Another bridgeport retrofit, need advise

[sohcpwr]

This should help me and a few others out.. I've been researching motors a lot and while I understand the basics, trying to match certain motors to other drives when their not in kit form is a little fuzzy. Can a larger motor be powered by a drive that won't supply current to its rates capacity? You'll obviously not get the rated torque but will it work?

And the doozie: brushed D.C., brushless D.C, and AC servos. DYNs low voltage drives are powered by a D.C. Supply. But their AC motors? How do you know what motor can connect to your particular drive.

Finally, Galils internal drives are either trapezoid communication type or sinosuidal. Can you again here explain how one would know what motor type you can hook up?

[Jim Dawson]

This should help me and a few others out.. I've been researching motors a lot and while I understand the basics, trying to match certain motors to other drives when their not in kit form is a little fuzzy. Can a larger motor be powered by a drive that won't supply current to its rates capacity? You'll obviously not get the rated torque but will it work?

It depends. If the drive can be programmed to limit the current output, then maybe. I have done this with VFDs by limiting the output voltage. But I have only done this for testing, never have actually tried to run a motor in an application this way. It is common to oversize a VFD and cut the max output current back to the motor nameplate current. There is not a lot of difference between a modern VFD and a AC servo drive.

And the doozie: brushed D.C., brushless D.C, and AC servos. DYNs low voltage drives are powered by a D.C. Supply. But their AC motors? How do you know what motor can connect to your particular drive.

That is a good question. All servo drives and VFDs that have an AC input convert the AC to DC internally then output either a PWM square wave in the case of a brushed DC motor, or in the case of brushless DC motors a 3 phase PWM sine wave. Brushless DC motors are 3 phase motors. Some brushless drives (Galil and maybe others) will run both brushed and brushless DC motors.

The motor specs need to be pretty much matched to the drive. Normally you need to match the voltage and the minimum inductance.

Finally, Galils internal drives are either trapezoid communication type or sinosuidal. Can you again here explain how one would know what motor type you can hook up?

The bottom line is that trapezoidal commutation requires that the motor have Hall sensors installed. Hall sensors are not field installable, they are inside the motor housing, buried in the windings I think. Most BLDC motors that I have seen have Hall sensors.

the following excerpt is from Trapezoidal vs Sinusoidal Brushless Servo Amplifiers | Galil

Trapezoidal vs. Sinusoidal Commutation

Trapezoidal commutation is the most cost effective way of controlling a brushless servo motor. It is perfect for higher speed applications and applications where the motor and mechanics will eliminate the torque ripple that occurs during switching current from one phase to the next. Hall sensors are required for Trapezoidal commutation.

Sinusoidal commutation is great for lower speed, direct drive or linear motor applications where the torque ripple of the motor phases needs to be minimized. Since the current to the motor phases are weighted as sine waves, the torque going through the motor is smooth and has minimal ripple. It also allows the mechanics to be simplified because Hall sensors can be eliminated.

[sohcpwr]

Funny I found the same article from Galil later that night. Its clear as mud now, think I had more questions after reading it.

That is a good question. All servo drives and VFDs that have an AC input convert the AC to DC internally then output either a PWM square wave in the case of a brushed DC motor, or in the case of brushless DC motors a 3 phase PWM sine wave. Brushless DC motors are 3 phase motors. Some brushless drives (Galil and maybe others) will run both brushed and brushless DC motors.

The motor specs need to be pretty much matched to the drive. Normally you need to match the voltage and the minimum inductance.

So what I think I am putting together from this research is that brushless motors are when things get interesting. Is it because the drive can read the position of the commutator from the brushes so things get "in phase" automatically with a brushed motor. Where as brushless motors have no contact so they need some external device (hall sensors) or fancy computer control to detect where it is in the magnetic cycle?


The most unclear, for me anyway, is AC motors. All of them seem to be brushless, but are they really running on AC current from the drive? Or does the drive convert AC to DC for these too? If thats the case how is it different from a brushless DC motor? You can tell me to shut up at any time

[Jim Dawson]

Funny I found the same article from Galil later that night. Its clear as mud now, think I had more questions after reading it.

So what I think I am putting together from this research is that brushless motors are when things get interesting. Is it because the drive can read the position of the commutator from the brushes so things get "in phase" automatically with a brushed motor. Where as brushless motors have no contact so they need some external device (hall sensors) or fancy computer control to detect where it is in the magnetic cycle?

A brushed DC motor is commutated by the........wait for it..........COMMUTATOR on the armature of the motor. The drive has no idea where the motor is at in the cycle, the drive just sends a pulsed DC voltage to the motor, and the motor figures out what to do with it. If the motor isn't doing what the drive or computer wants, then the drive sends more or less power as needed.

You are correct, brushless motors have no contact so require that the computer in the drive figure it all out with feedback from Hall sensors or some other method. Many VFDs are sensorless vector drives, and through some electrical magic figure out how to do it. That is above my pay grade. I have never seen a sensorless vector servo drive, probably not tight enough control without positive feedback.

The most unclear, for me anyway, is AC motors. All of them seem to be brushless, but are they really running on AC current from the drive? Or does the drive convert AC to DC for these too? If thats the case how is it different from a brushless DC motor? You can tell me to shut up at any time

All of the BLDC drives and VFDs convert DC to 3 phase AC output. In the case of VFDs, the incoming AC power is first converted to DC, then that is fed to the drive circuitry to create the 3 phase AC output. BLDC drives could have AC or DC input depending on what you buy.

[sohcpwr]

I found this to be a pretty good read and clear up most of the confusion between AC and brushless DC motors:

https://www.monolithicpower.com/Port...ndamentals.pdf

[Jim Dawson]

That is the clearest explanation I have seen, simple enough even for my feeble brain. Thank you for posting.

[sohcpwr]

Not a lot of updates but much going on behind the scenes. I spotted a deal on EvilBay and after much consideration, went for it. This seller had 3 Galil DMC 4080 controllers with the internal amp option. After going back and forth between a few different resources, all I could narrow down on the internal amps is that they are either 500w, 600w, or 750w servos. This coming from the input voltage listed on the board. Will have to wait until it gets here to power it up in boot mode or something to see everyting about how it was configured.

Initially even the 500w sounded like it would work, not to mention being a total steal getting 8 servo drives with the controller. But after some reading through the manual I found out they are 7amp continuous and only 10amps max. A typical bridgeport sized servo motor pulls around 30A to get the rated torque I presume. I figured I cant really go wrong at the price I got it for so I am going to give it a try. Worst case I have to buy a couple drives. The internal ones will still work for my quill axis and the "A" axis at bare minimum.

Here is the unit:


And a few other things I purchased through this seller's store: Dude was super cool, I dunno if it was the canadian conversion or just the fact that I already gave him $1000 for the 4080 but he sold me all of the following for less than $100









And 6 of these:

[sohcpwr]

Next step is to decide between these 2 for axis breakout boards:





The latter being considerably smaller, but the former having the option of a small PCB on top in case you had to modify the signals. I could see this being useful somehow... The next next step is to decide on motors. The internal amps are not made for AC servos, and I would really like brushless DC ones, but not really having any luck finding the right combination of price, voltage, and torque. If anyone has any suggestions please chine in!

Jim, servos with encoders will still work with your software right? Or is it only setup for a tach on them and encoder feedback from the axis? One more question, I found a pretty good deal on some high quality Euchner limit switches. The manual for the DMC sounds like I need 3 switches for each axis, + - and home. Is this correct? They have triple units available but the one I'm looking at is a double.

I forgot, also having a real hard time finding an enclosure in the size I need. Something like 14-15" wide by 20-30" tall. I wish I woke up earlier today I could have taken a ride to the local machinery resale place. Worried I'll get something full of holes for switches in all the wrong places that way.

[Jim Dawson]

Next step is to decide between these 2 for axis breakout boards:

The latter being considerably smaller, but the former having the option of a small PCB on top in case you had to modify the signals. I could see this being useful somehow... The next next step is to decide on motors.

You are going to need both, the 26 pin connectors have the encoder and command signals, the 44 pin has the I/O

Take a look at these also https://www.winford.com/products/cat_brk.php not the highest quality screw terminals, but I am using them.

The internal amps are not made for AC servos, and I would really like brushless DC ones, but not really having any luck finding the right combination of price, voltage, and torque. If anyone has any suggestions please chine in!

AC servos are all BLDC motors and all BLDC motors are 3 phase AC. Confusing isn't it. The internal amps will run AC (BLDC) servos.

Jim, servos with encoders will still work with your software right? Or is it only setup for a tach on them and encoder feedback from the axis? One more question, I found a pretty good deal on some high quality Euchner limit switches. The manual for the DMC sounds like I need 3 switches for each axis, + - and home. Is this correct? They have triple units available but the one I'm looking at is a double.

Encoders on the motor will work just fine with my software. It really doesn't care where the encoder signals come from, but keep in mind that the accuracy of the machine is dependant on the table position. I'm a real fan of putting the encoders on the load (table). Tach signals would only be found on brushed DC motors, and that signal is fed back to only the drive for internal speed control.

On the Z axis depending on how you want to run the machine, upper and lower limits are needed, but the home switch is optional. I am using only one switch on my Z axis also, triggered by what used to be the downfeed kick out mechanism. On the X/Y, a home switch is not needed unless you really want to home those axes, and I can't see a reason to do so. My software can set a parking position and/or zero anywhere you want with a mouse click. Only one travel limit switch is really required for the X and Y axes, you just put an operating cam on each end of the travel to hit the switch. I have pictures if needed.

I forgot, also having a real hard time finding an enclosure in the size I need. Something like 14-15" wide by 20-30" tall. I wish I woke up earlier today I could have taken a ride to the local machinery resale place. Worried I'll get something full of holes for switches in all the wrong places that way.

To mount the 4080 in, you need at least 24x30x6. Don't short yourself on panel realestate. Again, I have pictures if you need them.

[sohcpwr]

AC servos are all BLDC motors and all BLDC motors are 3 phase AC. Confusing isn't it. The internal amps will run AC (BLDC) servos.

For some reason when I read the article I posted earlier I was thinking that an AC servo was what they call a permanent magnet synchronous motor. Thats awesome I can use AC servos, like I said in the beginning the plan was to use the DMM ones. But unfortunately after reading their documentation, they wont work with Galil internal amps unless I hack up a brand new motor and swap out the encoder. Just did some research and it doesnt look like mine was ordered with the absolute encoder option.

To mount the 4080 in, you need at least 24x30x6. Don't short yourself on panel realestate. Again, I have pictures if you need them.

The controller is only 11.5" long, is xtra space required around the side of it? I wanted to mount the box to the side of the column and if i remember correctly its flat part is 16-17" at its narrow part at the top. I dont mind going longer, but wanted to keep the width inside the column footprint.

[Jim Dawson]

The DMM motors have absolute encoders? Those are not very common, I have never actually seen one.

The problem is that you need room for the cabling. Here are some pics of the Shizuoka panels. You won't have quite as much stuff because you're not using near as much I/O One thing you could do is plug the breakout boards directly into the Galil board, that would save a lot of real estate. No cabling that way.

Pardon the rats nest, I was looking for a wire the other day to make a wiring change. The Galil panel 24x20x6



The power panel 24x30x8

[sohcpwr]

From DMM's website...

he ABS-16 series absolute encoder is used for all DHT and DST series servo motors. 16-bit [65,536ppr] resolution and ultra-fast proprietary serial update to servo drive.

The ABS-16 encoder is suitable for OEM applications.

I just sent them an email, they also have an incremental encoder that uses a lot of the same parts as the one above. I would think the footprint would be similar and possibly adapted to their new servos. Its definitely not a factory option, but who knows maybe they can sell me an encoder-less motor and this incremental one in its place and I'll find a way to marry them.


As far as the enclosure goes, I was planning on putting the controller at the top on the back panel not the door. And hopefully finding <1' cables to make a ysmall loop to the boards. I will admit tho that I have yet to actually put some dimensions in Inventor/Acad yet and see if it is feasible.


Your method of limit switches is what I was hoping to be able to do. Since the controller had inputs for all of them, I wasnt sure if it was required for proper operation. Regarding homing/home switches- The manual for the Galil I think was saying using brushless motors (w/o hall sensors) you have to move them on startup and I was guessing homing usu accomplished that task? It sounds like your software is just doing this without actual sensors though huh? If you don't need them I prob dont either

I do still plan on getting the renishaw or equivalent encoders on the machine's axis. I thought if things start to go over budget, this would be one of the things to push on the "ill do it later" list.

[Jim Dawson]

You might also look at these motors. I'm thinking that the M-2311 and a 3:1 timing belt reduction would work well. These seem to be a lot of motor in a very small package, you'll have to look at the torque curves. https://www.teknic.com/products/huds...hudson-motors/

I can tell you that my motors at about 500 in/oz are way overkill, I only use a fraction of their max power.


On startup the motors only have to move a very small amount for the controller to figure out the commutation. If possible Hall sensors are the best way.

[sohcpwr]

Your motors are 500oz/in peak or continuous torque? Those are certainly a lot of punch in a tiny package... A nema 23 on a full sized bridggy would almost look weird. I'm all about smaller components though. Its odd they don't list the voltage or amperage on the spec sheet. The Vpeak/kRPM says 13 and 6 for the two different inductance options. A 30V power supply for the second one???

[Jim Dawson]

Your motors are 500oz/in peak or continuous torque?

The nameplate torque is 32 in/lb (512 in/oz) so I assume that is continuous. My timing belt drives are 1:1 as I recall, but they could be 1.5:1

Those are certainly a lot of punch in a tiny package... A nema 23 on a full sized bridggy would almost look weird. I'm all about smaller components though. Its odd they don't list the voltage or amperage on the spec sheet. The Vpeak/kRPM says 13 and 6 for the two different inductance options. A 30V power supply for the second one???

In the Galil manual there are all the formulas for calculating the voltages from the inductance and resistance. I can't remember all of that off the top of my head. Given that they are rated at 5000 RPM, with a 0.250 lead screw and a 3:1 reduction that would give you max rapids at 400 IPM. Way too fast for a BP size machine IMHO.

BTW, for cables I like thse guys http://www.cablesondemand.com/pcateg...UB)_CABLES.htm

[sohcpwr]

The manual for the DMC 4080 says the controller can provide sinusoidal commutation if an external amp does not. But it doesnt say anything about setting the controller to sine and using the internal trapezoidal drives. The specs for AMP D3040 says it can only control brushless motors that have hall sensors. Since none of these AC servos have them, do you think the controller can be configured for sine drive and still work with the internal amps? I would think if it was possible, there would be a note in the specs section that says something like "Brushless motors without hall sensors may be used by setting the controller to sinusoidal commutation." Have you wired AC motors to the internal drives before?

I found some great deals on Ebay for 500w Fuji and Mitsubishi servos. About $100 ea with the correct voltage/speed/amps and 8Nm peak tq.

[Jim Dawson]

The only AC servos that I have used with the internal drives have had Hall sensors. Never tried it without. To answer your question properly, I would have to get deep into the manual. Look at the ''BX" command in the command manual. Also look through all ''B'' commands, there is a lot of reading there. Pretty soon I'm going to be asking you questions I have also used the internal drives to run brushed DC motors, you connect A and B phases as I recall, the manual describes how to set it up.

Once you get the unit, then we'll figure out what drives are in it.

[sohcpwr]

Balls, using the controller to provide sinusoidal commutation requires 2 axis input to each amp. I dont think ill be opening it up and doing any homemade re-wiring!

Pretty soon I'm going to be asking you questions

Jim I saw your thread on that other machinist forum the other night about the shizouka conversion. Lets just say even with my insatiable quest for knowledge about how all these systems work and relate to each other, I am a looooooonnnnngggg way away from giving you any kind of practical advise.

The search for a brushless motor continues... DMM got back to me today (on a sunday btw very impressed) they can sub out the absolute encoder for an incremental one. I read the spec sheet for it, which might have solved one problem and created another. It has hall sensors in it that it uses for some kind of control algorithm. Not sure if these signals are accessible or not, waiting on a reply back. Downside is that its a differential encoder, another option the board does not have. Have to go back and re-read but I vaguely remember being able to still use just 2 channels of it.

[Jim Dawson]

Balls, using the controller to provide sinusoidal commutation requires 2 axis input to each amp. I dont think ill be opening it up and doing any homemade re-wiring!

I think that says it has 2 DACs in the controller to handle that. From the BQ command remarks: ''When using an internal Galil sine drive, each axis has two DACs (Digital to Analog Converter). BO sets the first DAC offset. BQ sets the second.''

Jim I saw your thread on that other machinist forum the other night about the shizouka conversion. Lets just say even with my insatiable quest for knowledge about how all these systems work and relate to each other, I am a looooooonnnnngggg way away from giving you any kind of practical advise.

Thank you for the kind words

The search for a brushless motor continues... DMM got back to me today (on a sunday btw very impressed) they can sub out the absolute encoder for an incremental one. I read the spec sheet for it, which might have solved one problem and created another. It has hall sensors in it that it uses for some kind of control algorithm. Not sure if these signals are accessible or not, waiting on a reply back. Downside is that its a differential encoder, another option the board does not have. Have to go back and re-read but I vaguely remember being able to still use just 2 channels of it.

On the contrary, the differential encoder is prefered, almost noise immune, and the Galil has the inputs for it. The signals are A, /A, B. /B, and some encoders have I, /I for the index pulse.which is also standard on the Galil board.

[sohcpwr]

Found another great piece of reading (or actually listening in this case). Galil's website has a bunch of great videos on all different topics of motion control. If anyone else is as interested in this stuff as I am, its worth your time to go check a few out. This link gave the real concrete answers to brushless motors and their commutation methods.

Access Denied / User log in | Galil


Basically he says that the motors are identical (other than the AC one usually lacking hall sensors) but its really the method of commutation that makes it a brushless DC or brushless AC servo. A brushless DC will use hall sensors and trapezoidal commutation, where an AC one will use sinusoidal. Further on he explains why 2 different analog outputs are required if the controller is to generate the commutation. Only 2 are necessary because the sum of the 3 phases should be 0 so the final phase is easily calculated by the amp he says.

''When using an internal Galil sine drive, each axis has two DACs (Digital to Analog Converter). BO sets the first DAC offset. BQ sets the second.''

So I feel weird disagreeing with you, but I'm pretty sure the BA command is for when your actually using one of the Galil internal amps that is of the sinusoidal type. I think the command manual is saying that the Galil sine amps do not require 2 axis to control the motor by using the 2 DAC's you speak of. But some 3rd party external sine amps might actually require 2 analog inputs to generate the 3rd phase. And any trapezoidal amp will not work with the controller providing the sine commutation due to the difference in construction of the amplifier itself.

I hope I'm wrong but I think my motor search just got narrowed down again. Time to try and find some AC servos that have hall sensors, Jim do you remember which ones you've worked with in the past?