Because, this blog has been so useful at keeping me on track with tasks, i'm going to lay down a few things for my own references:
Troy wants me to do stellar electrical wiring, so he referenced me to:
http://workmanship.nasa.gov/guidadv_recmeth_wusplice.jsp
Electrical Problems to tackle:
Opto-isolation
We are also discussing the prospect of integrating Opto-isolation into the circuit. From what I read on http://www.epanorama.net/circuits/parallel_output.html#circuithow the introduction references that simply installing a PCI Parallel port card should protect your motherboard from frying. Troy makes the good point that just because there is a card that it doesn't protect your PCI bus. (sadface) But from Newegg at 15$, it's not a steep price in the offhand chance that the guy is right. Maybe some inspection of the components of the card would do the trick, or questioning folks on CNCzone may do the trick as well.
http://www.newegg.com/Product/Product.aspx?Item=N82E16815166006
http://www.newegg.com/Product/Product.aspx?Item=N82E16815166007
Wiring in the CNC4PC controller.
Since Mach3 functions through the parallel port only (i think.) We cannot use the RJ45 port to control spindle speed (i think - i must verify this.) As such that means we will have to wire in that board into the specific pins we decide on the parallel port. Naturally, this means we will have to make our own circuit for this purpose. The interesting bit, is that we can integrate the opto-isolation components at this phase (hopefully.) The first 7 pins are reserved already so we are going to using the next couple pins for 6 limit switches (which I just bought from Fry's at 2$ a piece,) an estop button, and then finally the spindle control com pins.
Voltage Drop
Lastly, we need to figure out how to voltage drop from 49VDC to 36VDC to make full use of the steppers, currently we are at 24VDC. So our blunder of not accounting for Vrms of the toroid transformer is causing us problems. Fortunately, with the 18VAC line, we get 24VDC through the rectifier/cap. So we are at least functioning, and without opto-isolation. We considered using a bunch of resisters for a moment, but it looks like we would need literally about 100 to do the job. Frankly, that is not acceptable. Another option is to simply buy another toroid transformer with different output. I wish I had tried harder in fundamentals of electrical circuits. Now I must research more. Oh well, lesson to be learned, do well in school! Always! LOL.
Tuesday, September 22, 2009
CNC update: Skipped steps fixed, ballscrews in
Fixed Stepper missed/skipped steps
With my fixed 2500+ AMD desktop I was able to hook up and test the control board again. A drivertest of the parallel port showed that the computer could provide a steady 25 khz waveform. In fact is was around 26 khz for the test, not really sure why, but it was consistent and as such yielded a solid line with very minor blips in the waveform. I did have the system tuned in the services and applications down to a measily 14 processes in windows XP. So basically, the system was running bare bones.
With the breakout board hooked up, along with the steppers, I was able to move them around smoothly with mach3. Compare the video with the older video where you can visibly see the steppers doing wierd things.
http://www.youtube.com/watch?v=ywBT3VNERIg
Ballscrews in!
I also went to pick up a package from USPS, because they needed a signature and no one was at the door when they arrived at my house. I picked them up and was surprised to see the packaging a bit damaged. I was also concerned that something might have fallen out.Frankly, I was hoping the ballscrews would have been protected a little better. Oh well, to get those three complete ballscrews with bearings for 375$ shipped is a downright steal. The value well makes up for it.
The condition of all the components was agreeable. The threads on the ballscrew were a little rough, which looked like my first attempts at threading something in my machining class. I'm sure they will do the job of securing the ballscrew though. Surprisingly, some of the bearings were NSK! High quality stuff! Unfortunately, I don't think the angular contact bearings are though - they were some other unknown brand. I'll have to look into that later.
The end machining is the feature i'm most concerned with. The NSK bearings simply wouldn't go on the shaft. They were too tight. I measured them to be something like 9.91 +/-0.01mm and the end machining came in at about 9.99 +/-0.01mm. I think I may have to chuck them up in a lathe and use some emery cloth, or sandpaper to bring them down to an acceptable size. I don't feel like bashing the bearings on with a hammer. Something just doesn't feel right for an interference fit. The stepper couplers are interesting... They look like they will work great with set screws. The interesting bit is that they expand/contract vertically, but they do not twist at all to lose precision. Clever.
With the prospect of machining the ballscrews further, I was concerned about leaving the ballnuts on the ballscrews. From a quick search of cnczone I found this:
http://www.cnczone.com/forums/showthread.php?t=86614
From this thread I've learned that you can remove ballnuts by using a mandrel to hold the balls in place. I also learned that i probably don't need to remove the ballnuts from the ballscrew to machine them. Yay, I like it when things follow the KISS principle. I am contemplating how I will stabilize the ballscrew in a lathe. I'm thinking a live center should do the trick, I just don't want to have to drill the end of the ballscrew for that purpose. I do know they have those brass triple finger support type things (forgot the name) which may do the trick, I just don't want to risk damaging the screw. Perhaps more research is in order about chuking up a ballscrew for machining.
Saturday, September 19, 2009
The Beach Boys FTW
I saw THE beach boys this weekend at Chinook Winds Casino today. Don't worry, I didn't spend a dime on gambling. The main singer sounded exactly like you hear on the radio, it's amazing. His voice is unreal. Ok! Back to engineering stuff.
Stepper Problems
I've begun researching why there were weird things going on with the steppers as you can see in the youtube video in the previous post. CNCzone.com is where i've begun my quest for knowledge. I previously thought it was resonance, but after a cursory read I don't think that is what is going on. I suspect that the steppers are missing steps, or skipping them due to the slow 750 mhz computer. Here are some relevant links about steppers:
- Making Steppers run reliably
- Stepper resonance
==================================================
Old computer and Overclocking
I intend to restore my old desktop computer to it's former glory(dead OS harddrive) and see how well that drives the stepper motors. I suspect if anything it will help to have a faster computer anyways :) It has a laptop processor 2500+ in it, that I used to overclock to about 3200+ speeds. (The laptop CPU's used to be renowned for the overclockability)
When AMD made a batch of processors they would go through and test and catogerize processors. The laptop processors were required to run reliably 1.833 ghz at 1.45 volts, they would seperate these "good" processors from the other "average" processors that would run reliably at 1.833 ghz at 1.65 volts. The "average" processors were then labeled as desktop processors. They are the exact same processors, pin's and all. So somebody akin to this knowledge would *cough* *cough* me - would purchase the laptop processor for the exact same price throw it in their desktop system, and raise the voltage to 1.65 volts (since all those processors were rated for the same max voltages.) Raising the voltage .2 volts lets you overclock quite a bit. With a healthy heatsink I was able to squeeze an extra 367 mhz with just those extra .2 volts on air cooling, and maintain the exact same stock temperature of about 39'Celsius. That would put the effective performance at the equivelant 3200+ processor (2.2 Ghz). So basically you get a 3200+ upgrade from a 2500+ for free! Woohoo! The highest I ever pushed it was 2.4 ghz - The temp's and voltage were too high for me to comfortably run it for any period of time though.
====================================================
Airsoft PartsI also went to the shop and updated one of my airsoft part's dimensions, and cranked out 248 parts. I'm trying to prepare a healthy stock for while I am at school I will not have access to a laser cutting machine. I probably will make my way over to the shop a few more times this week to crank more out, and move other projects along.
CNC BallscrewsI hope the ballscrews from ebay come in this week. I'm amazed at the price we got them. A quick comparison to roughly equivelant ballscrews off of Mcmaster showed that it would cost 3000$ for equivelant, mounting blocks, bearings, ball nuts, etc... Not to mention they have more than twice the tolerance (this is a bad thing lol.) The C7 ballscrews, ballnut, mounting blocks, bearings, etc.. All costed 375$ shipped. Yay ebay! Talk about savings...
Organzing Your Desktop
In any event, I spent 15 minutes organizing my desktop into appropriate folders because i simply have too many projects. This made using my computer much easier to navigate :) I suggest you do the same.
CNC Status
I promised I would get more pictures up, so here's a good one. It's the current build of the CNC machine we are building. I think we may need to change it over to a gantry style CNC. I need to sit down and see if I can run some calculations to determine if the Z-axis will be rigid enough, and if the HSR rails can take the load. While i could be mistaken, I recall some arbitrary number that the rails could handle being around 1200 lbs.
Stepper Problems
I've begun researching why there were weird things going on with the steppers as you can see in the youtube video in the previous post. CNCzone.com is where i've begun my quest for knowledge. I previously thought it was resonance, but after a cursory read I don't think that is what is going on. I suspect that the steppers are missing steps, or skipping them due to the slow 750 mhz computer. Here are some relevant links about steppers:
- Making Steppers run reliably
- Stepper resonance
==================================================
Old computer and Overclocking
I intend to restore my old desktop computer to it's former glory(dead OS harddrive) and see how well that drives the stepper motors. I suspect if anything it will help to have a faster computer anyways :) It has a laptop processor 2500+ in it, that I used to overclock to about 3200+ speeds. (The laptop CPU's used to be renowned for the overclockability)
When AMD made a batch of processors they would go through and test and catogerize processors. The laptop processors were required to run reliably 1.833 ghz at 1.45 volts, they would seperate these "good" processors from the other "average" processors that would run reliably at 1.833 ghz at 1.65 volts. The "average" processors were then labeled as desktop processors. They are the exact same processors, pin's and all. So somebody akin to this knowledge would *cough* *cough* me - would purchase the laptop processor for the exact same price throw it in their desktop system, and raise the voltage to 1.65 volts (since all those processors were rated for the same max voltages.) Raising the voltage .2 volts lets you overclock quite a bit. With a healthy heatsink I was able to squeeze an extra 367 mhz with just those extra .2 volts on air cooling, and maintain the exact same stock temperature of about 39'Celsius. That would put the effective performance at the equivelant 3200+ processor (2.2 Ghz). So basically you get a 3200+ upgrade from a 2500+ for free! Woohoo! The highest I ever pushed it was 2.4 ghz - The temp's and voltage were too high for me to comfortably run it for any period of time though.
====================================================
Airsoft PartsI also went to the shop and updated one of my airsoft part's dimensions, and cranked out 248 parts. I'm trying to prepare a healthy stock for while I am at school I will not have access to a laser cutting machine. I probably will make my way over to the shop a few more times this week to crank more out, and move other projects along.
CNC BallscrewsI hope the ballscrews from ebay come in this week. I'm amazed at the price we got them. A quick comparison to roughly equivelant ballscrews off of Mcmaster showed that it would cost 3000$ for equivelant, mounting blocks, bearings, ball nuts, etc... Not to mention they have more than twice the tolerance (this is a bad thing lol.) The C7 ballscrews, ballnut, mounting blocks, bearings, etc.. All costed 375$ shipped. Yay ebay! Talk about savings...
Organzing Your Desktop
In any event, I spent 15 minutes organizing my desktop into appropriate folders because i simply have too many projects. This made using my computer much easier to navigate :) I suggest you do the same.CNC Status
I promised I would get more pictures up, so here's a good one. It's the current build of the CNC machine we are building. I think we may need to change it over to a gantry style CNC. I need to sit down and see if I can run some calculations to determine if the Z-axis will be rigid enough, and if the HSR rails can take the load. While i could be mistaken, I recall some arbitrary number that the rails could handle being around 1200 lbs.
I am proud of what we have accomplished with the CNC machine so far. As we are quickly aproaching a linear moving, computer controled system, I look at our project with caution. As I am going back to school in a week, I will have to take the components down with me and finish off the machine by myself - which isn't a simple prospect in of itself. The logistics keep cropping up problems.
A preliminary assesment puts the frame cost around 500$, while we could make it lower cost, we are looking to make a rigid, precise, modular system that will allow upgrades in the future. As such, the design is oriented for the future the cost accordingly goes up. It also means - I will likely have to make the spindle. Eeep. Thankfully our school has excellent machines for the task.
My goal for our team is to get all the electrical components, flawlessly interfaced with the computer and Mach3 before I head down for school. While we have the 3 steppers interfaced, there still is the skipped steps to address, and tuning up a computer for the CNC; there still remains the task of getting a powersupply working for the treadmill motor, hooking it up to the CNC4PC board, and interfacing that board with Mach3. A big, but reasonable goal. Even if it doesn't work out this week, we are all patient, and want to do this well and correctly.
Engineer on!
A preliminary assesment puts the frame cost around 500$, while we could make it lower cost, we are looking to make a rigid, precise, modular system that will allow upgrades in the future. As such, the design is oriented for the future the cost accordingly goes up. It also means - I will likely have to make the spindle. Eeep. Thankfully our school has excellent machines for the task.
My goal for our team is to get all the electrical components, flawlessly interfaced with the computer and Mach3 before I head down for school. While we have the 3 steppers interfaced, there still is the skipped steps to address, and tuning up a computer for the CNC; there still remains the task of getting a powersupply working for the treadmill motor, hooking it up to the CNC4PC board, and interfacing that board with Mach3. A big, but reasonable goal. Even if it doesn't work out this week, we are all patient, and want to do this well and correctly.
Engineer on!
Friday, September 18, 2009
It's Alive!!!
After buying a new Straight Thru DB25 M/F cable from fry's 4.99$ (again) and fiddling with mach3 configuration I was finally able to get movement!
After I bought the cable i came right home, plugged it in and attempted to run it. It didn't work. A moment of fear occured because I thought I had not identified the problem properly. I had to disable the E-stop, ensure the pins were sending the right signals according to the provided documentation, AND make sure they were assigned to the active motor port. So i had to change six different 0's to 1's, and assign pin numbers again. No biggie, after that the X stepper moved after a G0 X2 command! Yay.
I quickly recorded a video:
http://www.youtube.com/watch?v=ay-3UZTvABA
I was so caught up in the moment I didn't really notice the resonence, or the missed steps on the steppers. So that is the next hurdle to tackle, but at least all the steppers are moving in the direction I want them to :)
This CNC machine is proving to be like a child. You pour in love, pour in love, and get little back, but when you do get something back it is soooo rewarding. (not that I am an expert on children by any means.) So after a bit of tuning, the focus will shift to the CNC4PC board, and getting the spindle motor working as we want.
I also need to clean up the powersupply case, properly mount the control board, cut up some wire so we can easily attach the steppers where they ought to be on the actual machine. I'm not entirely sure on the layout we will take, but given enough time I'm sure we will find a solution. This was definitely the hurdle I feared most about this project, but it is quickly resolving down to the simple mechanical aspects, and money.
More to come later. I still owe you some pictures too.
After I bought the cable i came right home, plugged it in and attempted to run it. It didn't work. A moment of fear occured because I thought I had not identified the problem properly. I had to disable the E-stop, ensure the pins were sending the right signals according to the provided documentation, AND make sure they were assigned to the active motor port. So i had to change six different 0's to 1's, and assign pin numbers again. No biggie, after that the X stepper moved after a G0 X2 command! Yay.
I quickly recorded a video:
http://www.youtube.com/watch?v=ay-3UZTvABA
I was so caught up in the moment I didn't really notice the resonence, or the missed steps on the steppers. So that is the next hurdle to tackle, but at least all the steppers are moving in the direction I want them to :)
This CNC machine is proving to be like a child. You pour in love, pour in love, and get little back, but when you do get something back it is soooo rewarding. (not that I am an expert on children by any means.) So after a bit of tuning, the focus will shift to the CNC4PC board, and getting the spindle motor working as we want.
I also need to clean up the powersupply case, properly mount the control board, cut up some wire so we can easily attach the steppers where they ought to be on the actual machine. I'm not entirely sure on the layout we will take, but given enough time I'm sure we will find a solution. This was definitely the hurdle I feared most about this project, but it is quickly resolving down to the simple mechanical aspects, and money.
More to come later. I still owe you some pictures too.
Thursday, September 17, 2009
Serious CNC Progress
Sorry I haven't posted any updates in awhile. Women. lol.
We've received our CNC4PC board in the mail, it was delayed cause they were out of stock. It will control the spindle, it is controlled by an RJ45 port. (ethernet)
I've been working with Noobie on AirsoftRetreat to develop a new hopup solution. We've made considerable progress on that, I will release more information as we release it to the public. I made some prototypes and sent them to him for testing by USPS.
Troy tested a few of some old ATX powersupplies and used a jumper to quickly test the output of the powersupply. This website tells you how to jump it. Interesting...
http://www.duxcw.com/faq/ps/ps4.htm
===========================================
Troy and I fabricated the Stepper Power Supply from the various parts we've gathered and purchased. This consisted of:
-Antek Toroid (36VAC/18VAC/12VAC)
-Bridge Rectifier
-22,000 microFarad Cap
-7Amp Breaker.
-Old Pentium II Case
-Blade connectors
-12 Guage Wire (I think)
Troy was hoping to use a cap diode to condition the DC side of the power supply to cause a voltage drop to 36V. On our first test this resulted in a surprising catastrophic explosion of the diode. It was startling, and certainly opened our eyes a bit. We later realized this was because there was no load. Troy surmised to utilize that method would require a lot of resistors that could sap about 100Watts of power - not feasible. We also pondered Voltage Regulators - but they would require too many to properly work.
Forgive if my memory fails me for some of these stats... From the Transformer we were getting 36VAC, and through the Bridge rectifier we were getting 33VDC??. Through the capacitor we were getting 48VDC? This was too high. We measured the voltage in parallel to the capictor with the multimeter. We didn't completely think it through with the transformer because we thought the 36Volts would be DC not AC. Due to the nature of Vrms and the circuit we yielded the high 48VDC. Fortunately the transformer had a couple other power configurations. At 18VAC you get 24VDC through our circuit. So we undid our quick solders and switched over to blade connectors as Troy originally suggested as they are much easier to use. (Yup you were right Troy lol.)
This went off without a hitch, other than the scary part of discharging the huge cap. (as sparks fly) Something about it just feels wrong to drain the capacitor that way... (arcing it with a pair of insulated pliers.)
With the functioning 24VDC power supply we used that to test the connections on the control board. With the board layout properly verified with the multimeter Troy installed the remaining three driver chips to the PCB. I then pulled out an old Athlon computer I had located. We verified the potentiometer settings were correct with the power supply, control board, and multimeter. With the 2Amp rating for the steppers in unipolar mode that yields ~ 0.28VDC on the potentiometer. This stat was found on the specification that came from hobbycnc documentation. They were each set and the cap was drained. We waited several minutes and then refered to:
- http://www.kelinginc.net/KL23H286-20-08B.pdf
- http://www.cnczone.com/forums/showthread.php?t=61976
These were very helpful, and verified our interpretation of keling's documentation. Troy wrote a translation table between the controlboard and stepper documentation which I will take a picture of later. He wired one stepper up to the board after the capicator was less than 0.1Volts. After that, Troy had to leave. Bryan helped me out from then on. We installed Mach3, checked the pulse generation to be a bit shaky for the machine - it seems a bit slow. Sadly, the 750mhz computer has been tweaked for maximum performance already. I have it down to about 16 processes (this is very low.) We were ranging in the 25K to 15K PPS range.
We spent awhile configuring Mach3 attempting to jog the X-axis (the only stepper we had connected.) We figured out how to disable the E-stop break by switching the activelow signal. While this allowed us to green light the machine (make it ready,) we were unable to move the stepper. After a while, we decided to go buy food. Then we stood up and I noticed I hadn't plugged in the Stepper Power supply. LOL. We booted up again, and attempted to jog - no dice. Tired after a long day we got food and headed home.
I hope to figure out the issue by tommorow. I suspect software configuration is at fault.(and hope this too, I don't want to have to resolder the board.) I am certain patience, and persistence will get us through this hurdle.
The prospect of making an axis work is just too exciting! Pictures to come!!
We've received our CNC4PC board in the mail, it was delayed cause they were out of stock. It will control the spindle, it is controlled by an RJ45 port. (ethernet)
I've been working with Noobie on AirsoftRetreat to develop a new hopup solution. We've made considerable progress on that, I will release more information as we release it to the public. I made some prototypes and sent them to him for testing by USPS.
Troy tested a few of some old ATX powersupplies and used a jumper to quickly test the output of the powersupply. This website tells you how to jump it. Interesting...
http://www.duxcw.com/faq/ps/ps4.htm
===========================================
Troy and I fabricated the Stepper Power Supply from the various parts we've gathered and purchased. This consisted of:
-Antek Toroid (36VAC/18VAC/12VAC)
-Bridge Rectifier
-22,000 microFarad Cap
-7Amp Breaker.
-Old Pentium II Case
-Blade connectors
-12 Guage Wire (I think)
Troy was hoping to use a cap diode to condition the DC side of the power supply to cause a voltage drop to 36V. On our first test this resulted in a surprising catastrophic explosion of the diode. It was startling, and certainly opened our eyes a bit. We later realized this was because there was no load. Troy surmised to utilize that method would require a lot of resistors that could sap about 100Watts of power - not feasible. We also pondered Voltage Regulators - but they would require too many to properly work.
Forgive if my memory fails me for some of these stats... From the Transformer we were getting 36VAC, and through the Bridge rectifier we were getting 33VDC??. Through the capacitor we were getting 48VDC? This was too high. We measured the voltage in parallel to the capictor with the multimeter. We didn't completely think it through with the transformer because we thought the 36Volts would be DC not AC. Due to the nature of Vrms and the circuit we yielded the high 48VDC. Fortunately the transformer had a couple other power configurations. At 18VAC you get 24VDC through our circuit. So we undid our quick solders and switched over to blade connectors as Troy originally suggested as they are much easier to use. (Yup you were right Troy lol.)
This went off without a hitch, other than the scary part of discharging the huge cap. (as sparks fly) Something about it just feels wrong to drain the capacitor that way... (arcing it with a pair of insulated pliers.)
With the functioning 24VDC power supply we used that to test the connections on the control board. With the board layout properly verified with the multimeter Troy installed the remaining three driver chips to the PCB. I then pulled out an old Athlon computer I had located. We verified the potentiometer settings were correct with the power supply, control board, and multimeter. With the 2Amp rating for the steppers in unipolar mode that yields ~ 0.28VDC on the potentiometer. This stat was found on the specification that came from hobbycnc documentation. They were each set and the cap was drained. We waited several minutes and then refered to:
- http://www.kelinginc.net/KL23H286-20-08B.pdf
- http://www.cnczone.com/forums/showthread.php?t=61976
These were very helpful, and verified our interpretation of keling's documentation. Troy wrote a translation table between the controlboard and stepper documentation which I will take a picture of later. He wired one stepper up to the board after the capicator was less than 0.1Volts. After that, Troy had to leave. Bryan helped me out from then on. We installed Mach3, checked the pulse generation to be a bit shaky for the machine - it seems a bit slow. Sadly, the 750mhz computer has been tweaked for maximum performance already. I have it down to about 16 processes (this is very low.) We were ranging in the 25K to 15K PPS range.
We spent awhile configuring Mach3 attempting to jog the X-axis (the only stepper we had connected.) We figured out how to disable the E-stop break by switching the activelow signal. While this allowed us to green light the machine (make it ready,) we were unable to move the stepper. After a while, we decided to go buy food. Then we stood up and I noticed I hadn't plugged in the Stepper Power supply. LOL. We booted up again, and attempted to jog - no dice. Tired after a long day we got food and headed home.
I hope to figure out the issue by tommorow. I suspect software configuration is at fault.(and hope this too, I don't want to have to resolder the board.) I am certain patience, and persistence will get us through this hurdle.
The prospect of making an axis work is just too exciting! Pictures to come!!
Thursday, September 10, 2009
Quick post - tired
Worked on the CNC frame with Troy today, we made some real progress on the Z-axis. Learned a great deal from watching him model. Kept the tasks coming, as we were working efficiently. Troy couldn't start cutting his practice spindle because of the difficulties in cutting a section of the steel off to dimension. We probably will need to borrow Bryan's Chop saw for that task.
I also did some work on some airsoft parts today on the laser cutting machine. Various prototypes are being fabricated for bench comparisons of the different part geometries. (forgive my lack of proper english - i am tired.)
Lastly, I attempted to get the ball rolling on using the CNC machine again. While I was able to generate G-code and cut a part before the shop has changed some of the CNC's control software so we no longer have to remove G28's but now we need to change our post-processor in our CAM software to not use G43 commands. I made a post on cnczone to find a fix about this:
http://cnczone.com/forums/showthread.php?p=663482#post663482
I also did a bit of research on speeds and feeds for plastics here's what I got:
http://www.niagaracutter.com/techinfo/common_mat/polyprop.html
http://www.blurtit.com/q534090.html
I ended up choosing a speed of 766 RPM, with 7.5 IPM (~450 SFPM) with a 3/16 end mill. I'm going by memory for all those, so forgive me if there is a mistake. All that information was derived from the two websites above.
I also did some work on some airsoft parts today on the laser cutting machine. Various prototypes are being fabricated for bench comparisons of the different part geometries. (forgive my lack of proper english - i am tired.)
Lastly, I attempted to get the ball rolling on using the CNC machine again. While I was able to generate G-code and cut a part before the shop has changed some of the CNC's control software so we no longer have to remove G28's but now we need to change our post-processor in our CAM software to not use G43 commands. I made a post on cnczone to find a fix about this:
http://cnczone.com/forums/showthread.php?p=663482#post663482
I also did a bit of research on speeds and feeds for plastics here's what I got:
http://www.niagaracutter.com/techinfo/common_mat/polyprop.html
http://www.blurtit.com/q534090.html
I ended up choosing a speed of 766 RPM, with 7.5 IPM (~450 SFPM) with a 3/16 end mill. I'm going by memory for all those, so forgive me if there is a mistake. All that information was derived from the two websites above.
Saturday, September 5, 2009
Solidworks at the beach!
Oh yes, I went to the beach with the family this week, and spent some time just getting some old fashioned work done. I treated our deck with an oil, then after the chore sat down for some intimate time with Solidworks!
I sought out to roughly model up the frame for our CNC machine. The purpose for modeling the frame is so we are clear on our dimensions, plans, and specifically to ensure we buy the right ballscrews. I got it roughly halfway done. It was a lot of work switching between specifications looked up online, a slow solidworks computer, and the occasional realization that I would have to make my own assumptions in order to completely model it. This was a large amount of time and effort, here's some screenshots of what I was working on:
Despite all the effort i was only able to roughly map out the X and Y axis. But they move smoothly, and it was made with 75% actual dimensions! This sort of project makes you REALLY appreciate what parametric modeling does for you. I was able to get the design concept done despite a lack of a few dimensions. I turned over the files to Troy to see what he could do with the model. The Z-axis still needs to be done, and we are concerned about its deflection since the head would have to extend out 15" to reach the full working potential of our Y-axis.
Hopefully, we can order the ballscrews soon. I did receive the VXB spindle bearings from ebay last week. We are waiting on VFD esque controller Troy selected for our Treadmill motor, which essentially is a fancy potentiometer that we can control with the computer. I did a rough price estimate of the Bosch extrusions to be around 200$ with end machining - not bad. I imagine it will be a bit more by the time we are done. The mic6 plate will cost a good chunk too - i imagine like 200$. Oh well, so much is the price for extreme goodness!
I also spent some good time learning/grinding through some 3D milling as shown on the part above. Boy there is so much to learn! I feel almost overwhelmed with the magnitude of project potential that seems to consistently stare me in the face. I want to build the world! Alas, I am but one man. That won't keep me from trying though :P
I sought out to roughly model up the frame for our CNC machine. The purpose for modeling the frame is so we are clear on our dimensions, plans, and specifically to ensure we buy the right ballscrews. I got it roughly halfway done. It was a lot of work switching between specifications looked up online, a slow solidworks computer, and the occasional realization that I would have to make my own assumptions in order to completely model it. This was a large amount of time and effort, here's some screenshots of what I was working on:
Despite all the effort i was only able to roughly map out the X and Y axis. But they move smoothly, and it was made with 75% actual dimensions! This sort of project makes you REALLY appreciate what parametric modeling does for you. I was able to get the design concept done despite a lack of a few dimensions. I turned over the files to Troy to see what he could do with the model. The Z-axis still needs to be done, and we are concerned about its deflection since the head would have to extend out 15" to reach the full working potential of our Y-axis.
Hopefully, we can order the ballscrews soon. I did receive the VXB spindle bearings from ebay last week. We are waiting on VFD esque controller Troy selected for our Treadmill motor, which essentially is a fancy potentiometer that we can control with the computer. I did a rough price estimate of the Bosch extrusions to be around 200$ with end machining - not bad. I imagine it will be a bit more by the time we are done. The mic6 plate will cost a good chunk too - i imagine like 200$. Oh well, so much is the price for extreme goodness!
I also spent some good time learning/grinding through some 3D milling as shown on the part above. Boy there is so much to learn! I feel almost overwhelmed with the magnitude of project potential that seems to consistently stare me in the face. I want to build the world! Alas, I am but one man. That won't keep me from trying though :P
Saturday, August 29, 2009
Powersupply preparations
After looking over the components, documentation, and some a gecko document. I've been able to determine that we needed only 4666 micro Farads. This was determined with the following equation:
C = (80,000 * I)/V
C = (80000 * 2.1 Amps) / 36Volts = 4666 mF
Since we are running our steppers in unipolar mode the steppers will draw 2.1 Amps. Bipolar i believe was 2.8 Amps, so we calculated it with 3 Amps and that necessitates a 6,666 mF capacitor. The hobbycnc documentation recommends using a capacitor that is 20% higher than what is requested. From these calculations we see that our capacitor is more than sufficient.
This was not the 27,000 mF that was previously thought. Also from inspection the capacitor Troy purchased is rated for 22,000 mF and 50V. So we are good! We have almost all the components we want for our power supply. We would have built it tonight, but I believe we want some fuses, power switch, and wire. We also need to design and fabricate a power supply box.
I've made some progress on the frame, i'm now toiling with some geometry issues with the frame, and trying to ensure frame stability. The issues are occuring with mounting the THK rails to T-slot bars that will inevitably have to be drilled.
http://buildyourcnc.com/supply1.aspx
http://www.cnczone.com/forums/showthread.php?t=61976
http://www.geckodrive.com/upload/Step_motor_basics.pdf (page 11)
http://www.5bears.com/cnc03.htm
C = (80,000 * I)/V
C = (80000 * 2.1 Amps) / 36Volts = 4666 mF
Since we are running our steppers in unipolar mode the steppers will draw 2.1 Amps. Bipolar i believe was 2.8 Amps, so we calculated it with 3 Amps and that necessitates a 6,666 mF capacitor. The hobbycnc documentation recommends using a capacitor that is 20% higher than what is requested. From these calculations we see that our capacitor is more than sufficient.
This was not the 27,000 mF that was previously thought. Also from inspection the capacitor Troy purchased is rated for 22,000 mF and 50V. So we are good! We have almost all the components we want for our power supply. We would have built it tonight, but I believe we want some fuses, power switch, and wire. We also need to design and fabricate a power supply box.
I've made some progress on the frame, i'm now toiling with some geometry issues with the frame, and trying to ensure frame stability. The issues are occuring with mounting the THK rails to T-slot bars that will inevitably have to be drilled.
http://buildyourcnc.com/supply1.aspx
http://www.cnczone.com/forums/showthread.php?t=61976
http://www.geckodrive.com/upload/Step_motor_basics.pdf (page 11)
http://www.5bears.com/cnc03.htm
More purchases, and more to come
I've purchased the spindle bearings for Troy's spindle design. He ran some calculations on the angular contact bearings, and also found some ebay bearings that had an ABEC-3 rating. I also purchased a CNC4PC control board that will be used to interface and control the DC motor for the spindle.
Short of 1 Capacitor, I think we have all the major electrical components. I've begun frame design once again starting from scratch, as I've learned some assumptions were not correct that I was working with before. I am particularly concerned with of each axis, as I found in my research that it is difficult to find swarf guarding material of the appropriate size. I want to completely model the frame before making the ballscrew purchase, as we know where to do that. Although, i'm feeling the pinch of time. So i'm thinking I should get that done within the next week.
As I keep track of the fiscal performance of our project. We are in the clear so far, I hope we can stay on track. If anything we are slightly under our projected cost.
In other developments, I reassembled my M4 airsoft rifle, after Troy and I inspected the internals for future projects. You can see some points in the procedure from the pictures. It was actually a lot more complicated to re-assemble than the picture's show. With the anti-reversal latch, and some springs that just didn't want to stay put, I had to come up with some clever ways to hold them in place. Using my Allen wrenches, and some electrical tape, I was able to fix the components in place until I could close the mechbox and remove them. I also had to use the motor's magnets to magnetize my screwdriver, so I could manuever the screws into the pistol grip successfully.
I also setup and started up an old computer, that had problems in the past. I confirmed that the primary harddrive has some bad sectors, and cannot boot to windows. I used a bootable linux OS to take a gander at the drives. The Raid-0 array was fine, all the data on those drives were unscathed. Ironically, it was the single OS hard drive that failed. I was pleased to find no faults in any of the other hardware.
Tuesday, August 25, 2009
Gathering resources
As I keep my projects moving along I realize that one of the major aspects of projects that can easily be underestimated, is simply the selection and purchase of a given raw material or tool.
Several educated decisions must be made before making a purchase in the hopes of minimizing lost money or time. Today, to keep the ball rolling on one of my airsoft projects I had to make a quick decision about a ball-end mill purchase. As I was modeling my CAM file for the part, I had to choose between a 3/16" diameter or 1/4" diameter. I picked 3/16" because I can interpolate curves with the 3/16" that should be near to similar geometry of the 1/4" ball end mill. A quick run over to Hall tool, and I got the tools. I also purchased a VERY small double sided endmill (1/16"), for some detailed work.
I also received the Toroid Transformer the mail from Antek Inc. It looks like in good condition to me. It is a surprisingly heavy piece of equipment. I need to locate 1 more medium sized capacitor to run in parallel, and we will have all the components for our powersupply. If I can get all these things to Grant, we should be able to fabricate, and finish off the control board. Basically - we will be able to test the brains of the CNC machine. This will let us also test and configure the stepper motors. I'm sure if we can get the board, and the steppers running, the project will quickly accelerate.
There's always a defining moment in a project, where you find out if it is inevitably going to work. I believe we have one coming of those moments coming up with the power supply, control board, and steppers. I mean if we get that done, we essentially need to attach a few more mechanical components, ballscrews, THK rails, and spindle and we have ourselves a CNC machine!
Several educated decisions must be made before making a purchase in the hopes of minimizing lost money or time. Today, to keep the ball rolling on one of my airsoft projects I had to make a quick decision about a ball-end mill purchase. As I was modeling my CAM file for the part, I had to choose between a 3/16" diameter or 1/4" diameter. I picked 3/16" because I can interpolate curves with the 3/16" that should be near to similar geometry of the 1/4" ball end mill. A quick run over to Hall tool, and I got the tools. I also purchased a VERY small double sided endmill (1/16"), for some detailed work.
I also received the Toroid Transformer the mail from Antek Inc. It looks like in good condition to me. It is a surprisingly heavy piece of equipment. I need to locate 1 more medium sized capacitor to run in parallel, and we will have all the components for our powersupply. If I can get all these things to Grant, we should be able to fabricate, and finish off the control board. Basically - we will be able to test the brains of the CNC machine. This will let us also test and configure the stepper motors. I'm sure if we can get the board, and the steppers running, the project will quickly accelerate.
There's always a defining moment in a project, where you find out if it is inevitably going to work. I believe we have one coming of those moments coming up with the power supply, control board, and steppers. I mean if we get that done, we essentially need to attach a few more mechanical components, ballscrews, THK rails, and spindle and we have ourselves a CNC machine!
Subscribe to:
Posts (Atom)
