[frontrange]

I've seen the section of the manual that talks to glass scale feedback with stepper motors, and done some searches on the topic which lead me to believe that the only supported secondary feedback option using glass scales is with a stepper motor drive - just to be clear, is that correct?

I have a Lagun FTV-2 that was fitted at the Lagun factory with a Servo-II control system that had feedback from both the rotary encoders in the servo motors as well as glass scales on 3 of the 4 axis. (4th being rotary). The system is nice, but every time there is a problem, it costs a small fortune to repair. The last repair bill put me over the top as I have spent $4500 on repairs in just the last year alone.

I'd like to convert the motors and control platform to something less expensive to keep alive, but I'd hate to lose the glass scale feedback that keeps everything on coordinates. I do have precision factory ball screws, so it may be mostly a non-issue, but I'm used to the secondary feedback and don't want to leave it behind in a conversion unless I absolutely have to...

Is this doable with Kflop, or should I look elsewhere?

TIA/Steve

[TomKerekes]

Hi Steve,

I agree it would be a mistake to give up the linear glass scale feedback. With KFLOP you have several options. You didn't mention what type of drives you had or were planning on using.

The closed loop mode you were referring to can close a loop around any Step/Dir drive. This could be Step/Dir Drives for Steppers or Servos. See here.

The most common method is to close the velocity loop in the drive and close the glass scale position loop in KFLOP+Kanalog. In this case the Drive closes the velocity loop using feedback from the motor shaft. The feedback to the drive might be an encoder or tachometer. KFLOP+Kanalog then closes the glass scale position loop by sending analog velocity commands to the amplifier.

It is also possible to do full dual closed loop in KFLOP+Kanalog where both the rotary encoder and glass scale encoder are fed back to KFLOP. This mode consumes two of KFLOP's eight Axis Channels for each motor and of course two encoder inputs. The two Axis channels are connected together within KFLOP. This works out nice because a complete set of axis settings (ie PID+filters) are available for each loop and each loop can be tuned and tested separately.

HTH
Regards

[frontrange]

Thanks for the reply Tom, I guess I misunderstood the context of that manual section when I first read it. I assumed with all of the reference to stepper drive, well I took it literally as only applying to a stepper motor system.

The glass scales are a mixed blessing. Resolving physical position to the grain of the scale gives one bragging rights, but many problems accompany the second set of feedback information. As an example I've just had to replace my X scale because of a subtle problem that ruined a number of parts. I had a single bad spot on that scale that sometimes caused the signal to drop, and the machine would then skip about 50 thou when it hit that spot. I've had similar problems in that past. Sometimes just dead reckoning off of a home coordinate isn't such a bad thing.

As for motors, I'd really appreciate some guidance. The motors I now have are variable reluctance motors, directly shaft coupled and powered off of a grid tied 350VDC power supply with no isolation. They are very powerful but they gotta go.

If I had not already had terrible experience trying to tune a brushless setup, I'd go that way. But the tuning really scares me off. I like the simplicity of steppers, but worry about system performance. Can I feed at least 35 IPM and rapid at least 100 IPM (preferably 150+ IPM) on a machine this size with any sort of stepper configuration?

Then there's brushed DC servo that may work - I guess I'm leaning that way but am open to any suggestions. The machine is up and running now and I want to keep it that way as long as possible. Once I take it down for the swap, I really want to be able to get it all done quickly - thus the heightened anxiety with tuning a brushless servo...

[Need TECH Help!]

Hello Steve,
Only for information, i did a retrofit on a mill of same size as your Lagun with 960 oz/in stepper motors and 2:1 belt pulley reduction. My max rapid was 125IPM, thow i ran it at 100IPM for better reliability. If i did this same size mill again with steppers, i would go with a 1200 oz/in stepper motor. KFLOP was used to close the loop of stepper motors using the Glass Scales on machine. Used Gecko G201 stepper drives.VERY good results.

Another option you might look at is 750watt brushless AC servos, higher in price, but all the power and speed your mill can handle. Did a retrofit with this size AC servo.....wow fast. I think i stopped playing with rapids at 450IPM. And then set them to 150IPM.

The short... to me... a stepper motor system with glass scales closing the loop (on a mill this size) is much easier to tune and swap over from an older system, plus lower cost, with VERY good results. Depending on mechanical state of machine, this setup would be WAY better than factory.

HTH,
Troy

[frontrange]

Thanks Troy - very useful information. As I said, I'd love to go full throttle with brushless servos, but my last experience with tuning sucked. The amps were made back in the late 90's, maybe things are much better now and I worry too much? I don't know.

So with the stepper setup you describe, you went to a max 100 IPM rapid, that would be 1000 motor rpm with the setup you describe assuming my math is right. Do you recall what the torque curve looked like at 1000 rpm for the stepper you used? I was considering the same 2:1 belt ratio if I go that route, or maybe keep it all inline with bigger motors and direct drive as the machine is currently configured.

[Need TECH Help!]

There is some AC servos that have a "AutoTune" for there drives. At the time there was not any auto tune on the AC servos i got but, was well supported and got them tuned in a couple hrs. They were digital amps with easy to follow documentation with phone support. Note: This was my first and only servo retrofit. I have had a little experience with DC brushed servo tuning as a "maintenance" of machines. These AC servos where by far much easier to tune.

Got a typo on my steppers... should be 906 not 960 oz/in. Attached is the torque curve. I also use these same steppers on my Mill/Drill with glass scales and KFLOP of course. . Here is where i got the steppers...http://www.automationtechnologiesinc...95-43-8BT1.jpg
Used to be kelinginc.net.

Troy

[mactec54]

frontrange

It would be a step back, to go with a stepper system after having a servo system, no matter how old they are, I would not worry about the glass scales, as they most likely are no better than .0005 resolution, A good servo system if your ballscrews are good would work better, than what you have with the glass scales,
You could use the Dmm 750W AC Servo motors & Drives, depending on what Drive you went with your machine would do a min of 200IPM with Dyna2 Drive or a Max of 400IPM with Dyna3 Drive same motors for both drives, So if your ballscrews are .200 or 5mm Pitch then these are the speeds you could get with the Dmm system

DMM Technology Corp.

[Need TECH Help!]

I have ran some old machines with some energy hog servos and amps that dont run as well as a correctly set up hybrid stepper of today. I would never do away with the glass scales. For one you can have a dual closed loop with a servo setup, which is a costly option on the "big name" machines. Its always good to have feedback of actual table movement. Regardless of how good your ballscrews are you are still only tracking the servo motor shaft with encoder on servo.

Troy

[frontrange]

I wouldn't call what I have now a real servo system. The motors have encoders and the scales (5 micron) close the loop on the table but as I understand it a variable reluctance motor is a derivative of a stepper type motor. The max rapids this system is capable of is 150 IPM IIRC. The motors are beasts at 33 Lbs each, and no doubt could break lug nuts loose, but they aren't fast.

How difficult is tuning on that Dyna2 Drive system, and what kind of belt reduction do you have in mind for the size and the quoted rapid rates?

[mactec54]

How difficult is tuning on that Dyna2 Drive system, and what kind of belt reduction do you have in mind for the size and the quoted rapid rates?

For a belt drive 1.5:1 is enough 2:1 max, I have used 1:1 on some mills, this will give you 200 IPM, If your Ballscrews are 5mm pitch, with the Dyn2 Drives & the 750W motors, when using the Dyna3 Drive this gives you more flexibility, with gearing ratios & IPM

For tuning this depends on the machine, but most times very little is needed, there are only 3 parameters to change, very simple to set up

[frontrange]

Looking into that DMM system, I don't think I want to go down that route. Looks like it is tightly coupled to their own motors which do not have a NEMA footprint. I Don't want to replace one propritary system with another.

Any other easy to tune servo amps out there?

[mactec54]

frontrange

You won't find a AC servo system as easy as there's to set up, any good servo system will not be any different, from there system, there are very few that you can use different motors with different drives, most motors are much the same, but you have the encoder to deal with, a good encoder for your machine will be 12bit to 14 bit, as far as flanges go they are a standard for servo motors, most are ISO standard & metric

Quote Any other easy to tune servo amps out there?

If you find one, it most likely has Dmm technology inside, or made by Dmm with a different name on it

[keebler303]

frontrange

You won't find a AC servo system as easy as there's to set up, any good servo system will not be any different, from there system, there are very few that you can use different motors with different drives, most motors are much the same, but you have the encoder to deal with, a good encoder for your machine will be 12bit to 14 bit, as far as flanges go they are a standard for servo motors, most are ISO standard & metric

Quote Any other easy to tune servo amps out there?

If you find one, it most likely has Dmm technology inside, or made by Dmm with a different name on it

I've used Baldor Microflex drives on several projects and they have a very nice interface and autotune system. Allen Bradley also has one that works well.

If you're worried about future obsolescence, I would suggest incremental quadrature encoders, they are supported by practically every servo company on earth.

[frontrange]

If you're worried about future obsolescence, I would suggest incremental quadrature encoders, they are supported by practically every servo company on earth.

That was my plan - I was suprised to see the DMM drives do not use a standard quadrature input, instead they opted for an SPI serial encoder.

[mactec54]

If you're worried about future obsolescence, I would suggest incremental quadrature encoders, they are supported by practically every servo company on earth.

Strange you should say such a thing they do have a incremental quadrature encoder, Here is a photo of it, this new type of encoder, is a first for this type in the world & they do have a patent on it

[Al_The_Man]

The larger commercial type controls (Mitsubishi, Fanuc etc) have gone with a proprietary serial interface on their encoders for some time now.
This is fine when both motor and drive is purchased and the PID loop is closed in the drive, but when trying to mix and match drives and motors, and using controllers that are capable of closing the loop external to the drive, then this is where it is useful to use the universal quadrature encoder.
Also with controllers that close the loop, non intelligent drives are capable of being used, in this case the encoder is not required by the drive itself.
Al.

[frontrange]

Strange you should say such a thing they do have a incremental quadrature encoder, Here is a photo of it, this new type of encoder, is a first for this type in the world & they do have a patent on it

The drive manual on the web site shows input only for the SPI encoder unless I missed something.

[Al_The_Man]

Maybe its for an add-on?
I wonder if it is the same 'new' technology as the CUI Inc models?
Al.

[mactec54]

frontrange

There new encoder won't be used on those Drives for some time, but there is no need for change, as the servo loop is closed in the drive,(as most are) they work perfect the way they are, so why change, If you have a control that can use the encoder output, then you would use the Dyna3 drives, which have an output for closing the loop in the control

[Al_The_Man]

, but there is no need for change, as the servo loop is closed in the drive,(as most are) they work perfect the way they are, so why change,

The biggest thing for me is lower cost, less problems for a non intelligent drives, also when the loop is closed in the controller, features such as electronic gearing and electronic cam is possible, which is not possible when separate drives are used.
Al.