors switch and latch on and off, but electrical noise and in-circuit imperfections cause fluctuating output signals. To add to the problem, the magnetic field of the magnets interact with the sensitivity of the sensors as to cause multiple latching events when the sensor passes a single magnet!
How to make the sensors work properly?
1. Take multiple sample readings and average them into a moderated value; this will diminish the effect of short-lived extreme fluctated signals. Increasing the number of samples per averaged reading and decreasing the delay time after each sample helps, but will be limited by the processing speed of the electronics.
2. Use separate upper and lower trigger values for the microcontroller to determine when the sensor has passed a magnet. This ensures that fluctuations that pass a trigger value can be safely ignored to an extent by the microcontroller. 2 values that help determine the upper and lower trigger values are: minimum value in the high-latched sensor state and maximum value in the low-latched sensor state. If possible, have the lower triggger lower than the aforementioned minimum value and the higher trigger higher than the aforementioned maximum value.
3. Use post-latch delays. That can be as simple as putting a fixed delay after each latch-switch. This helps minimize the effect of sensor sensitivity to strong irregular magnetic fields but limits the operating speed of the sensor-controlled movement (ex. if the delay is 1 second, then the magnets need to be at least 1 second apart for the microcontroller to have enough time to pick up on the next latch point as it passes the second magnet).
Made significant revisions to the testing code for extreme troubleshooting ease. CEB Press Test Code
Tested the new test code to success after implementing all of the make-sensors-work-properly changes. Started to revise the operation code for continuous brick pressing.
Contacted Star Simpson about inverter design topologies.
Scraping in a Dovetail - Detailed
2/10/2012
Using arduino IDE 64-bit on ubuntu 12.04 - Guide to Arduino make sure you open arduino as power user with sudo ./arduino when in the correct folder! Otherwise the serial port will be greyed out.
Hall Effect Sensor Module V2 x2 complete
- Research for choosing between rod, dovetail, or boxed way for linear guide (whoa there's V-ways and half-round ways too!) Gah I need a free download of the book "Foundations of Mechanical Accuracy"
Dovetail
You are not under the impression that the dovetail cutter needs to be the full width of the female dovetail are you?The cutter only needs to be as tall as the dovetail.Your remark about not having large dovetail cutters prompts my question.
Dovetails only need 1 gib
Actually the dovetail is the simpler mechanism. It is simpler to make and simpler to adjust in use. The one gib at 45 or 60 degrees takes up slack both sideways and vertically. One adjustment and you are done.
I was afraid of the results before I made my first dovetail, but it is a piece of cake if you plan it properly. The setup is the same for both male and female parts. As mentioned above, you do not use a cutter that is the full width of the female dovetail. You use a standard milling cutter to rough out the center, usually to a depth slightly greater than the actual dovetail. then you cut each side individually. The cutter only needs to be tall enough to cut the whole angle at one time.
I used a CAD drawing and placed even sized circles tangent to both the angled sides and the flat top/bottom surfaces to give me a way of measuring the width of the dovetail. You can read the distance between them off the drawing to three or four places and them use drills or ground, round stock of the same size to measure your work. Just mike between them on the female dovetail and across them on the male. Real quick way of checking your work on the mill. Of course, with a gib, you don't need extreme accuracy.
Box
Boxed way more rigid than dovetail
Boxed way conventionally steel with turcite for anti-friction contact surface OR cast iron with oiling OR metal with high-pressure oil (hydrostatic)
Boxed way conventionally ground then scraped
Future
CNC Circuit Mill V2
CNC Multimachine
Universal Power Supply
Encoder Test
Mini Clamp
Battery Holder
Visual Display Module