Jun 13
admineOrbo replication 2 Orbo, Steorn
Finally I have received and installed the new 3D controller for the Profiler milling machine. Then I wrote a test file with G-codes, but it didn’t work right away. Seems that the controller takes feed rate as mm/sec, but the control program (ColiDrive) expects mm/min in files. I have contacted Colinbus support asking for a file format specification. I have also readjusted the Z-axis – everybody say it’s a bitch to do it and it really is. Now I believe it is as good as it can be and the nNext step will be to start milling….
Jun 01
admineOrbo replication 2 Orbo, Steorn
I went to a local glass shop today to get some Lexan. The pieces I need are relatively small and they had leftovers I got at a reduced price. I guess they knocked off 70-75 % which I really appreciate because Lexan is expensive. Plexiglas is less expensive, but not an option because it melts in the milling machine. If I can find the best milling bit, spindle speed and feed rate for the job I am quite sure the melting can be avoided. Lexan can also be difficult to mill, but I have found a way that works.
Yesterday I ordered and paid for a new 3D controller board for the Colinbus Profiler milling machine. This will replace the 2.5D controller board currently in the machine. As I said earlier there is a problem with the machine and it was either necessary to repair the 2.5D controller or upgrade to a 3D controller. The new controller can read G-code, it can be configured by software and I believe it can be flashed when there is a new firmware release.
May 30
admineOrbo replication 2 Orbo, Steorn
Yesterday’s post was a mistake. The hall switch is sensitive to a single magnetic pole on one side and to the opposite pole on the other side. I knew this and at the moment I was thinking that if the sensor didn’t sense I could just move it to the other pole. That is correct only if the same side of the hall switch is facing the other pole. This was not the case yesterday. I was lazy and didn’t think when I just moved the sensor around. The way to do it is to turn the sensor because it will work every time and does not depend on the situation at hand.
With my rotor the magnetic field is shooting up and down and not radially out the where the toroids are placed. This is not optimal, but having pockets in the rotor is attractive seen from an experimenter’s point of view. I can easily play with different size magnets just to see what happens. This may or may not be the best way to replicate eOrbo. It is, however, a good way to learn more.
The motor was started with a single toroid and with strong enough magnets the acceleration is good. I like running it with a load and the load this time was fan blades. It accelerated to around 3000 RPM. In my first replication the load was the fan motor (converted to a generator) and I used two toroids. This means the two replications cannot be compared at this time.
May 29
admineOrbo replication 2 Orbo, Steorn
I connected everything together, put magnets in the 8 pockets of the rotor and holding the hall sensor in my hand I moved it around to see if my circuit was switching the toroids on and off as expected. It didn’t, the hall sensor did not sense the magnetic field at the poles at all. I tried above and below. As I moved it between those two positions it turned on near the center of the magnet, but at this position it didn’t turn off when I turned the rotor. Anyway, I had some changes to make and removed all the magnets from the pockets. When I was ready to try again the hall sensor behaved as expected and I didn’t think more about it. At least not before it happened again. Thinking I had blown the sensor I checked with an extra magnet. No problem at all. This lead me to believe the magnetic field around the rotor had two different and stable configurations. One that behaves as I expect and another one. I experimented for a while and so far it seems that when I put the magnets in the pockets I get one or the other configuration, but I never know which one.
May 26
admineOrbo replication 2 Orbo, Steorn
I have mounted the PC fan on an MDF board. When I started replication 2 I wanted to be able to raise and lower the toroids to find the optimal position, but now the PC fan with the orbo rotor can be raised and lowered instead. As quickly as possible I will mount the two toroids at fixed positions and see if I can make this thing run. It will run, that is not the hard part, but I do not know how well it will run. At this stage I will not even test for excess energy.
May 24
adminAlone at Home II Humidity, Mold, Mould
I have been working on a new data structure for storing the IR commands for the air conditioner. The best format is the decoded IR stream. Each time a button is pressed on the remote control two command are sent – the first decodes to 8 bytes and the second to 19 bytes. These 8 and 19 bytes are what will be stored in memory and just before it is sent it will be converted to the much longer stream driving the IR LEDs. I have already written functions to decode the raw IR stream to command bytes and back to IR stream. To save even more space I can store only one command and modify it by patching in the function (dry, heat), temperature and generate a new checksum. The task now will be to change the software to use the new data structure but keep the functionality as it is.
May 23
admineOrbo replication 2 Orbo, Steorn
I have returned to my first location where I do not have a milling machine, but all my other tools and parts. I have picked up the magnets at the post office and fortunately they fit in the holes. I bought a digital vernier caliper so I knew that already. Oh boy am I glad I spent the extra money to get a digital one! The holes for the magnets are not holes but pockets. I can just drop the magnets in and they will stop at the bottom. The plan is to experiment with the number of 1mm magnets in the motor. I will try both with insulation between the magnets in each stack and without. The insulation must be very thin and is supposed to reduce eddy currents and make the motor more efficient. The rotor has been mounted to a PC fan and is fairly well balanced. The immediate task is to mount the fan to a board and to make something to hold the toroids. Some of that will have to wait till Tuesday when the stores open again after the holiday. Because I now don’t have access to the milling machine the toroids will be in fixed, non-adjustable positions.
May 19
adminAlone at Home II Humidity, Mold, Mould
To control the air conditioner the system needs to send the same infrared commands a remote control does. To do this I first need to record all the the commands I may need into my computer. That is what I did today – 60 different commands. They are quite long and I need to decide how to store them effectively because the Jennic boards have limited memory. So far I have stored the commands in the format requiring the smallest and simplest program. With so many additional commands I will have to optimize.
May 17
adminAlone at Home II Humidity, Mold, Mould
The system is connected to a “slug” (Linksys NSLU2) which saves information in a database. The information sent by the system is basically measured data and a time stamp. This information is passed on by the “slug” over the internet to a remote computer. So far it has been a one way street, I cannot send commands to the system from a remote location. Now I need to remote control the system by setting a temperature so I can arrive to a heated house during the winter months. To achieve this I need to change the strategy I use to lower the relative humidity by heating the air. As an added bonus I believe the new strategy will work better and maybe use less energy too. As always the problem is to start ripping the software apart…
May 16
adminAlone at Home II Humidity, Mold, Mould
Because the Jennic hardware is so good and close to what I need, the project turned out to be a software project. The main purpose was to make something quick that would do the job for me in my house. I was debugging to the moment I closed the house for the winter and left. Fortunately I have a neighbour who is a good and capable man. He had to reset the system every now and then, but it actually worked much better than I dared to hope for. As always when write version 1 of a software project I learn a lot and when finished I know how I should have made it in the first place. To get rid of the expensive Zigbee network stack I used and to make a more robust and extensible system I decided to write version 2. This new version is currently running and with some improvements it is finished.
Older Entries Newer Entries
RSS
Recent Comments