Machsupport Forum
Third party software and hardware support forums. => CS-Lab => Topic started by: jofriedl on December 29, 2014, 11:53:50 AM
-
Hello there guys,
Just like the title says, I'm working on a Matsuura retrofit that I'm pretty excited about. The machine is a 1979 Mc500v. The motion controller is a CSMIP/ip-a and an Automation DL06 will be running the tool changer. At this point, I'm well on the way to having the PLC programmed. I also finished ironing out the M6 macro last night.
I could definitely use your guys' help with the ethernet controller integration though. When I initially concocted the retrofit, I planned on using a digital encoder and swapping out the servos. However, after reading about Hood's success with the analog version, I took the leap and bought the IP-A in the hopes that I could simply wire it to my existing yaskawa drives.
I contacted Yaskawa hoping to obtain electrical prints for my servopacks. After bouncing from one department to the next, I was finally told by a Tech that all schematics are proprietary (never mind that they have other servopac prints available for download on their website ). It was suggested that I contact matsuura and obtain electrical prints for the machine. While Matsuura USA is more than willing to share their prints, they have been unable to locate the schematics for my machine. So my first question is, does anyone out there in Mach land have schematics for Yaskawa CPCR- MR 052 NB servo drives?
As a side note, the DL06 is a pretty cool device! I highly recommend them for anyone implementing a tool changer.
Jonathon
-
Have you looked at the standard CPCR- MR 052 manual? It may not be much different from the drives you have, I know some drives I had that were OEM for someone else were identical to the standard drives except for the numbering.
Hood
-
Hood,
You're just the guy I hoped would respond! I've read quite a few of your posts. Thanks for your response!
I haven't been able to find a manual for the cpcr mr 52, but I have found a cpcr mr 01C - 07c. ALso, Foothills Machinery out of Denver contacted me today to let me know they found electrical schematics for the mc500v. I'll be honest, I don't have the background to fully understand these wiring diagrams. There are a few things I can pick up, but I need to learn a bit more before I can figure out where to wire step and direction inputs and encoder outputs. In the attached image bellow, the wires are there and numbered, but not labeled in great detail. I'm assuming PC stands for pulse counter. As for T.G. stands for, not really sure. Also, the diagram depicts connections between the 3000g controller and X,Y and Z axis motors. I hope they mean X,Y and Z servopacks, because the controller will wire up to the servopacks alone. (step, dir and encoders).
This retrofit i my first experience with servo controlled systems, so If I am missing something, feel free to mention it. Thanks!
-
oops! I attached the wrong portion of the wiring diagram. Please see the attachment bellow for wiring of the servo motors. Thanks!
-
First thing is you do not have Step/Dir inputs on the drives, they will be analogue +/-10v, that is why you will need to use an IP-A controller.
The TG will be Tacho Generator, basically a means of the drive knowing the rotational speed of the motor and will be integral to the motor and will come out on two wires.
In the above drawing it looks like you have an encoder also on the motor, the B A C D E F wires, can you find them in the cabinet as they likely go to the drive and could well be Encoder signals. I would assume C and D are the power to the encoder and A and B are the A and B channels of the encoder but not sure about E and F. If there had been another wire I would have said it was hall signals, maybe the manual will shed some more light on it.
This seems to be the standard drives manual
http://www1.yaskawa.com/site/dmservo.nsf/link2/TKUR-5EJQ5Z/$file/TSE-C717-13E.pdf
I have not had a chance to look at it yet.
Hood
-
Thanks for clearing that up, Hood.
I was looking at pages 2 and 4 of the manual. The servopak wiring looks fairly simple. But I was under the impression that most servo drive to servo connections included the encoder feedback. Is that not typically the case? From the look of the diagram on page 4, the encoders are not connected to the servo drive. Once again though, reading electrical diagrams is not my forte.
Something else that caught my eye is that reverence input is rated from 4 to 50V. I suspect that the reference input is the analog voltage signal you referred to in your last response. If the servopaks truly require such a range, I'm thinking I might have to start looking for replacement drives. I remember reading another post in which you mentioned the brand AMC drives.
Another question; I know that CS labs sells an mpg module, but I'd like to use the original mpg in the yasnak control panel. I haven't done the research to know for sure, but I am almost certain that I can wire in the module through the discrete inputs on the ip-a. However, I also have a DL06 connected to mach3 via modbus tcp via ecom module and I'm wondering i there would be any advantage to running the mpg through the plc. Scott, AKA 'Poppabear' described the ecom module as blazing fast so I'm wondering what would be faster, digital inputs though the ip-a or inputs via the plc? This will also determine where the tool touch probe is wired.
Jonathon
-
Ok just realised these are DC servos (brushed), so no need for commutation. So all you will have going from the motor to the drive will be the Tach and the thermo switch (if it has one) and of course the A and B power wires for the motor.
If there is an encoder on the motor (rather than a resolver) then you should be fine.
Regarding the 4-50v you mentioned, I think that will just be the range that can be set ie 0-4v up to 0-50v.
I notice it also mentions differential of 4-10v, I think that is the one you will likely use and have it set to 10v so you will have a range of +/- 10v. I would need to see your drawings to be sure though.
In the drawing you attached above it does look like there is an encoder, going through from top to bottom I would say the connections are.
A and B are your main power wires for the motor, they will be from the drive.
H and G are your Tacho, they to will go to the drive.
B, A and C are your encoder signals (A, B and Index) and D is the 0v reference for them, its a single ended encoder rather than differential. These will go to IP-A
E and F are the power wires for the Encoder. These will come from IP-A
You will have to check the encoder is 5v square wave, it almost certainly is but best to try and confirm first.
Hood
-
Oh regarding the MPG, personally I would use the MPG module, you will get much better motion from the MPG if you let the IP-A control it.
If you use the normal Inputs or a PLC then it will be Mach itself controlling the motion via MPG and it will not be as nice or accurate.
Regarding the probe, you will probably be best to do it via the IP-A and use their macro, that way it is all hardware as the IP-A will do the probing rather than have to go via Mach. You may however have to alter their macro depending on what you want to do, I dont use it and have not looked, so not exactly sure what it is set up to do.
If however you have collet holders etc for tooling then maybe it would be an idea to use a setup similar to what I use. I have a height setter consisting of a linear scale and it is hooked into the Enc module and goes into Mach, I can then place a tool into the setter and then press a button in Mach and it will automatically set up the tool table to the correct height offset, below is a pic of the setter.
Hood
-
Hood,
That tool setter looks pretty impressive. Nice work! I like your brushed stainless switches. I just ordered a boatload of switches and a pair of joysticks from Automation Direct. The first machine i retrofitted I ran solely off the keyboard and mouse and I can tell you it gets old.
I made it to the shop tonight and started gutting the machine. For some reason, on projects like these, I plan to precisely excise only the components and wiring I know I won't need, but I end up getting carried away. And tonight, I got carried away! There is little left of the original machine innards and wiring, Limit switches, encoder signals, transformers, power wiring, wiring to the drives and the drives themselves. The rest is piled on benches beside the machine.
This brings me to tonight's question: Is it really worth it to salvage the original drives? If I were to upgrade to new drives like Advanced Motion Control Drives, what could I expect to pay? I would have called AMC myself today but being that today is a holiday I didn't even bother. I'm sure they were closed.
Per the encoder wiring, I have attached a picture of the encoder ID plate on the outside of the motor. On the wiring diagram there are 9 connections and I'm assuming that two of them are the tachometer. Since it is a single ended encoder, D will be the common ground for each channel. Connections labeled E and F I'm assuming are the power connections. Not sure about the other connections though.
Jonathon
-
The encoder will be as I mentioned above, ie it is the A B and Index pulses.
One thing that concerns me however is it takes 12v power, that may mean that it outputs 12v square wave. That is no use for the IP-A, it requires 5v. Likely you could make a small circuit to convert the 12v to 5v but best check it out with a scope to see if it is the case.
As for AMC drives, afraid I dont have a clue, I have never bought any.
Hood
-
I'll order an oscilloscope this tonight. Are you saying that If the 5V converted encoder signal resembles the raw 12v encoder signal wave then I should be good? I feel I might be overthinking this, but I wonder if the voltage regulator would slow down the signal from the encoder to the controller. Thanks for your help!
Jonathon
-
If the encoder does output 12v square wave then you would have to use something like a fast opto to convert it to 5v square wave.
I am not great with electronics so cant really help too much in that regard I am afraid.
Hood
-
A quick update......
After many hours spent talking with tech support for yaskawa, pouring over the nearly useless electrical prints provided by matsuura and lots of twisting wires together, I was able to move the servos using a 1.5v battery. While the manuals mentioned earlier in this thread seemed promising at first, the servopacks turned out to be significantly different. Only sheer trial and error yielded the correct wiring.
The encoder issue is another problem though. I bought three fast optocouplers from Automation Direct to convert the 12v pulse to a 5 volt pulse. These seem to work. However, something seems awry with the output from the stock encoders. Three Small LEDs built into the optocouplers light up when the coupler receives a signal from each of the channels on the encoder. For some reason or another, only two of the channel LEDs light up- A and B. I don't have my osciliscope up and running yet but my plan is to test the raw encoder output on the Z channel.
Now, I'm pretty sure I Know the answer to my next question here, but I have to ask anyway. First though, I have tried to run the autotune function with what I am presuming is only the A and B channel. The result is a moving axis that oscillates back and forth around a point like it were make of warm jello. The autotuning program only manipulates the P and I gain and leaves the D value at 0. This seems odd. But the most troubling thing is that when I hit the e-stop, the axis starts to drift. I don't have the servopack wired such that the IP-A can disable the axis upon e-stop triggering. I know that I should, but why would the controller continue to send out a reference voltage after the estop has been hit?
And the last question of the night will has to do with tuning the servo amplifier. The IP-A manual states that the servoamplifier must be tuned prior to attempting to tune the controller. While I am using both the original servos and servopacks which should already be tuned for the servos, I'm not quite sure that the original reference voltage was + or - 10V. What kinds of problems can this cause and what are some things I can do to alleviate them?
Again, thanks for any and all help Hood.
Jonathon
-
You will not likely see the Index pulse without a scope, it is very short in duration and your only chance is to know roughly where the Index is and very slowly rotate the motor back and forth until you find the exact point. Just a fraction of a degree either side and it will be gone.
When you hit E-Stop the IP-A is not controlling the analogue output, you will get noise on it which will make the motor drift, it will just be mA but that is enough. What you need to do is have an enable connected to the drives from the IP-A so that when you E-Stop the drive is disabled.
If the original analogue command voltage was other than +/- 10v then it will spin slower or faster at 10v depending on whether the input was lower or higher on the original.
Usually there is a way to scale the input top the output, ie so many volts per 1000rpm. So say you want a max of 4000rpm then you would adjust the drive so that 2.5v = 1000rpm.
Hood
-
Hood,
Thanks for the response. Thats great news to hear the encoder is working properly. I looked over a yaskawa manual for a servopack that is not completely dissimilar from mine. The wiring is different, but there is enough similarity between the two to follow their reference voltage adjustment procedure. As you mentioned above there is a way. On the front of the drive there is a variable resistor which can be used to set the proper response to the reference voltage. There is also a vr to adjust the zero movement voltage. Obvciously, I probably shouldn't mess with that one though.
As far as sending drive enable signal, there are four voltage inputs to the board: +15V -15V 24V and 5V. I'm fairly certain given the design of the similar yaskawa drive for which I have a manual that the 24V is the enable signal. The 24V controls 2 large relays on the base of each servopack that enable the drive to move in the positive or negative direction. According to the manual, the NC limit switches interrupt the 24V signal to each of the relays when triggered. I will install a relay (because I don't know the amperage draw) rather feed 24 directly from the IPA and enable the servos this way.
Tonights questions is a shot in the dark, but I think I'll ask anyway. On the front of the drives , there are three small contact relays. Two relays control 24v and when active trigger the larger relays that I mentioned above. The third controls the 110V ac which I presume is the power to the drive. Beneath each relay, there is a teardrop shaped component labeled SUP. I'm fairly certain this stands for voltage suppressor, but I'm not sure. When I google voltage suppressor nothing like whats on the board comes up. I attached a picture in the hopes that someone will recognize this part. Thats the first part of the question. The second part is a bit embarrassing. When I was fiddling around in the cabinet and testing the 24v input, I accidentally touched the 24V positive wire to a 110V ac source. FWAP!! Five out of six of those little babies sparked and blew apart; the large relays mentioned above clicked off. Lemme tell ya, that took the wind out of my sails! After more research trying to identify the parts, I discovered similar components wired into the back of a contact relay board I had previously removed from the machine. I clipped 5 of them out and soldered them in place of the blown ones. I hoped this would solve my problem, but no such luck. I can turn the drive on with the 24 volt signal, but the suppressors immediately short out. Unlike the first time when I blew them, the drive stays on, so I assume that continuity is maintained. However, the suppressors take on a charred appearance and even initially spark. I've replaced them twice now and the problem persists. I know this is outside the scope of the original post, but I'm not even sure where I would post a question like this one.
Thanks for any and all help!
Jonathon
-
No, wouldnt bother with the adjustment for the zero movement, it will just end up moving anyway and you would have to continually adjust it. Using the Enable is the way to do it.
Cant really be any help with the suppressors maybe the CNC Zone has a section for electronics?
Hood
-
Its been a while so I thought I would post an update.
I haven't fiddled with the servo drives since I posted the problem with the voltage suppressors. I did contact yaskawa again hoping to identify the mystery component. While they told me they are surge protectors, the exact specifications are proprietary. Go figure. At this point, I'm considering replacing them with an off the shelf radio shack part that won't fail under the current load.
I ordered an MPG module from CS labs per Hood's recommendation. Because the mpg encoder emits a 12V signal, I got another Fast optocoupler from automation direct. They convert the 12v signal to 5v. As you can see in the picture, I haven't wired it up yet, just snapped it onto the din rail earlier today.
I also bought an ENC threading module. I'm hoping to do rigid tapping, or semi rigid tapping or synchronous tapping, whatever the heck they call it when the z axis follows the rpm of the spindle. I admit, I've done zero research on setting up rigid tapping on a cs labs controller so I'm not even sure I need the module. I was just ordered it with the MPG module to save on shipping. If I don't use it on the matsuura, I'll use it on a lathe retrofit I hope to do next.
When I started the project, the first thing I did was connect the DL06 plc to mach3 and test out modbus tcp. It took a while to simply connect the pc to the PLC. Hours of research revealed that oem firewall settings in windows interrupt the communications and must be deactivated. You want an unfettered signal between the two. After a solid connection was made, I learned how to work the modbus tcp setup page. This step was extremely frustrating as only through trial and persistent error did I discover that mach has to be shut down and rebooted every time you change a setting on this page. (very much like the brain set-up). Next, I experimented with using brains to activate outputs and internal contacts in the PLC. I learned basic ladder logic by watching endless you tube videos on the subject and reading the exhausting and thorough DL06 Manual. I can't thank Scott Schaffer aka Poppabear enough for turning me onto the idea of using a DL06 for the toolchanger and other accessories. Using modbus tcp, brains and a plc all in conjunction allows awesome control over the machine and all its functions. For example, I'm thinking of creating a program in the plc which will allow me to flip a switch on the control panel which will activate a program that will send periodic blasts of air via pneumatic solenoids at the tool to help with chip clearing, but only if the spindle is active. This will help tremendously with deep pocket milling.
Another cool thing you can do with the DL06 is control the spindle vfd. This is imperative for the tool changer, as the first step in the process is to rotate the spindle at very low rpm until the dog locks it into the correct orientation. Did I mention that I love my DL06? AD customer support has been super helpful as well. As a first-timmer, its not easy integrating all this hardware. Without this forum and their customer support, I'd be bald from pulling my hair out.
Lastly, I redesigned the control panel. I'm having it waterjet by a company here in town and then I'm taking it south to Denver where another company will laser etch the button labels. Ive included every button I can imagine needing, but if i missed one, please speak up. The two unlabeled circles in the center are joysticks and the small rectangle box above them is a USB port for loading programs. The large box in the lower left hand corner is for the MPG of corse and the unlabeled circle above it is the estop. I'll post a pic of the control panel the minute it gets back from laser etching. ;D ;D ;D
-
You will need the Enc module for rigid tapping, you will also need an encoder connected directly to the spindle.
Personally I prefer the old serial (Non-plugin) interface for the PLC but others seem to prefer the TCP.
I do not have to restart Mach after changes to the PLC but maybe that is a quirk with the plugin/TCP.
With Brains you definitely do not need to restart, you just need to reload them after any changes.
On my machines now I have a power button that switches on the computer, there is then a relay connected to the computers power supply that brings in main contactors which in turn power the 24v power supplies and the logic to my servo drives. This means there is a single button to switch on the whole control and then I have another On/Off integrated button to switch the mains to the servo drives.
Regarding your spindle and coolant, you really only need 1 button for each as in Mach they are toggle buttons, ie if off a press will switch on, and vice versa.
Hood
-
Just a quick update. I got the control panel back from the waterjet and laser guy. The laser etching didn't show up as well as I;d hoped, but I've been told that I can rub acrylic paint into the etching to make it pop a bit more. I installed the myriad of buttons in the panel and mounted it in the machine. Since I don't have room for all the buttons I want and a screen, I'm mounting the screen and keyboard console off to the side.
I also spent a few hours yesterday convincing the DL06 to talk to the GS2 drive. The first problem I encountered was differentiating between necessary and unnecessary ladder logic to get the dl06 to write to the drive's modbus registers. After a few hours I was finally able to write to registers in the PLC via the TEST MODBUS window in mach3 and have the PLC then write to the registers in the GS2. I then wrote a simple brain using the spindle LED and the commanded speed DRO and modbus. Unfortunately the GS2 drive speed reference uses HZ rather than RPM. To achieve 1750 RPM you would send 60HZ to the speed reference register. I'll have to write a program in the ladder logic to perform the appropriate calculations to convert the commanded rpm to hz.
The one odd thing that I'm not liking about the DL06 to GS2 interface is the lag time between when I tell the spindle to power off and when it powers off. Precious seconds pass after I click the spindle button and when the spindle actually starts decelerating. Whats odd is the spindle is much more responsive powering on than off.
-
Is there a reason you are using the PLC for the spindle control rather than using the IP-A?
Hood
-
Yes. The dl06 runs the tool changer and also orients the spindle for the tool change. There is a pnumatic dog that locks out the spindle in the correct orientation. I decided to have the dl06 run the drive because it would be easier to program the tool changer. I suppose I could connect the spindle and pnumatic solenoid for the dog to the iPa and write that portion of the tool change in the m6 macro.what do you think would be easier? Thanks.
-
Have you tried the spindle orientation in the IP-A?
I have never tried it with the IP-A but I use it with the IP-S on the wee lathe so that the chuck stops with the key in a suitable position.
I assume you have an encoder on the motor?
Hood
-
I think I do have an encoder on the motor. I haven't looked into installing it yet because I'm not at that point yet. To be honest Hood, I didn't even know the ip/a had a spindle orientation feature. Does this need programming in VB to work with a tool changer or does it automatically perform the operation at the activation of the M6 command? I'll do more research tonight.
You know, another option would be to have the IP-A run the spindle but also have the DL06 send a low rpm signal to the drive during a tool change operation. I have the drive set to communicate via modbus with the DL06, but I could get an analog output expansion module for the dl06 and send the signal that way. I definitely like the idea of having the IP-A run the spindle during programs. This way, if there is every a problem with Mach to DL06 communications the program won't start with a dead spindle.
-
When you stop the spindle it stops at the defined position, so for example if you have 20 degrees set it will stop at 20 degrees from the encoders Index.
As said I have never tried it with the IP-A but it does work well with the IP-S, so no reason to think it would be different in the IP-A, then again ::)
Yes you could do that with the PLC and toolchange but I would definitely want it via the IP-A for the actual spindle, it should be more accurate in speed I would think as well as I am sure even with a VFD you can set it up closed loop but best check on that as I have never really looked into it as I use a servo.
Hood
-
Just looked in the plugin and maybe the angle stop does not work with a VFD as when you disable the Spindle Axis option it greys out. Will have to download the latest plugin here and see if it is the same.
Hood
-
Nope looks like it is still greyed out if you choose the DAC option rather than spindle axis :(
Same with the closed loop option.
Of course if your VFD can accept +/- 10v commands rather than 0-10v then you could set it up as a spindle axis.
Hood
-
Thats too bad. You had me all kinds of excited with talk of spindle orientation! I looked over the IP-A manual and from my cursory glances I gather that only with a servodrive running the spindle can you orient the spindle upon stop. That settles it. Next time I retrofit something it will have a servo for the spindle motor.
As far as the drive accepting + or - signals, the gs2 drives from Automation Direct only accept 0 to+10v.
Can you explain the difference is between DAC and spindle axis? If the drive will accept an analog input and has an encoder, I don't understand why the IP-A could not perform a closed loop operation.
-
To be honest I am not sure either why it can't. I was almost certain it could as I seem to recall a guy from Greece, I was helping set up a lathe, said that CS-Lab had said it was possible for closed loop with VFD. I must have been mistaken but it may be worth an email to them to see if there is a way.
Hood
-
Well, I think for now I'll keep the spindle under dl06 control while I get the rest of the machine on line. I'll engineer some kind of alarm that will trip an estop is the spindle doest respond to Mach commands to prevent problems. Theoretically it should work. Before the machine is finished I'll get the analog expansion module and place the drive under IP-A control as I mentioned above.
When I started researching options for the retrofit, I talked to a company stateside here called PMDX. They had provided the breakout board I used for a brideport retrofit I did a few years ago. They didn't have an ethernet control board, but they told me Brian Barker had recently purchased two of these very same machines and was planning to retrofit them both. I emailed Brian and he put me in touch with the guy who was designing his retrofits, Scott Schaffer. Scott planned to have the spindle under DL06 control. From the reading I've done, Scott is extremely knowledgable on machine design and mach retrofits. He's done quite a bit with this particular PLC. So, long story short, this is the primary reason my DL06 controls the spindle drive.
Just out of curiosity, do you know of anyone who makes a vfd that takes a + or- 10V signal?
Jonathon
-
I'll get the analog expansion module and place the drive under IP-A control as I mentioned above.
Is that a module for the VFD?
Regarding the Alarm, just input it to the IP-A and set it up as the spindle alarm in the plugin.
Yes, Scott knows his way around the 06 PLCs, thats for sure.
I know the Yaskawa spindle drive I have has +/- 10v, it is basically a big VFD but as for other "normal" VFDs afraid I do not know offhand any that do the +/- 10v, I dont use VFDs really so not too familiar with them.
Hood
-
Was curious what options were available for VFD with +/- 10v so did a Google, seems quite a few, the GS3 variant from Automation Direct is one, also saw the Delta VFD's seem to have that capability. There were quite a few others as well.
Hood
-
Nice conversion! :) Just wondering, looking at the pics of your control cabinet, what servo drives did you go with? The ones in my mc 500 are huge. Getting ready to convert so trying to find as much info as possible.
-
Thanks! I ended up using the stock servo drives. I think they're CPCR MR052-nb made by yaskawa. Its a pretty cost effective way to go about a retrofit. Servo drives are quite pricy and the available selection of analog drives is pretty slim. What year is your MC500?