Build log-drhanger's DM4400 electronics retrofit

Page 1 of 5 1234 ... LastLast
Results 1 to 20 of 97

Thread: Build log-drhanger's DM4400 electronics retrofit

  1. #1
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Build log-drhanger's DM4400 electronics retrofit

    Hey all--

    I bought a couple of the DM440's over the last holiday, so thought I would post a build log, even though I'm partially into it already. I got them from a shop near where I work in the Phoenix area--here's a couple pix, getting loaded to move to the rigger's yard temporarily:

    Build log-drhanger's DM4400 electronics retrofit-2016-12-27-12-19-09-jpg Build log-drhanger's DM4400 electronics retrofit-2016-12-27-12-19-40-jpg

    I left one in storage with the rigger and had the other moved to my garage another 35 miles away--here she sits:

    Build log-drhanger's DM4400 electronics retrofit-2016-12-29-13-40-20-jpg

    Since that picture I've done some cleaning up. Many people have already extolled the virtues of these machines, so I won't say a lot, except wow, what a cool little machine. It takes up less than a quarter of my 2 car garage, but it's built like a brick outhouse, a full on industrial cnc in a nice compact package. Needs a lot of work, though, and when I found these I had no idea about the SKIP controller. If I had known beforehand I might have passed them up, but by the time I realized I was already charmed by them.

    One of the first things I did was start hunting down as much information as I could about these--got on CNCZone and quickly discovered this Dyna forum and read through all the posts. I already knew that the one I had in the garage had an oiler problem, and the forum quickly helped me find out the culprit. I removed the covers over the Y ballscrew and found the disintegrated lube line that everyone else with this machine discovered. That was an easy task--I used the technique 'olfoundryman' did (I think it was him) and wrapped the tubing around a mandrel and heated it with an air gun for a few minutes, which gave it a nice spiral to go around the ballscrew like the original. Worked out nice. I didn't take any pix because I hadn't thought of a build log at that time, but there's plenty of similar pics on this forum, nothing unusual about mine.

    The next thing was to get it up and running and check out all the operations to see what else might need to be done. I only had single phase 240V but I went ahead and connected two of the transformer coils and got some life out of it. At the time I assumed that my only option was to eventually get three phase, but thanks to a couple other members (Jim Dawson and Marty Escarcega) I discovered that wasn't so--everything on the machine except the spindle amp used single phase, and even the amp would run the spindle motor on single phase, but I could not get the spindle to initialize even though everything else would, including the ATC (yes!). There's a lengthy thread (http://www.cnczone.com/forums/dyna-m...converter.html) that goes into a lot of the details, but it came down to a ground fault between some component and the frame. I finally tracked it down to the coolant pump--once I left that unplugged the spindle amp finally stopped showing a ground fault and the spindle would initialize. After my initial excitement of getting the last item running the excitement was quickly tempered by really loud spindle bearings. So another item on the to-do list.

    So what am I gonna do with these? Well, it wasn't originally in my mind to do a retrofit, but after I discovered the SKIP controller and trying my best to get a handle on it (I've used conventional G code for too long now) it's turned into a retrofit project. Apart from that, I have an amateur astronomy (my other hobby) product I've developed and hope to eventually commercialize, and most of the parts would fit nice on this machine. I had a cnc router I had built myself a few years back and I was able to prototype a lot of the parts on it (acetal and aluminum) but it really wasn't up to any kind of production runs, and of course the accuracy of a home built isn't going to come close to a real metal cutting machine.

    So down to brass tacks (whatever that means, but it sounds nice). I've already stripped out all the electronics from the cabinet and started doing some research into what I would need. I've divided the project into three main areas in my mind, in the following order:

    1. Axis motion

    2. Spindle operation

    3. ATC operation

    The first item I feel pretty comfortable with. Since my experience with building the router I've learned a lot about stepper motors and drivers and limit/homing switches, so I'm pretty comfortable with that part. I plan to use the existing steppers for now since they seemed to be in fine working order, and I'm not so desperate for really fast rapids since it doesn't have far to go. I've already ordered a Gecko G540 which is due to arrive tomorrow (Saturday) so my weekend is going to be taken up with that. I plan to use the existing power supplies from the machine, which seemed to be in good working order. I haven't decided on what controller software to use, but for my router I used LinuxCNC so I will use that to get started until (if) I run into any insurmountable snags with it. I'm thinking of getting a full motion controller of some sort, and some additional I/O board(s), but for now this will get the machine moving. I found an old PC with a parallel port and fired it up, and discovered I already had LinuxCNC on it. I finally recalled that there was some issue I was having with that machine and had replaced it with another, but I don't recall what the problem was. I guess I'll soon find out and may have to find another PC--I just sold the router and included the working Linux box with it.

    The other two items after axis motion are going to be daunting for me. As evidenced by the other thread I'm no expert on electrical/electronics matters, so it's going to be a tough row to hoe for me. But, I love to learn new things, so I feel pretty good about forging ahead with it. As far as the ATC is concerned, even if I don't get the full system operating (I hope it doesn't come to that, what a waste of one of the nicest features on it) I need to at least get the 'presser bar' working in order to change tools manually. That's the AC reversing motor/worm gear/presser that pushes on the drawbar at the top, with two sensors to track the presser bar. One thing I don't like about it is you have to use two separate buttons to clamp and unclamp, and I have to make an effort to make sure I don't leave it in the unclamp position and pressing on the drawbar washers. It would be nice to re-work it to use one button like most machines I've used so that the presser moves back after releasing the button.

    I think that's enough to get started--I'll report on my progress with the Gecko soon I hope.

    Similar Threads:
    Last edited by drhanger; 02-03-2017 at 08:53 AM. Reason: fix broken picture links


  2. #2
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Yesterday I spent most of the day prepping the power supplies for when the Gecko driver arrived. After cleaning out the cabinet of all the old electronics (I left some of the old cables, like the axis cables and the ones for the ATC mechanisms since I don't know yet how to handle that) I re-installed the shelf that the power supplies had been on and started doing some wiring for it. I had also left some of the old terminal strips and the conduit/wire loom and took advantage of those. I took an old construction grade heavy gauge extension cord and removed the socket end, then stripped the ends to attach to one of the terminal strips as a take off for main power supply. I then routed the original cable from the smaller PSU back to the main terminal strip and connected everything up, plugged in the main, and both PSU's fired up.

    One thing about the smaller power supply--the labeling calls for 170-250VAC as input. I wanted to get away from using 240V so I decided to use 120 on it and see what happened. I don't know if it will ultimately damage the PSU, but I checked the outputs--the 5VDC read a steady 5.02V and the 12VDC read a steady 11.98V. The 48V PSU had a pair of terminals that you could short to convert it to the lower input power so that wasn't a problem. After I got them working I wired up the two cabinet door fans and all three axis motor fans to the 12VDC supply (using the old terminal strips again) and plugged in the main power. All 5 fans came to life just fine, and with those all running I checked the voltage at the small power supply terminal--it was 11.95, barely moved from the unloaded state. Well, if anyone can tell me if this is not good for the small PSU (being fed a lower incoming voltage than called for) I will have to come up with an alternative, but so far so good.

    This took me most of the morning, but the Gecko still hadn't arrived so I concentrated on setting up LinuxCNC and ran into a problem with it right away. The problem is not Linux, but the PC box itself--I don't know if this is why I had abandoned it earlier but possibly. LinuxCNC includes a latency test that checks for any jitter in the system that would affect the pulse timing. Even though the version of Linux that's paired with LinuxCNC is a real-time OS, apparently there are things that go on in the bowels of any given PC that can still cause latency, so the utility runs for a while and checks for what the minimum time between pulses is needed in order to account for the jitter and not have any lost pulses to the motors. The jitter was so high that the highest speed I'll be able to get from the motors is about 36 IPM, and that's not nearly good enough for me--it's not even near the original 100 IPM the machine was capable of. So I will definitely have to do something else--either find another PC with better performance, or go with an external motion controller, which I suspect will be the best alternative. The reason I'm not sure this was the same problem I had with the PC way back when is because with my router I was using 1/4 microstepping on the drivers, and the Gecko uses 1/10 microstepping. The number of pulses needed were only 40% of what the Gecko needs, so that alone would be enough to speed up the motors considerably. I think there was some other problem that I had when using it for the router, like memory issues, which only highlight further how bad this PC will be.

    The G540 finally arrived about 3 in the afternoon, and for a product that has such a high regard in the CNC community I'm a bit underwhelmed by the fact that you're required to do a complete Dsub connector assembly for each motor. I don't know what's wrong with using screw terminals like just about everyone else. Sure, the Dsub connector (if done properly) is a really nice way to connect and disconnect the motors, but how often does a person need to do that? Well, in spite of not being an electronics expert I'm really quite good at soldering, so I expect to be able to make a good, secure connection. I decided to call it a day and tackle it the next day (today), but I'm not sure I'll be able to get to it today. Being a dyed in the wool nerd I have no interest in the Super Bowl, but I do have other chores to attend to (like getting a new TV after our 12 year old Sony gave up--maybe I'll get a good deal on SB Sunday).

    I will attach a couple of update photos, even though there's really nothing of interest to see, but maybe it will be interesting to someone else.

    Until the next time.....
    Dan

    Attached Thumbnails Attached Thumbnails Build log-drhanger's DM4400 electronics retrofit-img_20170204_100035-jpg   Build log-drhanger's DM4400 electronics retrofit-img_20170204_100324-jpg  


  3. #3
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Dan, I wish you nothing but the best of luck.

    How are you going to handle the extra I/O? The G540 via the parallel port doesn't have enough.

    I would have suggested trying the Clearpath SD brushless DC motors from Teknic, but good NEMA 34 Steppers sized appropriately for torque and your voltage should do nicely because of the gear reduction from the motor to the ball screws.

    If you retain the original spindle amplifier, you will need a +/- 0-10VDC signal to drive it forward and reverse. Most controls only give 0-10VDC for a VFD. Maybe you can find a tall skinny motor like was used on the late DM4400M and just go with VFD. You will need to instantly brake the spindle for spindle orient.

    Did you happen to take the time to take voltages of the ATC motor and the electro mechanical drawbar? Their control boards are a bit complex, its why I abandoned them and went with traditional air cylinder on the ATC and air over hydraulic original power drawbar from the later model machines. I supposed you could machine up a mount and do a more traditional pneumatic power draw bar.



  4. #4
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Hi Marty--

    Sorry I haven't responded yet, been pretty busy. I'll get back by the weekend.

    Dan



  5. #5
    Community Moderator Jim Dawson's Avatar
    Join Date
    Dec 2013
    Posts
    5716
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Lookin good Dan!



  6. #6
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    I've had an epic battle with LinuxCNC through 3 computers, 2 different distros of Linux, and several pieces of hardware. Finally won this morning and have all 3 axes in motion with home/limit switches. But, I will detail that in a later post. I want to get to Marty's questions first.

    Quote Originally Posted by Marty_Escarcega View Post
    Dan, I wish you nothing but the best of luck.

    How are you going to handle the extra I/O? The G540 via the parallel port doesn't have enough.

    I would have suggested trying the Clearpath SD brushless DC motors from Teknic, but good NEMA 34 Steppers sized appropriately for torque and your voltage should do nicely because of the gear reduction from the motor to the ball screws.

    If you retain the original spindle amplifier, you will need a +/- 0-10VDC signal to drive it forward and reverse. Most controls only give 0-10VDC for a VFD. Maybe you can find a tall skinny motor like was used on the late DM4400M and just go with VFD. You will need to instantly brake the spindle for spindle orient.

    Did you happen to take the time to take voltages of the ATC motor and the electro mechanical drawbar? Their control boards are a bit complex, its why I abandoned them and went with traditional air cylinder on the ATC and air over hydraulic original power drawbar from the later model machines. I supposed you could machine up a mount and do a more traditional pneumatic power draw bar.
    You are correct, the G540 does not have nearly enough I/O's to get everything done, and there are other things about the G540 I'm not thrilled about, but more on that in the later post. After the Linux battle I'm more convinced than ever that I want to go with an external controller with additional I/O's. I've been looking at UCCNC as opposed to Mach 3/4; I hadn't heard of it until talking with Arturo at CNC4PC, but some videos on Youtube seem to be saying pretty good things about it. So far it only works with their hardware, but that's not necessarily a bad thing. I'm considering their ethernet stepper with motherboard which come with a large number of I/O's and support up to 100 kHz pulsing, which should be more than adequate I would think.

    I don't know whether that hardware can play a supporting role for the spindle motor; if you have time to check it out maybe you can inform me about that. I have decided to not use the SD3060 servo amp--I would rather go with new hardware as much as possible, but as I said before the spindle and the ATC are big hurdles for me in terms of understanding how everything works together and needs to be designed. Arturo recommended another of CNCDrive's products, one of their servo amps. As for the +/- 10VDC signal, I understand that's a problem, with Arturo mentioning using relays to reverse the signal, but he didn't go into detail about that and suggested there was some danger to that. I know this very issue was discussed in some detail in one of the older posts on this forum, ckm's very informative thread on this machine. Post 17 begins some of the discussion about it, and the idea of using relays for reversal is brought up there. Arturo also suggested a C6 which sounds as though it does just that, provides a +10V signal as well as relays to reverse it. When I asked Arturo for more information about it I didn't really understand his response, so any insight on that would be helpful.

    Ever since you first told me about the Clearpath motors I've thought about them and how cool they are, but at this point I can't really justify the cost. This machine is not what I would call 'mission critical' right now, and I'll be satisfied with what I have for the moment. For now I can only get a top speed of 48 ipm, but that's mainly because of the high latency of the PC I finally got working and the 10 microstepping Gecko, which is not configurable (another of my issues with it).

    As for the ATC, that also is where I want to get rid of all the old hardware and go with something like you suggest. In fact, Timothy (2ferrous) sent me some pix and a teaser about how he went about it and sounds similar. He said he would post the info here, but he's been busy with spindle motor problems and hasn't had time to get to it. I'm looking forward to getting ideas for that.

    Okay, on to the latest project updates and how I got the axes configured and moving (omitting most of the battle sequences).



  7. #7
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    After getting the G540 it was time to set about wiring the motors and soldering up the connectors. There didn't seem to be any good wiring diagrams for the motors so I found a page that details how to determine the phases on an 8 wire motor. Timothy (2ferrous) had posted a motor wiring diagram earlier, but that diagram didn't match my motors as his diagram had some striped wires, while mine were all solid colors. Anyway, after going through the procedure this is the setup I came up with (using the author's convention of naming the coils P Q R and S):

    PHASE A:

    GREEN - P1
    ORANGE - P2
    BLUE - Q1
    RED - Q2

    PHASE B:

    BLACK - R1
    GRAY - R2
    YELLOW - S1
    WHITE - S2

    Deciding that I was more interested in lower speed torque I chose to wire them in series, so my wiring was like this:

    GREEN AND RED TO PHASE A
    ORANGE AND BLUE CONNECTED
    BLACK AND WHITE TO PHASE B
    GRAY AND YELLOW CONNECTED

    Series current was rated at 2.3A and parallel at 4.6A. Since the G540 was limited to 3.5A the series wiring was a good match for that as well, and I started by soldering in a 2.2K resistor for the current setting on the Gecko Dsub connectors. I had left the motor cables in place so I just cut the ends off inside the cabinet and soldered and connected the cables to the Dsub connectors and plugged in. Later on, after getting the motors going I discovered that by luck I had gotten all the phases in agreement with LinuxCNC as I didn't have to reverse the phases on any of them, they all went the right direction from the start.

    I decided for now that I would just use the NO limit switches (only one switch per axis) as homing/limit switches and rely on software limits for a while. I know that's not the ideal, but it's just for now until I have time to get some details done--just want to get in operation right now. The cables for the limit switch and motor encoder were bundled together, so I had to determine which two wires were for the limit switch (black/white) and ignore the rest. Once I determined that I rigged up a simple perf board circuit and connected everything up. After a bit of tinkering with Linux I got everything in motion with the homing working perfectly. Very happy with the results so far, except for the 48 ipm I mentioned last post.

    I'm thinking of returning the Gecko and going with something different for stepper control, maybe something like this. This adds higher amperage and supply voltage than the Gecko, but the tradeoff is one less axis, as the Gecko has 4 drivers in it. It has similar I/O's to the Gecko, and adds more configurability with the microstepping and amperage, and easier connections. I wasn't very happy with the 1/10 microstepping in the Gecko and not being able to reduce it. I was used to using 1/4 in my router and it seemed to work fine, and it reduces the pulsing rate considerably, which I may need. The CNCDrive ethernet stepper I mentioned has a top rate of 100 kHz, which I said earlier ought to be adequate, but I may find myself bumping into that ceiling without lowering the microstepping. I suppose the Gecko is a bit smoother, but it may not be different enough to matter. This driver also gives me the option of going with more powerful steppers in the future if I find I need it--60V as opposed to 48V with the current motors, and current of 6A instead of 3.5A with the Gecko.



  8. #8
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    If you are going to run Linux, why don't you use the Mesa boards? Mesa Electronics
    Same boards that run the Tormach and its tool changer. You are simply treating the ATC as an axis so that's easy.

    I think you need to figure out the power drawbar and making it work. Then you can integrate it. That's one hurdle. OR figuring out the throw and building a pnuematic power drawbar.

    Yes, you can use the DUOGONG drives from Arturo if its big enough. I will be building some of those boards that with a reverse signal input will invert the 0-10VDC signal to -0-10VDC, so you can reverse. Honestly, I would suggest seeing if you can replace the motor with a new C face mount 3 phase motor and a vector drive VFD and move away from the DC Brush motor & drive and all its troubles. You can reverse the vfd easily. Look hard at this option. If you decide to part with the motor and the spindle amp, let me know. I could probably use them as spares for my machine.

    If you resolve the power drawbar and spindle, you have a usable machine until you sort out the ATC. That is the route I would go.

    Oh, these are really nice for DB9 connections for a G540: EZ-G540 - Set of 4 - Soigeneris.com



  9. #9
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by Marty_Escarcega View Post
    If you are going to run Linux, why don't you use the Mesa boards? Mesa Electronics
    Same boards that run the Tormach and its tool changer. You are simply treating the ATC as an axis so that's easy.
    I don't think I really want to stick with Linux. I had really good success for several years with the computer I had attached to the router, but that's gone now and I went through quite a lot of hurdles getting another one to cooperate (that has a parallel port). I haven't looked much at Mesa but I would assume they have something that works without a parallel port, like an external motion controller? I know almost nothing about Linux OS and I don't really want to try--not a priority for me, although once it's going I don't need to know that much. It's something to think about though.

    I think you need to figure out the power drawbar and making it work. Then you can integrate it. That's one hurdle. OR figuring out the throw and building a pnuematic power drawbar.
    I think getting rid of the current presser bar and associated hardware is probably the best thing. I'm pretty sure Timothy's setup was right along the line of the pneumatic solution.

    Yes, you can use the DUOGONG drives from Arturo if its big enough. I will be building some of those boards that with a reverse signal input will invert the 0-10VDC signal to -0-10VDC, so you can reverse. Honestly, I would suggest seeing if you can replace the motor with a new C face mount 3 phase motor and a vector drive VFD and move away from the DC Brush motor & drive and all its troubles. You can reverse the vfd easily. Look hard at this option. If you decide to part with the motor and the spindle amp, let me know. I could probably use them as spares for my machine.
    I'd like to know more about the boards you're building.

    I'm having a hard time grasping the difficulty of the +/-10V issue--is using the servo motor for a spindle motor really that different from using it for an axis? I mean, they both have to change directions so it doesn't make sense that it's such a high hurdle to make it work. Well, this is where my ignorance really shines, so to speak. Changing motor types may be better, I don't know, but I can see issues with adapting another motor what with the motor mount and changing the pulley over. Plus, the current motor has a brake built in which I would need for the ATC, right? And, I don't really want to go back to 3 phase either.

    If you resolve the power drawbar and spindle, you have a usable machine until you sort out the ATC. That is the route I would go.
    My thinking exactly, at least get to the point of being able to do manual tool changes, then go from there.

    Oh, these are really nice for DB9 connections for a G540: EZ-G540 - Set of 4 - Soigeneris.com
    Thanks, but it's a moot point now. I've already got them all done, and on top of that I've decided to go ahead with exchanging drivers. Something to remember for future, though.

    Dan



  10. #10
    Community Moderator Jim Dawson's Avatar
    Join Date
    Dec 2013
    Posts
    5716
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by drhanger View Post
    I'm having a hard time grasping the difficulty of the +/-10V issue--is using the servo motor for a spindle motor really that different from using it for an axis? I mean, they both have to change directions so it doesn't make sense that it's such a high hurdle to make it work. Well, this is where my ignorance really shines, so to speak. Changing motor types may be better, I don't know, but I can see issues with adapting another motor what with the motor mount and changing the pulley over. Plus, the current motor has a brake built in which I would need for the ATC, right? And, I don't really want to go back to 3 phase either.

    Dan
    Controlling the spindle with a +/- 10V signal is exactly like controlling an axis, in fact it becomes an axis in that case. But in order to apply the signal correctly, you need a motion controller providing the signal. I can't imagine that a 0-10V signal and a relay would well at all, but maybe I'm not understanding how it works.



  11. #11
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Here's part of what I got from Arturo--maybe this can shed some light on a direction to go with the spindle motor. I don't really understand it, so maybe one of you guys can interpret it for me:

    Dan,



    OK, awesome when you have all the info. OK, that servo can take +/-15vdc. We could generate that with an external C6 board connected to the C62. Note that to generate the negative values, we would need to use additional relays to swap the polarity of the signals. While all is doable, it is complex and may be a little dangerous. But I did notice that drive for the spindle can easily be replaced by a DG4S-16035: <https://cnc4pc.com/dg4s-16035-dc-servo-drive.html> https://cnc4pc.com/dg4s-16035-dc-servo-drive.html . You may even be able to use the encoder currently mounted on the motor. Note that with this setup, you can use the spindle as a true servo, and program threading operations and even do true rigid taping. I am not sure how much of this is currently been supported by the motion controllers, but it is all moving in that direction. That is my suggestion. You may want to check if the current servo has the power supply built-in, or if it uses and external power supply/transformer that you can use. If not, you can source one from Antek.



    The UCx00 controller currently are only compatible with UCCNC, but Mach4 plugins for them are in the works are about to be released, so they should work with Mach4 or UCCNC. The ESS and PoKeys will only work with Mach3/4.



    Yes, with the C62 you can control the spindle on any of the systems. If you go with the DG4S for the spindle, you will connect the drive to the connector for the 6th axis and not use the connection for the spindle.



    Arturo Duncan




  12. #12
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by Jim Dawson View Post
    Controlling the spindle with a +/- 10V signal is exactly like controlling an axis, in fact it becomes an axis in that case. But in order to apply the signal correctly, you need a motion controller providing the signal. I can't imagine that a 0-10V signal and a relay would well at all, but maybe I'm not understanding how it works.

    Jim,
    You are correct, The spindle motor is a SEM Servo motor, It contains a Tachometer. It has NO encoder, it also contains a 110VAC brake. The Servo Dynamics SD3060 is just a large servo amplifier. I tried to reverse the signal polarity with a relay. It did not work as I recall. With a help of another user, we came up with a simple circuit that takes the standard 0-10VFD signal and the motor reverse logic input, an LM741 Op amp and a -12/+12V power supply to give the proper signal to the SD3060, that worked. Then I added logic with to take the spindle orient input ANDED with the spindle orient signal, along with a fixed voltage and a small relay to turn the spindle and fixed RPM until the spindle orient signal went low and then BRAKED the spindle. Then continued with the tool change process. This to 4 revisions of my control board. I had to add the spindle orient logic because my flashcut signal generator was not fast enough to respond to the orient signal and brake the spindle consistently. Fixing the voltage during spindle orient, did this. Spindle orient, and the tool change process under control is a fairly complex process. The original Dyna control also had a spindle board that seemed to do this very thing.

    I'm not sure how Timothy handle the ATC.

    After doing that work, my ATC has worked very consistently.

    Many have successfully converted the DM4400's but few have gotten the tool changers to also work reliably that I know of. Its easy to move the axis around. The subsystems that complicate the conversion are:
    Spindle control
    Electromechanical ATC
    Electromechanical drawbar.

    I should note that it took almost a year of weekends consistently working on my machine to get it working consistently. I chose NOT to use Mach, I think Linux would have been a better choice, but the Flashcut system worked wonderfully and their support team is top notch. Their hardware is not cheap. The support is the reason for that. Don't get me wrong, Mach has its place and I think was a contributor to starting the DIY CNC Revolution.

    Marty



  13. #13
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by drhanger View Post
    I don't think I really want to stick with Linux. I had really good success for several years with the computer I had attached to the router, but that's gone now and I went through quite a lot of hurdles getting another one to cooperate (that has a parallel port). I haven't looked much at Mesa but I would assume they have something that works without a parallel port, like an external motion controller? I know almost nothing about Linux OS and I don't really want to try--not a priority for me, although once it's going I don't need to know that much. It's something to think about though.



    I think getting rid of the current presser bar and associated hardware is probably the best thing. I'm pretty sure Timothy's setup was right along the line of the pneumatic solution.



    I'd like to know more about the boards you're building.

    I'm having a hard time grasping the difficulty of the +/-10V issue--is using the servo motor for a spindle motor really that different from using it for an axis? I mean, they both have to change directions so it doesn't make sense that it's such a high hurdle to make it work. Well, this is where my ignorance really shines, so to speak. Changing motor types may be better, I don't know, but I can see issues with adapting another motor what with the motor mount and changing the pulley over. Plus, the current motor has a brake built in which I would need for the ATC, right? And, I don't really want to go back to 3 phase either.



    My thinking exactly, at least get to the point of being able to do manual tool changes, then go from there.



    Thanks, but it's a moot point now. I've already got them all done, and on top of that I've decided to go ahead with exchanging drivers. Something to remember for future, though.

    Dan
    I explained the heart of the spindle control circuit to Jim.

    It sounds to me as if you are going the Mach3 route? I'm personally done with Mach3 and its quirks...and no more factory support.

    If memory serves, I just took a 1.5V battery and connected it to the analog input of the SD3060 and then flipped it. I think while it worked well, doing the same with the VFD signal output and a relay did not at all. I think it had something to do with all the power supply grounds being connected together. So the circuit was born. VFD's can brake a motor, (this sort of setup is used on most all new CNC mills) but it HAS to be timed precisely. Therein lies the challenge.

    So, are you running Mach3 on a parallel port computer and feeding that signal to a G540?
    Look at the POKEYS stuff? They could provide you with more I/O. Oh and per my advice during our visit, don't neglect to connect up the ATC crash/tilt sensors. During testing, I am very grateful I had them. If you stayed with Steppers, they might just stall out, servos aren't so forgiving. (As you know, I went with DC Brush servos)

    Good luck on your continued progress....

    Marty



  14. #14
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by drhanger View Post
    Here's part of what I got from Arturo--maybe this can shed some light on a direction to go with the spindle motor. I don't really understand it, so maybe one of you guys can interpret it for me:
    This may work provided you add an encoder to the spindle motor. This is doable. If you pull the tachometer off the shaft, then measure it (I would ask Timothy for the size of the shaft where the encoder armature mounts, I have not had my motor apart). Some motors had both the tach and an encoder. There may be a stub on it for an encoder and you just need a mounting plate. I would ask Tim to see whether or not after removing the Tach armature, the shaft is smooth, what size and whether or not it can accept an encoder.

    It may work because it takes step and direction signals directly. The SD3060 provides 30amps continuous current with 60amps peak while the Duogong will max at 35A. You will likely never push it this hard so the amp will survive.
    Price is pretty good too. One thing, when you rectify 120VAC you get 169VDC which is over the nameplate. You'd better ask Arturo about this. Why they didn't design for 180VDC is beyond me. (probably cost) Now if your AC voltage was constant 110 (yeah right) then no problem.....

    This one similar but better aligned with the original specs: Viper 250 H (230Vdc 30A), ViperServo
    .
    However you would need at least a 5 axis control, 6 axis drive if there was a 4th axis in your future. Wouldn't work for me....Flashcut only has 5 axis output and I have a 4th axis.
    X, Y, Z, ATC, Spindle, Future 4th Axis


    Marty



  15. #15
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    You know, to really make things easy, you could just use the VFD output and run the spindle forward and if you ever want to tap, tapmatic head...this would get you going even quicker until you sort things out.

    Hand tool change and just run the spindle in forward....you aren't running production, you just want to learn and make a few parts. As time and money permits, you can work on the other machine and sort things out, perfect and then convert the first machine. You have 2 machines.....what is your end goal? Get a CNC capable CNC machine up and running quickly do you can make your prototype parts right??

    Then its just a matter of sorting out your power drawbar....



  16. #16
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by Marty_Escarcega View Post

    It sounds to me as if you are going the Mach3 route? I'm personally done with Mach3 and its quirks...and no more factory support.
    I'm not necessarily going with Mach, but I'm not ruling it out either. As far as support goes, there are many years of community support with both Mach and LinuxCNC, but with Mach I get the privilege of paying for it I am also looking at UCCNC, which seems to be a pretty good competitor to Mach, and a lot cheaper. I've had trouble with support from them, though, as well--I sent three different messages with three different questions, and only got a reply to one of them. Seems there's a real need for somebody to come up with a good control software and good support at the same time.

    So, are you running Mach3 on a parallel port computer and feeding that signal to a G540?
    No, I'm currently using Linux on a parallel port. I was using the G540 but returned it and got the MX3660 from CNC4PC (well I haven't got it yet, will be here next week. Terrible customer service experience with them, I hope it's not always that way). It's similar to the G540, except one less axis, but more power and options (I covered all of that in post 7). I'm still considering an external ethernet type of controller that would give me more I/O, but for now this setup is adequate.



  17. #17
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by Marty_Escarcega View Post
    This may work provided you add an encoder to the spindle motor. This is doable. If you pull the tachometer off the shaft, then measure it (I would ask Timothy for the size of the shaft where the encoder armature mounts, I have not had my motor apart). Some motors had both the tach and an encoder. There may be a stub on it for an encoder and you just need a mounting plate. I would ask Tim to see whether or not after removing the Tach armature, the shaft is smooth, what size and whether or not it can accept an encoder.

    It may work because it takes step and direction signals directly.
    What does? The motor or the amp or both? Not sure what you're saying.

    The SD3060 provides 30amps continuous current with 60amps peak while the Duogong will max at 35A.
    When you say Duogong are you referring to the DG4S I mentioned?

    You will likely never push it this hard so the amp will survive. Price is pretty good too. One thing, when you rectify 120VAC you get 169VDC which is over the nameplate. You'd better ask Arturo about this. Why they didn't design for 180VDC is beyond me. (probably cost) Now if your AC voltage was constant 110 (yeah right) then no problem.....
    The nameplate on my motor says 180V.



  18. #18
    Member
    Join Date
    Apr 2010
    Location
    United States
    Posts
    125
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Quote Originally Posted by Marty_Escarcega View Post
    You know, to really make things easy, you could just use the VFD output and run the spindle forward and if you ever want to tap, tapmatic head...this would get you going even quicker until you sort things out.

    Hand tool change and just run the spindle in forward....you aren't running production, you just want to learn and make a few parts. As time and money permits, you can work on the other machine and sort things out, perfect and then convert the first machine. You have 2 machines.....what is your end goal? Get a CNC capable CNC machine up and running quickly do you can make your prototype parts right??

    Then its just a matter of sorting out your power drawbar....
    Marty, this makes really good sense. In fact I've kinda been coming to that conclusion myself. Like you said, getting the axes moving is the easy part (already got that part taken care of). So what exactly do I need to make the spindle work for this simplified setup? The MX3660 driver seems to have the VFD output you mentioned but don't understand how to make that connection (manual for the driver, page 8, pin 14). I'm guessing that goes to the Duogong driver, but don't understand how to control it from LinuxCNC. Can you give me a general diagram of what I need and how to connect it?



  19. #19
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    Dugong is the name of the servo drive. It is rated for 160V & 35A https://cnc4pc.com/dg4s-16035-dc-servo-drive.html ask if you can run it without an encoder?
    What your motor nameplate says is that its maximum terminal voltage is 180V (as supplied by the servo amplifier) but your servo amp is only good for up to 160V INPUT. If you rectify 120VAC you get 169V. Larken's Viper is good to 170V Viper Servo Motor control products



  20. #20
    Member
    Join Date
    Dec 2009
    Location
    USA
    Posts
    521
    Downloads
    0
    Uploads
    0

    Default Re: Build log-drhanger's DM4400 electronics retrofit

    If your servo dynamics amplifier is good you have what you need to drive the spindle. All the amp wants to see is 0-10VDC. Heck you could rig a power supply up and feed 0-10VDC to the amp and your spindle is running.....

    Please describe your current setup
    Computer & software
    Parallel port, USB or Ethernet port?
    Motion controller? Smoth Stepper?Pokeys? No external motion controller?
    Break Out board?
    Axis stepper driver (I think you were using G540 but now you went with the Leadshine knockoff)



Page 1 of 5 1234 ... LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Build log-drhanger's DM4400 electronics retrofit

Build log-drhanger's DM4400 electronics retrofit