Brother TC229 Linux CNC retrofit

Hi

Four month ago I bought a perfectly good TC229 with the conversational control for about $7000 with the conversational control. I went to some trouble to get one that was in good shape mechanically, I couldn't measure any backlash using a tenth indicator.

My original hope was to keep the original Sanyo Denki motors and drives but I found that the Brother control was talking to the drives with some sort of serial format so I gave up on the Sanyo denki drives. The motors are four pole brushless servos with incremental encoders (5000 lines for XYand Z, 1024 lines for the spindle) and no hall sensors. After spending 40 hours of research and another 40 hours of experimenting I managed to get all the axis motors running with AB Ultra3000 which has an algorithm that allows it to determine its commutation based on the encoder pulses without the use of hall sensors. I made a bunch of educated guesses for motor values such as inductance, resistance, rated current, torque constant etc, if anyone needs help with the specifics they should email me, it took me a long time to figure it all out. Overall I am happy with drives they allow a lot of flexibility in configuring the IO and the Ultraware software works well. Many of the used Ultra3000 drives on Ebay have the SERCOS network feature but I found that the software allows you to disable this and use the either step and direction or analogue commands. I have a mix of 2098-010 and 2098-020 models. I will need the higher current 020 for the z axis.

The z axis is not counterbalanced, the z axis motor has a brake, one the brake is disabled (24V) the z-axis drops at an alarming rate, I found this out the hard way. Unfortunately the commutation routine needed for the Ultra3000 does not occur until a few seconds after the enable pin goes high, the brake is released as soon as the enable pin goes high. I did find that the drive was able to successfully perform the commutation routine with the brake on, this surprised me but I think I will try to control the brake with the ladder in LinuxCNC, as the drive can provide a signal that the commutation routine was successful. I have not actually tested this yet.


My original plan was to use an Ultra3000 to run the spindle as well, the problem with this plan is that drive runs out of voltage before the motor gets to its top speed of 10000 rpm. The drives I have can run with an input voltage of 120-240v. When I put 120V in I could only get the motor to go to about 3600RPM, meaning 240V would get me 7200 Rpm max. The other problem is that an Ultra3000 that would be capable of driving this motor to full capacity would be costly ($1500 on Ebay). I am now researching VFDs that are capable of driving permanent magnet motors, this will hopefully allow for field weakening that will allow me to drive the motor to 10000 rpm. Orienting the spindle will be more of a challenge with a VFD instead of a servo.

I have a Mesa 5I25 and 7I76 daughter card, my plan is to run the axis motor in step and direction mode, and use the classic ladder to control the tool changer. It looks like the toolchanger has some sort of 5 bit sensor that indicates position and a three phase motor driven cam indexer to rotate the turret. loading and unloading the tool appears to be done by a mechanical linkage powered by the motion of the z-axis.

With some luck I hope to have the control panel finished and the various motors and sensors wired up within a month. I will post more info as I struggle through.

I might be able to help on the atc i Do right now an Bridgeport 412 interact

Gesendet von meinem SM-N9005 mit Tapatalk

Thanks, I am hoping the ATC programming will not be too hard, I think orienting the spindle using the VFD will be the trickiest part.

If you use linuxcnc there are 3 ways i know of orienting it i also Need both today iWork on 2nd Way to Orient ,.. I use a Toshiba vfs 11 vfd


Gesendet von meinem iPad mit Tapatalk

Quote:

---

Originally Posted by sbaer

Thanks, I am hoping the ATC programming will not be too hard, I think orienting the spindle using the VFD will be the trickiest part.

---

There is nothing tricky really, just use VFD which is capable to orient the spindle (...like Tkamsker said). I use Delta Electronics C2000-series VFD, these have also PLC build-in, which give some possiablities to make safety features. There is lot of other manufacturers (Omron, ABB, Vacon, Allen Bradley, Toshiba...) who sell exactly same type VFDs. I have same type Brother in progress (link to topic: http://www.cnczone.com/forums/vertical-mill-lathe-project-log/213078-brother-tc225-retrofit.htm ), but I don't use VFD (Delta C2000) orient feature, CNC controller does whole process by +/10V speed command.

Tool changer is pretty easy. I don't just can't remember was that coder 4- or 5-bit with or without parity checking, but you can just figure it out with normal voltage meter. Also there is some signal from Malta's cross mechanism when it is in lock position (if right tool is in place, motor turret motor stops to this signal). I probably put copy of PLC to my Brother topic, I think there is everything what you need and you can translate to your PLC, whatever you are going to use. Only shorts rotation path calculation is made in my controller by special command, but it is not that complex maths anyways.

Tricky part with your ATC can be limits, you need two sets of soft limits, I'm not sure how that is done LinuxCNC (is there any status/control bits in PLC?), because I don't have that much experience of it. Normally you can't run to tool changer area, so when you have tool changing in process (M6), second limit set is in use. Also there is (physical) limit switch for tool changing area, which activates EM-stop circuit. This switch is overriden when tool change is in process.

I really recommend you think little bit how you can avoid accidental spindle run which can be cause of wire breaking. Take some signal back of spindle run to you CNC controller's PLC (or do you use some external PLC?) and make sure your it can drop whole system down (for real, not some just enable signals). If CNC gets signal "spindle running" and there is no affecting M3, M4, M29 or orient process going on --> whole system down and fast. Also you can use signal vice versa, if spindle is not running when it should, actived feed hold and show alarm. It is pretty useless to try to mill something if spindle is not rotating ;)

Then about your voltage/rpm problem... C2000 (like many other) is capable to run synchroneus motors. This VFD have sensorless PM mode and also you can buy additional cards like PG01U with U/V/W phase detection. Voltage is not problem, if you have 240VAC, it is after VFD's rectifier far over 300VAC. But if your motor is not rated to 10k rpm, I think it won't run there without field weaking (and that's impossible, because motor have permament magnets far as I know..)

But what is my solution for this motor problem, is induction motor (which is for made for spindle use by Advantech-LNC) which runs up to 10krpm. Controlling by closed-loop FOC (drive same C2000), it is not going shame original PM motor at all (we have used C2000s and induction motors as servos on big horizontal milling machines with success)


Ps. By the way, do your machine have 10 000 rpm spindle? Is there any coolant system for bearings..?

Thanks for the reply. My machine came with a 10 000 rpm spindle, I did not see any provision for cooling. We played around with the toolchanger on Friday. One thing we found is that it takes a lot of current from the the Z-axis motor to move the lever that compresses the tool retainer spring. We ended up having to "take a run at it". The problem may be that we have damaged the z-axis servo motor by overheating it while we were figuring out how to get the Ultra3000 to commutate without the z-axis falling. The magnets in the Z-axis may have been overheated to the point of lowering the motor's torque constant. My plan is to replace the z-axis servo motor with one that has hall sensors.
Another issue we are facing is how to get the turret to stop consistently in the dwell are of the cam. I am not sure I am using the correct motor contactor for the turret index motor, there is one contactor that has some extra features that I do not know the purpose of.
I will look into the Delta VFD, it is probably less expensive than the Yaskawa, at the moment we are using an Ultra3000 servo drive for the spindle but the continuous current rating on this drive is only 7.5A.

Stan

I have two Pratt and Whitney ancestors to your Brother/Robodrill sitting in my garage right now, and am about ready to start my retrofit.
The hardest part to me seems to be the tool change functionality, I am wondering how the ATC functions on your machine? Mine is completely mechanical, and uses the excess Z travel to operate some drive dogs, I saw mention of two sets of limits in this thread so I assume yours is the same. What I am wondering is how your machine rotates the turret, mine appears to have a big gear on the spindle that interfaces with the ATC, so rotating the spindle will rotate the carousel, does that sound like how your machine operates? I have a replacement AC servo for the spindle so if I need positioning for my tool changes that shouldn't be a problem, but I would like to know how to "wire" that all in, is the spindle set up as my A axis for the ATC?

Hi on the Bridgeport it is an 110 v ac Motor and some gear. There is an Relais thing which changes polarity of an Elko so it is one Relais for on of and a 2nd for left right which are in the Old cabling going to 2 Relais one is turn left the other turn right and i Control them with my modern Relais ..

Quote:

---

Originally Posted by sbaer

My original hope was to keep the original Sanyo Denki motors and drives but I found that the Brother control was talking to the drives with some sort of serial format so I gave up on the Sanyo denki drives. The motors are four pole brushless servos with incremental encoders (5000 lines for XYand Z, 1024 lines for the spindle) and no hall sensors. After spending 40 hours of research and another 40 hours of experimenting I managed to get all the axis motors running with AB Ultra3000 which has an algorithm that allows it to determine its commutation based on the encoder pulses without the use of hall sensors. I made a bunch of educated guesses for motor values such as inductance, resistance, rated current, torque constant etc, if anyone needs help with the specifics they should email me, it took me a long time to figure it all out. Overall I am happy with drives they allow a lot of flexibility in configuring the IO and the Ultraware software works well. Many of the used Ultra3000 drives on Ebay have the SERCOS network feature but I found that the software allows you to disable this and use the either step and direction or analogue commands. I have a mix of 2098-010 and 2098-020 models. I will need the higher current 020 for the z axis.

---

Hi, it's been a lot of years, but maybe you still remember or have documentation of what parameters you entered into Ultraware for sanyo denki engines. I have a BROTHER 312 that has crashed and I want to control it with ULTRA3000 drivers, but after a few days of trying I give up. The motors I have in the machine are 700W 3.36A 4-pole with a 5000 PPR encoder. Please help.


Please forgive the mistakes I translate using googletranslate

Quote:

---

Originally Posted by szary11

Hi, it's been a lot of years, but maybe you still remember or have documentation of what parameters you entered into Ultraware for sanyo denki engines. I have a BROTHER 312 that has crashed and I want to control it with ULTRA3000 drivers, but after a few days of trying I give up. The motors I have in the machine are 700W 3.36A 4-pole with a 5000 PPR encoder. Please help.


Please forgive the mistakes I translate using googletranslate

---

i have successfully control spindle TC-22A. TC-S2A with Vector Control Drive the price only 370 USD from china. and able to run at 12000rpm without problem., and also i can control spindle for Spindle Orientation (for ATC Purpose)