Mach3 Retrofit to 1989 CNC Lathe

Hi all,

I'm looking for a little advice. I recently bought a 1989 Somab Transmab 400 CNC lathe in great condition for next to nothing. It has a Gamet power chuck, powered tail stock, Sauter 8 turret tool changer, swarf conveyor, 10kW spindle motor and came with 40 tool holders! It is controlled by an old Num 720 system which I'm looking to replace with a Mach3 setup.

Keeping it short and to the point, electronics currently consist of:

2 x Indramat TDM 3.2 AC servo drives
2 x Indramat MAC 090A AC servo motors
1 x Indramat KDV 1.3 power supply (100amp)
1 x Indramat KDA 3.2 spindle drive
1 x Indramat TBM 1.2 Bleeder
2 x litton G58LSLDBI encoders (not sure on spec)

I'm thinking I have the following potential options:

1. Keep AC power supply, servos, encoders and drives (no doubt analog - need Step2Linear). Mach3 via PP and breakout board? This would be easiest option if it was lucky enough to be that simple.

2. Keep AC power supply, replace drives with suitable AC step/dir units, keep servo motors, new encoders?

3. Removal ALL, use new servo motors, new step/dir drives, new encoders and power supply?

Machining of parts / modifications to fit new components is not a worry.

Which ever method you guys deem most suitable I then need to consider spindle speed / direction control through Mach3 (hopefully I can mate it up with the KDA spindle drive which I assume will be analog 0-10v), the Sauter 8 turret tool changer and numerous relays controlling the powered tail stock (on/off), coolant etc etc.

I have already converted our 3-axis Bridgeport Mill to Mach3 using CNCDrive DG4S drives, encoders, breakout board etc so I have an idea of whats involved. However I know this conversion will require most likely ALOT of work.

Any thoughts much appreciated.

Thanks,
Jim

It would be a bonus if you could keep the motors at least, and possibly the drives, if they take analogue then reusing you would stay ahead of the game, you could look at Kmotion with a Kanalog card.
Al.

Thanks for the reply Al.

I've just had a quick look at Dynomotion' KFLOP and Kanalog combination package, and it looks impressive! My only worry is that KmotionCNC software looks very basic and has minimal functionality to support Lathe use (mostly seems to be mills / routers). Whereas Mach3Turn seems to be much more accommodating if I use the Kmotion Mach plugin.

Plus, Mach3 is has a huge range of supporting CAM software, which is a must for me as the majority of my work is done via CAD.

My goal is to have a relatively simple to use setup in which I can transfer drawings to CAM to create G-code to upload to Mach3Turn. As well as the standard lathe functionality it is essential I am able to incorporate the tool changer, spindle speed / dir and most importantly single point threading reliably.

While I look into whether I can use the existing drives, can someone recommend replacement drives capable of running my Indramat MAC 090A AC motors? http://manuali.eltex.biz/Manuali_Indramat/Motor_MAC.pdf

Dynomotion can also use the Mach front end.
Also it has true spindle servo sync for threading, unlike Mach which is just a single pulse/rev detector.
Al.

Good evening jim_cliff
I just bought a lathe the same, but the wiring of the electrical panel, the tool-holder turret and the control panel have cut cables, could you tell me how I can put it all?
Thank you very much

Toni, I'm picking the programming manual up next week so hopefully I can share this with you then.

Thanks for insight so far all. At this stage I'm trying to get my head around the servo drive I/O to make sure KFLOP & Kanalog are compatible. I've spent hours going through the TBM pdf and taken notes of information which could be useful. I've attached diagrams showing servo drive I/O.

Notes I've made from documentation & part identification on machine:

• Drive operated as vari-speed drive in position control loop of NC. A speed set-point is fed to drive from NC.
• Drive set-point input via a differential input or two summing inputs with zero voltage reference.
• Drive servo can be run at controlled speeds by injecting controller enabling signal at RF input & applying an analogue set-point voltage, that is proportional to the motor speed, across inputs E1 & E2 or E3 & 0VM or E4 & 0VM.
• Drive input E1-E2 (analog) is a differential input, inputs E3 & E4 (analog) are summing inputs referenced to 0VM
• Tacho feedback to drive only
• Motors have no encoder, just tacho.
• Encoders are mounted separately on each axis and will run to NC (NUM)

The info I have for the drive pin-outs are as follows:

• MA - Master output
• E1/E2 - Differential input for set-point (-10 to +10vdc)
• E3 - Summing input for set-point with respect to 0VM (-10 to +10vdc)
• E4 - Summing input for set-point with respect to 0VM (-10 to +10vdc)
• RF - Controller enable (0 to 3vdc)
• Ired - External torque & current limiting (0 - 10vdc)
• Bb - Ready state of drive
• Tsense - Actual speed valve from tacho (0 - 10vdc)
• 0VM - Reference potential for all inputs / outputs except E1-E2 & Bb-Bb
• TU-TU - Outputs if drive overheating

Couple of questions:

1. Why are there three separate analog inputs (E1/E2, E3,E4)? From what I understand only one of the three is used for complete control, but in my case which input method would I use to link a breakout board / motion controller with DAC (KFlop / Kanalog for example)?

2. Given encoder signals go back to the NC (soon to be replaced with PC) and the drives rely solely on tacho feedback, can I scrap the encoders and stick to the tacho's? My drives don't have the right programming module to accept encoder signals and I know Mach3 is essentially an 'open loop' controller. The loop is only closed between drives and encoders / tacho's. Could this cause me problems?

Typically where the encoder goes back to the CNC controller on older systems used velocity drives which had a tach input, now the general tendency is to use torque mode drives which do not use a tach.
They are BLDC drives and it looks like the tach goes back direct, not through the terminal inputs so all you should need is the analogue input from the controller.
You cannot use tach only for positioning, you do need the encoder, at least.
If using Kanalog, the encoders go to the Kmotion card, if using mach only, then they would have to go to the drives, but Mach output is step/dir, so using those drives may be out if just Mach control only.
Al.

Centroid Oak or All-In-One-DC board? Their software & support sure is good!

I agree the tach goes back directly to servo drive from motor. The servo drives have no communication with the encoders on this particular machine. Even though the TBM drives are capable of having encoder feedback as shown in the circuit drawings, they are only programmed to use tacho feedbach in this case (I think - and I have no way of reprogramming it).

Al I'm slightly confused by your last post. If I was to stick with velocity mode control and therefore keep the tacho to drive config, surely I would just need a controller to inject -10 to +10vdc into the drives? The KFlop would accept step/dir signals from mach (using dynomotion mach plugin) which the Kanalog would convert to analog -10 to +10vdc signals which would then be fed to the drives? Are you saying this would not work?

Is Kflop capable of taking axis encoder inputs then? If so, could I run the existing or compatible encoders to the Kflop while using the setup described above?

Kflop accepts encoder input and Kanalog is capable of ±10vdc
Send Tom an email, he is very helpful.
Al.

Hi Jim,

I think KFLOP+Kanalog would be a good fit for your system. You can run Mach3 or our KMotionCNC Software.

I agree the tachometers might be left connect directly to the drives so that the drives remain operating in Velocity mode.

The differential encoders can be connected back to Kanalog's differential encoder inputs.

I often recommend a simple test with a battery to see if you can get the amplifiers enabled and the motors can move at controlled speed in both directions. See:
Kanalog Hardware Info - Dynomotion

If you have any specific questions let us know.

HTH
Regards

Tom, thanks for the reply!

I've noticed the drives are currently supplied their analog control signal through input E1/E2 which is a differential input. Inputs E3 and E4 are summing inputs as I mentioned previously. I will try the battery test later on today but am unsure of which input to use?

E1/E2 would be my choice according to the sheet.
I use a 9v battery with a 5k or 10k pot across it and use one end and the slider to test/vary rpm.
Al.

Quote:

---

Originally Posted by Al_The_Man

E1/E2 would be my choice according to the sheet.
I use a 9v battery with a 5k or 10k pot across it and use one end and the slider to test/vary rpm.
Al.

---

Sorry for the continuous questions. I want to be sure I can control the drives via +-10v before I go any further.

I'm using a 9v battery varied through a pot as you suggested Al. I'm getting no output what so ever from the drive. All power lights are on, the tachometer and encoder feedback lights are on and I've applied 24vdc to the RF pin to enable the drive.

I've connected the varied +ve from the pot to E1 and ground to E2. Adjusting the pot has no effect and the axis does not move.

Reading further into differential inputs it seems that they require two seperate inputs and a ground. That way their is no noise or interference from the 0v line. This is opposed to a single ended input which is a 2 wire configuration consisting of a +ve and ground.

https://en.m.wikipedia.org/wiki/Differential_signaling

In the data sheet for the tdm it only references E1 and E2 as differential (should I connect the negative of my battery to the 0v pin on the drive too?) and then summing inputs E3 and E4 which are each individual and referenced to 0v. One of the 3 puts is used only. The existing wiring config for the NUM controller has just two wires: 1 to E1 and the other to E2. All other inputs and pins on that plug are disconnected. But there must be a common ground somewhere surely?

I'm confused. Any help appreciated.

Thanks again

Was there any original connections to the other inputs?
Usually analogue input does not always need a reference to ground as the signal is differential.
Is there any Ready Light on the drive?
Al.

Great, works perfectly.
+/- 10v across both axis varies rpm accordingly.

There was a motor brake requiring 24v to release on the X axis.

Thanks for the help all, will keep you posted regarding progress.

Before I go ahead and order the Kflop & Kanalog from Dynomotion I just need to clarify spindle control.

Is Kflop / Kanalog combo capable of Spindle speed / direction control through the Mach3 plugin? The main spindle motor drive I'll be using is an Indramat KDA3.2 with -10v to +10vdc analog input. Would I wire this up as an additional axis or does Kanalog have a port dedicated to driving the spindle?

I would suggest you email Tom direct, he usually answers all questions promptly.
Al.

Hi jim_cliff11,

Quote:

---

Is Kflop / Kanalog combo capable of Spindle speed / direction control through the Mach3 plugin? The main spindle motor drive I'll be using is an Indramat KDA3.2 with -10v to +10vdc analog input. Would I wire this up as an additional axis or does Kanalog have a port dedicated to driving the spindle?

---

Yes. Normally it is best to configure the Spindle as a motor Axis and connected to any of Kanalog's 8 +/-10V DAC Outputs. That way it can be controlled with normal Axis Jog Commands to control the speed and this will allow you to control the acceleration as with any motor Axis. See the example SpindleMach3Jog.c as to how Mach3 may control a Spindle Motor Axis using Jogs.

If the Spindle has a quadrature encoder that can be used for full closed loop control. If the Spindle does not have feedback it can still be used as an open loop motor axis by using a velocity feed forward only method described here:

KMotionCNC Spindle Control

and shown here:

Attachment 380666

Quote:

---

Tool changer wise, I guess this is wired up as a basic relay output?

---

It depends how your Tool Changer works. If it can all be controlled with a few relays then you would need to interface the relays to KFLOP/Kanalog and code up the sequence to do the Tool change. Kanalog has 8 24V Relay Driver Outputs.

Quote:

---

If ordered tomorrow, will the units be with me for Christmas? Delivery to UK.

---

We provide USPS International Express Mail at a flat rate of $50USD. I think they claim 3-5 days. But I think it is often a few days longer and then whatever custom delays there are. But I would expect it to arrive before Xmas.

Regards

Thanks Tom I'll look into it when my Kflop & kanalog arrive :D

I've been going through wiring again tonight.

It turns out each TDM servo drive receives not only a tacho feedback, but also a rotor position signal (BLC1, BLC2, BLC3) - see attached. From what I can gather these signals simply tell the drive what degree of rotation the servo motor is at (120degree intervals), therefore the drive can optimise current and torque. I'm guessing this rotor position detection is NOT a form incremental or absolute encoder, and does a completely separate job? My understanding is that my drives are tacho only.

But each axis including the spindle motor, does have an encoder mounted separately from the motor. These encoders are not linked to the drives, and I presume go back to the NUM720.

Just to clarify Tom as I'm confusing myself.....

If I replace the NUM encoders on the end of the ball screws (belt driven for spindle encoder) with encoders suited and wired to the Kflop/Kanalog, can I leave my drives as they are now with tacho and rotor position feedback? So encoder feedback to the Kflop, which then outputs the analog signal to the drives. Then the drives look after themselves as they currently are doing. Is this right?

Thanks,
Jim


http://www.cnczone.com/forums/attach...d=380882&stc=1