[engnerdan]

I have had many inquires about how to do this, and a few people telling me it can't be done. In the past I promised to do a full write up on how to run a high speed induction motor like a Fanuc on a common VFD, but its time to admit I just had too much going on in my life and that is a low priority. So this thread is starting off as a information "dump" of everything I have figured out so far and what I have accomplished along with the documents I have gathered.

I have my 8000 RPM Fanuc spindle motor running on a Hitachi VFD in sensor-less vector mode right now. I would say the setup is 80-90% done.
The good
-Very stable RPM operation
-Good amount of power
-Very smooth operation
-Very quiet

The less glamorous
-My power meter shows a high kW draw across most of the RPM range, like 1.5kW to 2.2kW free spinning (with and without the spindle belt in place).
-Motor runs a bit warm for my headstock enclosure at 8,000 RPM for an extended period of time (hours)
-Occasional error on the VFD during braking (yes I have a braking resistor)

Google docs link to all the files I have compiled
https://drive.google.com/open?id=0B-...3FuNndMUVhxeTQ



Cliff notes type summery of what I have done:
Here is the label of my Fanuc spindle motor.
Attachment 338724

What I took from the label is this

RPM range 1500-8000 which corresponds to 50hz to 266.67hz
Voltage range 110-216volts which I took to correspond to the above RPM and frequency range.

Attachment 338726

This was my starting point, with this I did my first test with a manual voltage/frequency scale in the VFD and this worked to get the motor turning across the RPM range without overloading.
Here it the motor running under this configuration
https://www.youtube.com/watch?v=UeOB3rCyNLM


I then moved on to use the 50hz as the base frequency and 267hz as the max and sensorless vector control. This is where I am at now.
Here it the motor running under this configuration
https://www.youtube.com/watch?v=WKS98U51bXo


My next test will be to set the base frequency at 267hz and just let the VFD apply whatever V/F curve it wants. This is basically what a person does with one of the high speed, water cooled, Chinese spindle motors so I think this will work too and be much simpler. If this works than I believe it should also work with many of the other high speed induction spindle motors out there, like Yaskawa, and Lincoln.


I know this is cutting it very short, but its either this or I will never get this posted and never be able to share the information.
-Dan

[mactec54]

I have had many inquires about how to do this, and a few people telling me it can't be done. In the past I promised to do a full write up on how to run a high speed induction motor like a Fanuc on a common VFD, but its time to admit I just had too much going on in my life and that is a low priority. So this thread is starting off as a information "dump" of everything I have figured out so far and what I have accomplished along with the documents I have gathered.

I have my 8000 RPM Fanuc spindle motor running on a Hitachi VFD in sensor-less vector mode right now. I would say the setup is 80-90% done.
The good
-Very stable RPM operation
-Good amount of power
-Very smooth operation
-Very quiet

The less glamorous
-My power meter shows a high kW draw across most of the RPM range, like 1.5kW to 2.2kW free spinning (with and without the spindle belt in place).
-Motor runs a bit warm for my headstock enclosure at 8,000 RPM for an extended period of time (hours)
-Occasional error on the VFD during braking (yes I have a braking resistor)

Cliff notes type summery of what I have done:
Here is the label of my Fanuc spindle motor

What I took from the label is this

RPM range 1500-8000 which corresponds to 50hz to 266.67hz
Voltage range 110-216volts which I took to correspond to the above RPM and frequency range.

This was my starting point, with this I did my first test with a manual voltage/frequency scale in the VFD and this worked to get the motor turning across the RPM range without overloading.


I then moved on to use the 50hz as the base frequency and 267hz as the max and sensorless vector control. This is where I am at now.
Here it the motor running under this configuration


My next test will be to set the base frequency at 267hz and just let the VFD apply whatever V/F curve it wants. This is basically what a person does with one of the high speed, water cooled, Chinese spindle motors so I think this will work too and be much simpler. If this works than I believe it should also work with many of the other high speed induction spindle motors out there, like Yaskawa, and Lincoln.

The most important thing you should of done from the start was to set the max frequency and base frequency, this may be why you are getting more heat than you expected, 50Hz should not even be in the picture, unless that is the minimum frequency parameter setting

No matter what 3ph Ac motor you are running with a VFD Drive, having wrong Frequency Parameter settings will damage your motor

[engnerdan]

mactec54, Did you even bother to read my full post before hitting reply? I appreciate helpful feedback, and even constructive criticism. But I am beginning to feel like your goal is to refute everything I post about motors and drives.

The most important thing you should of done from the start was to set the max frequency and base frequency, this may be why you are getting more heat than you expected, 50Hz should not even be in the picture, unless that is the minimum frequency parameter setting

No matter what 3ph Ac motor you are running with a VFD Drive, having wrong Frequency Parameter settings will damage your motor

If you had read everything you would have found that I had the base frequency and max frequency set the whole time. "I then moved on to use the 50hz as the base frequency and 267hz as the max". Which I had actually done on the first test with a free form voltage/frequency curve. So can you tell me how the 50hz doesn't come into play? That is the base frequency I used. Without formal specifications from the manufacturer one has to make an educated guess based on the information on the label, hence the information I presented.

So to clarify for you, the heat is not a big issue as there is a thermal sensor built into the motor and connected to the drive. But the heat I believe is caused by a combination of two factors. 1. The very small enclosure on my mill around the motor that causes the hot air to basically just circulate around continuously. 2. I have the constant torque mode/ Variable torque mode parameter set to constant torque, which I think is a mistake but I have not had time to read through the VFD manual again and see if this is automatically set by using the sensor-less vector setting or if it was a error on my part. Additionally while the motor is warmer than I would like, it by no means is anywhere close to overheating and I have 20-30 hours (maybe more) on it in this configuration without a problem. Looking at the manual I now see that I can't use Variable Torque mode when using sensor-less vector control.
Attachment 338806

I could set the base frequency and the max frequency both to 266.7hz but that will provide a flat line voltage/frequency "curve" which is not ideal. The frequencies below the base frequency are the constant torque area and the frequencies above at considered constant horsepower. So that could cause either poor performance at anything but max speed or severe overheating at low speeds.
Attachment 338804

-Dan

[mactec54]

mactec54, Did you even bother to read my full post before hitting reply? I appreciate helpful feedback, and even constructive criticism. But I am beginning to feel like your goal is to refute everything I post about motors and drives.

Exactly correct, why because this is not difficult to do or understand how it works, you just posted some information, do you understand it and how it works, then you need to setup you motor name plate spec's with the right parameter settings, it is very simple, I have done many of these spindle motors, have about 10 in stock from 3Hp to 15 Hp all new Yaskawa motors, have had some of the Fanuc motors as well, they all program the same way, I always use the encoder feed back to the VFD drive, if you don't have this feature, you are missing a big part of what these motors are all about, and why they are used the way they are

[mike_Kilroy]

Exactly correct, why because this is not difficult to do or understand how it works, you just posted some information, do you understand it and how it works, then you need to setup you motor name plate spec's with the right parameter settings, it is very simple, I have done many of these spindle motors, have about 10 in stock from 3Hp to 15 Hp all new Yaskawa motors, have had some of the Fanuc motors as well, they all program the same way, I always use the encoder feed back to the VFD drive, if you don't have this feature, you are missing a big part of what these motors are all about, and why they are used the way they are

Mactec, you are not correct at all and should not be stating such stuff. The Fanuc motor capable of 8000rpm max speed is very unique in its voltage ratings. If you set it up as you suggest, you will loose a large portion of the HP capability - even if you can get to 8000rpm.

And no, he looses nothing of value not using the proprietary Fanuc serial output encoder - unless he needs to do tapping at very low speeds.

Please stick to stepper motors where I am sure you are an expert.

engnerdan, I have a teamview session with a customer @9am that may take a couple hours on servo tuning and such, but when finished I will post why both methods you have used so far leave a lot to be desired and how to do it to get full motor HP.

[engnerdan]

engnerdan, I have a teamview session with a customer @9am that may take a couple hours on servo tuning and such, but when finished I will post why both methods you have used so far leave a lot to be desired and how to do it to get full motor HP.

Mike,
I am quite intrigued, and look forward to the information you have. I spent a fair deal of time trying to find engineering level information on these high speed motors and always came up empty handed.

I never expected this setup to be perfect, but it is much more affordable than a entire factory Fanuc motor and drive setup.

Some more information on my setup, I have tested down to 50RPM up to 8,000 RPM with this motor. If I remember the low end was low on torque, but the voltage going into the motor is very low like in the teens (read via Modbus from the VFD) so I think if I tweak the torque boost voltage up some I can improve that. I have retrofitted a US digitial 500 line optical encoder to this motor in place of the factory magnetic encoder, which is how I have verified the RPM. Its not a quantifiable number but I have taken some decent cuts with my 2.5" facemills in both aluminum and steel without a hiccup, which would have been a problem for the previous Chinese 6000RPM/2HP motor.

The factory position device had a reluctor wheel and a dual pickup coil which output a A and B signal in sinusoidal form, so would that be a resolver?

-Dan

[Al_The_Man]

The factory position device had a reluctor wheel and a dual pickup coil which output a A and B signal in sinusoidal form, so would that be a resolver?
-Dan

Fanuc and Mitsubishi/Mazak have used this system on their spindles.
It is not exactly a resolver, which is a 3 winding device.
A resolver does however output a sine/cosine signal.
Even optical quadrature encoders start out with two A B sine waves, on some such as Heidenhain encoders this sine wave is used as an high definition encoder by using the arc-tangent math function.
The other style squares up the sine waves and makes them the more common dual pulse quadrature version.
You could do the same thing with the two AB signals you have. But they are usually very coarse signals.
They are usually used for velocity FB rather than position on a spindle.
Al.

[mike_Kilroy]

Sorry for rushing, but paying customers you know....

So I drew a quick what you get with your two methods of settings: I did not even bother with the other person's settings as they would simply overload the hitachi and cause it to shut down on overcurrent with 50hz/240v base speed settings as you know.

Sorry for the paper I used: we have put on servo/spindle training classes for the R&D engineering department at corporate P&G and I stole a bunch of cool paper for stuff like this...

So your first way was to sete base as 110v & 50hz. This is the dotted line of course in the pix. So from 0-1500rpm you get all the motor is rated for, but above 1500 you loose everything! you are stuck 110v@ all higher speeds so your full rated output goes down from 2.2kw to 0.97kw max! look at all the area above the curve you lost.

Your second way was to set base at 266hz/8000rpm: again, look at all the HP you throw away! it is much whimpier than method 1 from 0-1500rpm, but does gain at higher speeds.

Remember the PURPOSE of the chinese POS hi speed spindles: MACHINE AT 18000RPM! So they don't CARE that you loose all that capability from 0 to 10,000rpm....

What YOU want is to set base at 110hz & 240v and GET IT ALL. NOw you can machine steel and stuff from 1500-5000rpm and not have whimpy-ville.
this is 50hz*240v/110v. same v/hz curve but don't give away the store.

You mentioned wimpy at 50 rpm: SHOULD NOT BE - i suspect that was with the 8000/266hz base settings. We sell Hitachi. It should be able to produce full rated torque down to like 2 hz on sensorless vector....

Adding an encoder will just get the 0-2 hz full quick torque back - and position if you need tool changer or something. Otherwise it should not make your performance any better.

I will not go into how from about 5000 to 8000rpm you loose potential torque due to breakdown torque area of the motor - more complex. Not appropriate for this discussion. Just realize that when we design high speed spindle motors, we MUST use reduced voltage at "base speed" in order to not get bit by the V^2 loss of torque at the high end.

'snuff for now - gotta get back to it. It is again miller time and I get to play with my 40 mile range 16db M900 rocket 900mhz links to get cable to my house thru 1 mile of solid trees back in the woods.... the important stuff!!! Hope you enjoy this.

[mactec54]

Mactec, you are not correct at all and should not be stating such stuff. The Fanuc motor capable of 8000rpm max speed is very unique in its voltage ratings. If you set it up as you suggest, you will loose a large portion of the HP capability - even if you can get to 8000rpm.

And no, he looses nothing of value not using the proprietary Fanuc serial output encoder - unless he needs to do tapping at very low speeds.

Please stick to stepper motors where I am sure you are an expert.

engnerdan, I have a teamview session with a customer @9am that may take a couple hours on servo tuning and such, but when finished I will post why both methods you have used so far leave a lot to be desired and how to do it to get full motor HP.

You are full of BS as usual, the only part that you have posted that is correct, is the use of the encoder, the rest is BS

[mike_Kilroy]

You are full of BS as usual, the only part that you have posted that is correct, is the use of the encoder, the rest is BS

haha. thank you.

[mike_Kilroy]

engnerdan, just so you understand the fanuc nameplate fully, the line showing 110-216v motor input means just that: they are telling you the 4 pole, 50hz "base' voltage is 110v, but that the motor will accept upto 216vac as frequency continues to increase... I used 240v as it will work to that just also since you have it available. Note the Fanuc drives typically have 200vac fed to them, and they do not like 240v, so they guarantee their motor performance curve to 216vac; they are telling you that the breakdown torque will get in your way at 8000rpm if you have less than 216vac input to the motor. You have 240 available so you should use it all; hence my giving you the 240v ratio frequency for your "real' base speed setting in the drive.

[engnerdan]

Sorry for rushing, but paying customers you know....

So I drew a quick what you get with your two methods of settings: I did not even bother with the other person's settings as they would simply overload the hitachi and cause it to shut down on overcurrent with 50hz/240v base speed settings as you know.

Sorry for the paper I used: we have put on servo/spindle training classes for the R&D engineering department at corporate P&G and I stole a bunch of cool paper for stuff like this...

So your first way was to sete base as 110v & 50hz. This is the dotted line of course in the pix. So from 0-1500rpm you get all the motor is rated for, but above 1500 you loose everything! you are stuck 110v@ all higher speeds so your full rated output goes down from 2.2kw to 0.97kw max! look at all the area above the curve you lost.

Your second way was to set base at 266hz/8000rpm: again, look at all the HP you throw away! it is much whimpier than method 1 from 0-1500rpm, but does gain at higher speeds.

Remember the PURPOSE of the chinese POS hi speed spindles: MACHINE AT 18000RPM! So they don't CARE that you loose all that capability from 0 to 10,000rpm....

What YOU want is to set base at 110hz & 240v and GET IT ALL. NOw you can machine steel and stuff from 1500-5000rpm and not have whimpy-ville.
this is 50hz*240v/110v. same v/hz curve but don't give away the store.

You mentioned wimpy at 50 rpm: SHOULD NOT BE - i suspect that was with the 8000/266hz base settings. We sell Hitachi. It should be able to produce full rated torque down to like 2 hz on sensorless vector....

Adding an encoder will just get the 0-2 hz full quick torque back - and position if you need tool changer or something. Otherwise it should not make your performance any better.

I will not go into how from about 5000 to 8000rpm you loose potential torque due to breakdown torque area of the motor - more complex. Not appropriate for this discussion. Just realize that when we design high speed spindle motors, we MUST use reduced voltage at "base speed" in order to not get bit by the V^2 loss of torque at the high end.

'snuff for now - gotta get back to it. It is again miller time and I get to play with my 40 mile range 16db M900 rocket 900mhz links to get cable to my house thru 1 mile of solid trees back in the woods.... the important stuff!!! Hope you enjoy this.

Mike,
I hope you got your internet connection fixed.

I may have muddled something up in posting the info I did. Here is a better breakdown of how I have configured things. I ended up pulling the parameters from the drive because I was getting things mixed up and it turns out I was mixing things up in my head. I did not realized that on the current configuration I had set the base frequency to 266.6hz, which as I knew and and you have stated screws up the constant torque/constant power breakpoint on the voltage/frequency graph. So I will be going back and fixing that. Plus I just realized I may have messed up the spreadsheets of the configuration parameters I shared (I will fix them shortly).

BTW - this mental mixup might just explain the heating, high kw draw and lackluster performance at 50 RPM.


Test 1:
Constant Torque mode (no encoder feedback)
Base Frequency = 50hz
Base Frequency Voltage = 100 (Done Via Parameter A045, 50% voltage gain at base frequency)
Max Frequency = 267hz
Max Frequency Voltage = 200 volts


Test 2:
Free form Voltage/Frequency (no encoder feedback)
Base Frequency = 50hz
Base Frequency Voltage = 110
Max Frequency = 267hz
Max Frequency Voltage = 160 volts (done as a CYA)


Test 3: - Current Configuration
Sensorless Vector Mode (no encoder feedback)/ Constant Torque (HD Rating)
Base Frequency = 266.6hz
Max Frequency = 267hz
Max Frequency Voltage = 200 volts


Mike what do you think about this instead of the 110hz base frequency setting, I have not checked the Hitachi manual to verify it can be done?

Sensorless Vector
Base Frequency = 50hz
Max frequency = 267hz
Max voltage (AVR) = 200 volts
A045, 55% voltage gain at base frequency = 110volts at 50HZ


-Dan

[mike_Kilroy]

Dan,

Couple things...

1) how are you setting "max freq voltage?" I do not believe there is such a parameter in the Hitachi...

2) I think your AVR settings are meaningless? I think AVR just tweaks the output voltage a small percentage based on the AC input voltage fluctuations? I could be wrong on this...

3) Your final setting proposed test seems to be 50hz base with 200 or 220 or something at that point and will just be high current, oversaturated motor and not run. You either need to set A045 to a lower voltage as you have done OR use the higher base freq that corresponds to 240v as I suggested. I think all your tests are basically at the 3 different curves I drew.

Again too, there is NO REASON to limit your motor to 200v instead of 240v you have available; you are simply giving up that (20%) HP performance. Remember, Fanuc could not handle 240v for many years thus they stayed at their happy space point of 200v just like they have in japan.

[engnerdan]

Dan,

Couple things...

1) how are you setting "max freq voltage?" I do not believe there is such a parameter in the Hitachi...

2) I think your AVR settings are meaningless? I think AVR just tweaks the output voltage a small percentage based on the AC input voltage fluctuations? I could be wrong on this...

3) Your final setting proposed test seems to be 50hz base with 200 or 220 or something at that point and will just be high current, oversaturated motor and not run. You either need to set A045 to a lower voltage as you have done OR use the higher base freq that corresponds to 240v as I suggested. I think all your tests are basically at the 3 different curves I drew.

Again too, there is NO REASON to limit your motor to 200v instead of 240v you have available; you are simply giving up that (20%) HP performance. Remember, Fanuc could not handle 240v for many years thus they stayed at their happy space point of 200v just like they have in japan.

I just looked in the Hitachi manual over my lunch break (didn't have time before).

The A045 setting will not help, I was thinking it would go from the A045 setting*the AVR voltage at the base frequency setting up to full AVR voltage at max fmax. But looking at this chart it does not do that.
Attachment 339516


1). When using the Free form Voltage/Frequency control method (see picture) I can create whatever curve I want, ideally to match the magnetic saturation Voltage Vs Frequency curve. So that is how I was able to set a voltage at the max frequency.
Attachment 339522

2). Here is the AVR settings page, so I have been taking it as the manual says and that it is regulating the peak waveform voltage. I will bump it up, at least to 215 if not 220. I was just going the same or less than what the nameplate listed.
Attachment 339524

3). No, what I thought A045 was doing and what it actually does are two different things. What I thought at the time of that post was 50hz/110 volts (.55*200), and 266.66hz/200 volts. But as I already covered I was wrong.

Mike here is a plot I made, using your 110hz suggestion, my take on the basic version of an ideal voltage/frequency (basic because I think the true ideal needs to actually form a curve). The advantage your 10hz setting has is I can still use the sensorless vector setting. But it might be a little high on current close to the 110hz area. In order to try my plot I have to take the drive out of sensorless vector mode and go back to free form voltage/frequency.
Attachment 339532

Here I added in my three test configurations.
Attachment 339530



I hope everybody reading long realizes this is a "working" thread, as in all this stuff is still being worked out. I started this to share my current progress and maybe help someone get a start and or get some help from others as I am terribly short on time over the last year.

-Dan

[mike_Kilroy]

I just looked in the Hitachi manual over my lunch break (didn't have time before).

The A045 setting will not help, I was thinking it would go from the A045 setting*the AVR voltage at the base frequency setting up to full AVR voltage at max fmax. But looking at this chart it does not do that.

I was not aware you used the custom v/hz curve maker: by doing that you got what you expected! Good job.

But do realize: A045 DOES do as I said and as you originally thought! It limits the voltage MAX at base speed hz setting of A003 or whatever it is! AND YOU WILL NEVER GET ANY MORE VOLTS OUT THAN THAT! So, had you NOT used the custom v/hz curve routine, you would have limited the MAX voltage output above base speed to the A045 value.


1). When using the Free form Voltage/Frequency control method (see picture) I can create whatever curve I want, ideally to match the magnetic saturation Voltage Vs Frequency curve. So that is how I was able to set a voltage at the max frequency.

YES

2). Here is the AVR settings page, so I have been taking it as the manual says and that it is regulating the peak waveform voltage. I will bump it up, at least to 215 if not 220. I was just going the same or less than what the nameplate listed.

NO COMMENT AS I DO NOT BELIEVE THAT IS HOW IT WORKS BUT IT IS NOT WORTH IT TODAY FOR ME TO ASK HITACHI; Their answer will be confusing.

3). No, what I thought A045 was doing and what it actually does are two different things. What I thought at the time of that post was 50hz/110 volts (.55*200), and 266.66hz/200 volts. But as I already covered I was wrong.

IF YOU DO NOT USE THE CUSTOM V/HZ CURVE, THEN YOU WILL NEVER GET TO 200V OUTPUT.

Mike here is a plot I made, using your 110hz suggestion, my take on the basic version of an ideal voltage/frequency (basic because I think the true ideal needs to actually form a curve). The advantage your 10hz setting has is I can still use the sensorless vector setting. But it might be a little high on current close to the 110hz area. In order to try my plot I have to take the drive out of sensorless vector mode and go back to free form voltage/frequency.
Here I added in my three test configurations.

I AGREE SINCE YOU USED CUSTOM V/HZ CURVE MAKER! Had you not, then your voltages above base speed would be horizontal lines as I showed. I think you are the first person in 20 years outside our office that I ever encountered who used the custom curve!



I hope everybody reading long realizes this is a "working" thread, as in all this stuff is still being worked out. I started this to share my current progress and maybe help someone get a start and or get some help from others as I am terribly short on time over the last year.


-Dan

GUESS my replies ended up inbedded in your post copy above.

Good job!

[engnerdan]

GUESS my replies ended up inbedded in your post copy above.

Good job!

Mike,
What do you think it is a big disadvantage to use the free voltage/frequency setting instead of the Sensorless vector? My biggest concern is establishing the voltage for low speed, too low and I have no torque, too high and I cook the motor.

-Dan

[mike_Kilroy]

Mike,
What do you think it is a big disadvantage to use the free voltage/frequency setting instead of the Sensorless vector? My biggest concern is establishing the voltage for low speed, too low and I have no torque, too high and I cook the motor. -Dan

Not sure of your question: did you mean what or why?

[ramble_on]

nothing wrong with using the custom v/hz curve. although I would scale hz to go upto 240v that you have instead of limiting it. We almost always define the 'base' speed and volts to match the supply voltage rather than leaving that major hole in performance from 1500 to 3270rpm. About the only time we leave it at 1500rpm base speed on these high performance spindles today is for a few aerospace customers who we KNOW try to run at 120% load all day long. If we give them extra HP they will use it and often exceed the mechanical limits of the spindle itself. And for you to leave 20vac on the table unused is not necessary either in my opinion.

On the low end torque, I assume you are not doing tapping or your need at low speed would be a lot stiffer. I would recall how v/hz mode works vs. sensorless vector and then stand back and think on what I need at the affected low speeds...

v/hz of course will droop in speed from basically synchronous speed no load to the rated slip at full load, at ANY speed upto base speed.... Assume it is 50rpm on your motor. So if you tell it to run 40rpm at no load and then load it, it will stop. Tell it 50rpm, and it will stop under load. tell it to go 51rpm and it will run 1rpm at load rating, call for 100rpm and loaded it droops to 50, etc. So if you need torque below 50rpm you need sensorless since hitachi states they will give you full load torque at around 1hz. So command 30rpm and supposedly it will run at 30rpm loaded or unloaded. Below that it falls off. Just the nature of the beast. Decide what you need.

I have not played much with the Hitachi in sensorless mode to know, but I would assume it is not super wonderful for a system running up over 5x the 1500rpm base speed: another reason we let the base speed go up on our fancy drives to the supply voltage - reduces that constant HP range down to, in your case, 2.4x the new 3270rpm base. I can tell you the algorithms in our fancy drives work a LOT better with this smaller field weakening range. Not sure how any of this effects your 50rpm performance tho as we do not typically have those issues - we use feedback when we need low speed performance. Or high speed for that matter. Working on a 70hp system right now that has 6 flute cutter running 500rpm in hard steel: motor goes from 0 to 100nm torque in 10msec each time a flute makes contact. sensorless control would never be able to do this job.

Another thought to confirm you have your v/hz curve set right would be to run 1500 rpm with 110v output no load and see what your current draw is: this is the magnetizing current for your motor. You once said 5 amps no IIRC? Bet that is too low - I would expect your no load to be around 7 amps? But you can measure it this way. Then you know for sure what volts you should be putting out from 0 to 110hz where it flattens out. With this knowledge you will never have to wonder or guess at what current is right from your curve again.

[/rambling_off]

[engnerdan]

Not sure of your question: did you mean what or why?

[ramble_on]

nothing wrong with using the custom v/hz curve. although I would scale hz to go upto 240v that you have instead of limiting it. We almost always define the 'base' speed and volts to match the supply voltage rather than leaving that major hole in performance from 1500 to 3270rpm. About the only time we leave it at 1500rpm base speed on these high performance spindles today is for a few aerospace customers who we KNOW try to run at 120% load all day long. If we give them extra HP they will use it and often exceed the mechanical limits of the spindle itself. And for you to leave 20vac on the table unused is not necessary either in my opinion.

On the low end torque, I assume you are not doing tapping or your need at low speed would be a lot stiffer. I would recall how v/hz mode works vs. sensorless vector and then stand back and think on what I need at the affected low speeds...

v/hz of course will droop in speed from basically synchronous speed no load to the rated slip at full load, at ANY speed upto base speed.... Assume it is 50rpm on your motor. So if you tell it to run 40rpm at no load and then load it, it will stop. Tell it 50rpm, and it will stop under load. tell it to go 51rpm and it will run 1rpm at load rating, call for 100rpm and loaded it droops to 50, etc. So if you need torque below 50rpm you need sensorless since hitachi states they will give you full load torque at around 1hz. So command 30rpm and supposedly it will run at 30rpm loaded or unloaded. Below that it falls off. Just the nature of the beast. Decide what you need.

I have not played much with the Hitachi in sensorless mode to know, but I would assume it is not super wonderful for a system running up over 5x the 1500rpm base speed: another reason we let the base speed go up on our fancy drives to the supply voltage - reduces that constant HP range down to, in your case, 2.4x the new 3270rpm base. I can tell you the algorithms in our fancy drives work a LOT better with this smaller field weakening range. Not sure how any of this effects your 50rpm performance tho as we do not typically have those issues - we use feedback when we need low speed performance. Or high speed for that matter. Working on a 70hp system right now that has 6 flute cutter running 500rpm in hard steel: motor goes from 0 to 100nm torque in 10msec each time a flute makes contact. sensorless control would never be able to do this job.

Another thought to confirm you have your v/hz curve set right would be to run 1500 rpm with 110v output no load and see what your current draw is: this is the magnetizing current for your motor. You once said 5 amps no IIRC? Bet that is too low - I would expect your no load to be around 7 amps? But you can measure it this way. Then you know for sure what volts you should be putting out from 0 to 110hz where it flattens out. With this knowledge you will never have to wonder or guess at what current is right from your curve again.

[/rambling_off]

Mike,
I will be tapping and possibly drilling with 1/2" and larger drills in steel or other hard materials. But all of those operations should be done in excess of 50 RPM, so maybe 50 RPM is a useless data point.

I finally tore into my mill to do a bunch of wiring to finish connecting my operator control panel http://www.cnczone.com/forums/vertic...cnc-posts.html

While I was in wiring I decided to quickly change the base frequency to the 110HZ like you recommended, because I was in the electrical cabinet and knew the parameter to change. I didn't change anything else. I then did a bunch of machining this weekend and everything worked well. A good portion of the machining was all done at 8,000 RPM (using carbide in Delrin and Aluminum I am RPM limited for most everything) and the motor was running cooler than before. One thing I did notice was the RPM droop when loading up with a heavy cut, just like the motor should in open loop. That was something that didn't really happen before with the motor always in the constant torque range of the drive.

I now need to figure out my random error trip on the VFD. I have had this problem for a couple years now, even on my old motor. I can run 30 parts each with 4 tool changes and one part out of the 30 will throw and error or two and then it will not happen again. Then the next part I run it might error out on the same tool but not every time. It always shows up after I change tools and hit cycle start, the Z starts to rapid down and the spindle never starts. A toggle of the reset input and the drive starts up like nothing happened. I don't actually know what the error code is because in order to run my mill the electrical cabinet has to be tight against the wall, so I can't see the display on the drive. I need to get my PLC code fixed so I can read the error codes via MODBUS 485, but that is one other thing that is not working right now.

I have an external braking resistor installed (the lowest ohm resister Hitachi said could be used) which is either a 10ohm or 100ohm (don't recall) 1000 watt import special. Which I doubt the 1000 watt rating I know it checked out at the labeled ohm rating. I have
A051 - DC BRAKING ENABLED = 01
A053 - DC BRAKING WAIT TIME = 0.0
A054 - DC BRAKING FORCE AT DECELERATION = 100
A055 - DC BRAKING TIME FOR DECELERATION = 1 SEC
A098 - DECELERATION TIME = 1 SEC
B090 - Dynamic Braking turned on to 100%
B095 - Dynamic Brake Control (01) - Enabling (disabling while the motor is stopped)

I am looking at setting B130 - "DECELERATION OVERVOLTAGE SUPPRESSION ENABLE" which would fix the error if it is a over voltage condition happening during braking.

My Current parameter settings
Attachment 340884

-Dan

[mike_Kilroy]

Well Done!!

[engnerdan]

Well Done!!

Mike I was stepping though the RPM range of my motor and hit a spot at 800 RPM (700-900) that acts really weird. The RPM jumps around, it sounds terrible (much like a helicopter coming in to land) and it chugs quite a bit and then the drive will trip out.

My educated guess as to what is happening is, there is some sort of harmonic in the sensor-less control loop. Where the drive can't correct fast enough or maybe too fast. I would guess this is the processing speed you are talking about the Hitachi just not being fast enough to handle.

https://youtu.be/1-0H_OWuWjo

What do you think?

Not that it would fix this but I am leaning more towards closing the loop on this spindle. I already have the A and B channels of the encoder fed back into my KFLOP motion controller. I could take the Z index pulse and wire it into the VFD or just go back to Frequency/Voltage control on the VFD and let the KFLOP close the loop.

-Dan