3ph VFD for 3kW mill spindle motor

[OxxAndBert]

Since I'm a VFD virgin, I'm after a sanity check of my plans to install a VFD on the spindle motor of a 1980's TOS mill.

The mill has quite a bit of electrical control. 5 motors (spindle, feed, coolant and 2x lube motors) as well as control for rapid traverse and a supply for the DRO. The electrical cabinet resembles something from the space shuttle. Complexity is low, there's just lots of of it!

I'm trying to make it home friendly with the least modification. I don't have 3-phase at home, but have a decent size rotary phase converter that should be good for over 5HP so have decided to just supply the mill completely off the RPC but use a 415v 3ph VFD for the spindle motor.

I've got access to mains 3ph at work so the plan is to get modified and tested on a known good power source before it goes home.

Main reasons for fitting the VFD are:
- to soften the load on the RPC when the spindle starts
- speed control (max spindle speed is 2000rpm standard. A 50% increase on that - 3000rpm - would be nice)
- possibly some braking/deceleration control

So I think I'm after a 3kW 415v 3ph ==> 415v 3ph inverter with a remote control panel to mount on the existing mill control console (keeping the high voltage power stuff in the electrical cabinet/VFD).
The cable from the inverter to the remote panel would need to be about 2m long so I'm thinking it has to be a CATV style cable or similar rather than the ribbon style.
Would need the ability to add a braking resistor if required, and definitely will also need to do forward and reverse and jog (which I assume is all pretty standard).

Does that sound reasonable?

I'm keen to give one of the Powtran units a go since BobL has given them a good rap and others seem to be having success with them, but a bit confused by their huge range. The particular ones that have been recommended previously in the forum are typically 1ph to 3ph.
Looks like their PI500A model might be suitable for what I need, but I could be way off the mark so any guidance there would be appreciated too.

Steve

[BobL]

One problem you might have is that Powtran don't seem to make a 3kW VFD - they seem to jump from 2.2 to 4kW on all their model ranges.

I have been trying to get a copy of the english manual for their 500 series VFDs for a few months without any luck. I think they are still away on chinese new year holidays so maybe wait a few days before trying to contact them.

[OxxAndBert]

Thanks Bob. I'm not in a hurry - have plenty of other projects on the go so will leave it until next week to contact them.
Although practically I'd probably get away with a 2.2, I'll get a 4kW model.

Steve

[OxxAndBert]

Having a look through the other Powtran models in the meantime. Have a few more questions:
Download links for the manuals etc on the website don't seem to work. Are the links just broken or do you need to be registered to access them?

Is there a particular terminology they use for the CatV style remote cable so I can make sure it has that style?

Input voltage: My RPC is ~400v output but runs quite unbalanced unloaded and may be up around 440v P-P. Is a VFD going to cope with that or is it likely to trigger an overvoltage situation? Possibly it depends on the particular inverter capability which brings me onto....

There seems to be a bit of variation in the input voltage specs for the different 3ph input models.
The PI500A for example is listed as 380v(-15%)-440v(+10%), but the PI160 is 380v (+-10%). On the face of it the PI160 wouldn't be suitable. Is it that simple or is there a bit of Chinglish interpretation in the specs.

Steve

[BobL]

Having a look through the other Powtran models in the meantime. Have a few more questions:
Download links for the manuals etc on the website don't seem to work. Are the links just broken or do you need to be registered to access them?
Their website links have never all been completed - we need to keep in mind that they are not a retail business. They sell VFDs by the container load so they don't particularly care about one of a kind sales.


Is there a particular terminology they use for the CatV style remote cable so I can make sure it has that style?
If you are only using short distance cable you can pretty well use any cable - its the plugs that you have to be compatible with - I will take some pics of mine and post them.


Input voltage: My RPC is ~400v output but runs quite unbalanced unloaded and may be up around 440v P-P. Is a VFD going to cope with that or is it likely to trigger an overvoltage situation? Possibly it depends on the particular inverter capability which brings me onto....

There seems to be a bit of variation in the input voltage specs for the different 3ph input models.
The PI500A for example is listed as 380v(-15%)-440v(+10%), but the PI160 is 380v (+-10%). On the face of it the PI160 wouldn't be suitable. Is it that simple or is there a bit of Chinglish interpretation in the specs.
Yeah I would not use the 160 models in your case.

All my HY VFD's are listed a 220V +/- 10% so 242V is the supposed max. I have panel meters all over the place in my shed so I can see what the mains V is and its almost always over 240V and I often see 248V and even 250V here - some of these VFDs are now 7 years old and still working well.

[jhovel]

Just a quick note about the imbalance of your RPC: the inverter doesn't care at all about that. It rectifies whatever you feed it into a single DC voltage and current anyway. It then chops it into 3-phase sinusoidal AC with controllable/variable frequency. The output will be well-balanced.

[OxxAndBert]

Just a quick note about the imbalance of your RPC: the inverter doesn't care at all about that. It rectifies whatever you feed it into a single DC voltage and current anyway. It then chops it into 3-phase sinusoidal AC with controllable/variable frequency. The output will be well-balanced.

Thanks Joe - good to know. So as long as the particular inverter can handle the max voltage it will get from the RPC all should be fine on the input side.

Steve

[malb]

Just a quick note about the imbalance of your RPC: the inverter doesn't care at all about that. It rectifies whatever you feed it into a single DC voltage and current anyway. It then chops it into 3-phase sinusoidal AC with controllable/variable frequency. The output will be well-balanced.

Not so sure about imbalaced supplies Joe. The input stage is going to be a three phase rectifier in some form. Any rectifier will only conduct current when the instantaneous input voltage applied is greater than the instantaneous voltage of the storage capacitors connected at its output. If the phases are badly imbalanced, the highest voltage phase will conduct for a significantly longer portion of the cycle than for the lower voltage phases. It is quite possible that the lowest voltage phase may never conduct. For the VFD to function properly, the diodes for the highest voltage phase may be seriously overloaded (current and thermally), while those for the other phases are underworked. This could lead to premature failure of the VFD, depending on how much headroom there is in the rectifier specifications.

A lot of the 220-240V input VFD's utilise a three phase input stage that is also compatible with single phase supplies, suggesting that the rectifier block in them has sufficient overhead to cope with all of the supply power being derived from a single phase. In a 380-415V input VFD, the designers may anticipate that the unit would not be subjected to a significantly imbalanced three phase or 380-415V single phase supply, and may not allow the headroom to cater for such in the rectifier block. I note that many supplier offer specific models for use with 480V SWER derived supplies, presumably because they have a single phase rectifier block designed to cope with the higher currents and thermal loads that a single phase input imposes. Being a sceptic, I suspect that they do this because they don't have confidence that the equivalent 380-415V three phase input units would cope well with the effective imbalance that the 480V SWER derived supply represents. Then it comes down to just how tolerant are the rectifier stages of the VFD, and how imbalanced is the output from the RPC. The only effective way to tell is to buy the unit, commission it, and use it. If it has a usable lifespan, then the variables are in your favour, it it dies prematurely, then they were not.

I do not dispute the argument that VFD with an imbalaced input should produce a well balanced three phase output, just point out that the selection of the components in the rectifier block may have a significant impact on the reliability of the system when faced with significantly imbalanced supplies and loadings approaching the devices ratings. One reasonable approach to overcoming the issue may be to to derate the VFD, i.e buy and use a VFD capable of handling a load twice that likely to be imposed, to ensure that the rectifier can handle the increased peak currents and power associated with the imbalanced supply.

[OxxAndBert]

IIRC, going back a few years when 1ph-3ph VFD's weren't so commonplace, it was standard practice to just use a 3p-3p unit of double the power rating and run it off a single phase.

Steve

[malb]

IIRC, going back a few years when 1ph-3ph VFD's weren't so commonplace, it was standard practice to just use a 3p-3p unit of double the power rating and run it off a single phase.

Steve

At moderate power ratings where the rectifier block probably can handle all of the input current coming from a single phase, you may not even have to allow for derating to half rated power out, though it is a good rule of thumb for higher powered units.

The issue with cheaper VFD units is that normally the output voltage is limited to the input voltage, so 240V in generally means 240V 3ph out, which in turn means motors max out at 50% rated power, need to be reconfigured for a different winding connection or need to be rewound/replaced to achieve maximum output. There are some exceptions that can make 415V 3ph from 240V single phase input by use of a different voltage boosting input stage. The more commonly available units were based on Asian VFD's and modified, available from the UK but were more than twice the price of the units they were based upon, and from memory limited to about 2KW output. I am not familiar with the Powtran offerings, so I don't know if the stepup VFD situation has improved any.

[peterbilt]

Download links for the manuals etc on the website don't seem to work. Are the links just broken or do you need to be registered to access them?

Steve

Steve, I found the same problem trying to download Powtran manuals but found them on alternate sites simply by googling e.g. "Powtran PI9000 Manual". Here's one site that I found but there are numerous others: https://www.manualslib.com/manual/1192464/Powtran-Pi9000-Series.html#manual

(https://www.manualslib.com/manual/1192464/Powtran-Pi9000-Series.html#manual)

[OxxAndBert]

Steve, I found the same problem trying to download Powtran manuals but found them on alternate sites simply by googling e.g. "Powtran PI9000 Manual". Here's one site that I found but there are numerous others: https://www.manualslib.com/manual/1192464/Powtran-Pi9000-Series.html#manual

(https://www.manualslib.com/manual/1192464/Powtran-Pi9000-Series.html#manual)

Thanks, found the manual for the PI500 series and have ordered a couple of those.
The Western Union transfer to Powtran is a bit unsettling, but since others have already been down the same road I'm hoping its all good.
Will report back on how it goes.

Steve

[BobL]

Thanks, found the manual for the PI500 series and have ordered a couple of those.
The Western Union transfer to Powtran is a bit unsettling, but since others have already been down the same road I'm hoping its all good.
Will report back on how it goes.

Steve

WU are very reliable - that's why some less than honourable folks have been found to use them :D

[OxxAndBert]

The VFD just arrived. Well packed and labeled and in good condition.

Much smaller than I expected though.

https://uploads.tapatalk-cdn.com/20190313/2c8ff3e71b14b329a238152b1d7aca58.jpg

https://uploads.tapatalk-cdn.com/20190313/cd7c1c9afd463d67ea71a05a70b723c7.jpg

The plug on the removable control module is RJ45 which makes it nice and easy to extend.

https://uploads.tapatalk-cdn.com/20190313/b5c7c7dad59b3d36183405d02b434c6d.jpg

Steve

[OxxAndBert]

Finally had a chance to do some basic settings and hook up the VFD temporarily for testing. Running on real 3-phase supply in the workshop at work for testing, so I know its OK before attempting to run it from my RPC at home.

Started by leaving it at the default 50Hz max frequency, and 10sec accel/decel setting. Starts and stops nicely (but obviously very slowly). I can decrease the speed by winding the pot back on the control panel.
Positive first steps :)

I'll need to be able to run it in reverse, and it would be nice to be able to jog - but can't see a logical way of commanding either of those via the control panel.
How are those things normally done? Set up some form of simple external switch arrangement and find the appropriate terminals in the VFD to connect to or is there a better way?

Steve

[BobL]

I'll need to be able to run it in reverse, and it would be nice to be able to jog - but can't see a logical way of commanding either of those via the control panel.
How are those things normally done? Set up some form of simple external switch arrangement and find the appropriate terminals in the VFD to connect to or is there a better way?

That's it - its all in the manual,
Step one; setup switches and wiring as per page 19 in the manual.
eg DI1 terminal is normally set as default to Run.
One terminal of teh switch gies to DI1 the other goes to COM.
Switch closed VFD starts, switch one VFD stops.
379324

Step 2: Then you have to set F0.11 to 1 so that the external switches will work
(default for F0.11 is 0 - front panel)

Step 3 Set the Correct codes for the DIs
P33 shows what to set the DIs to using F1.00 to F1.08
See how F1.00 is already set at 1 and chart below shows that 1 = Run/FWD,
F1.01 is set for 2 ie REV.
I do this on the same switch with 3 positions, eg up = FWD, middle = stop, down - REV
No D! need be used for a stop.

To set a FWD jog set up a switch connected to any unused DI terminal (other end to com) and set the corresponding F1.0x to 4, or 5 for rev Jog.

379325

[OxxAndBert]

Awesome - I certainly didn't expect that level of detail!!
Thanks Bob :)

My next question was going to be how to control the speed if the front panel was no longer active, but looks like I can set F0.11 to 4 and have the front panel and the terminal block enabled at the same time.
So should be able to still use the pot on the front panel to control the frequency.

Steve

[BobL]

Awesome - I certainly didn't expect that level of detail!!
Thanks Bob :)
My next question was going to be how to control the speed if the front panel was no longer active, but looks like I can set F0.11 to 4 and have the front panel and the terminal block enabled at the same time.
So should be able to still use the pot on the front panel to control the frequency.
Yep

[Ropetangler]

At moderate power ratings where the rectifier block probably can handle all of the input current coming from a single phase, you may not even have to allow for derating to half rated power out, though it is a good rule of thumb for higher powered units.

The issue with cheaper VFD units is that normally the output voltage is limited to the input voltage, so 240V in generally means 240V 3ph out, which in turn means motors max out at 50% rated power, need to be reconfigured for a different winding connection or need to be rewound/replaced to achieve maximum output. There are some exceptions that can make 415V 3ph from 240V single phase input by use of a different voltage boosting input stage. The more commonly available units were based on Asian VFD's and modified, available from the UK but were more than twice the price of the units they were based upon, and from memory limited to about 2KW output. I am not familiar with the Powtran offerings, so I don't know if the stepup VFD situation has improved any.
Certainly not trying to undermine or argue with you Malb, I never have been an engineer, not even his bootlace, but as the OP is going to use a 4Kw inverter on a 3Kw motor, and intends to use very soft starts and stops, I would imagine that the rectifier diodes will most likely remain fairly unstressed. Bearing in mind, a milling machine in hobbyists hands would rarely be fully loaded, (unlike some production machines, which are often loaded to capacity), and the possibility that the rectifiers could be specced for 480V or single phase operation, the chances are that he should be pretty safe. Of course a crash, or workpiece coming loose and causing a jam could change all that in a few milliseconds:o. Hopefully the rectifier bridge is fairly easily replaced if it becomes necessary, and not firmly soldered to the main board, should my theory be proved quite wrong.

[BobL]

There are some exceptions that can make 415V 3ph from 240V single phase input by use of a different voltage boosting input stage. The more commonly available units were based on Asian VFD's and modified, available from the UK but were more than twice the price of the units they were based upon, and from memory limited to about 2KW output. I am not familiar with the Powtran offerings, so I don't know if the stepup VFD situation has improved any.

Powtran still don't make such a beast but Sako does, their SP220 to 3P380 models are called SKI670.

Here is a 7.5kW unit - note the AU$ 415 price and $135 shipping !
https://www.aliexpress.com/item/Sako-7-5KW-VFD-Input-220V-1ph-to-Output-380V-3ph-High-Performance-AC-to-AC/32963416805.html?spm=2114.search0104.3.1.7ed47cb2clgSVb&ws_ab_test=searchweb0_0,searchweb201602_3_10065_10130_10068_10547_319_317_10548_10696_453_10084_454_10083_10618_10307_537_536_10131_10132_10133_10059_10884_10887_321_322_10103,searchweb201603_51,ppcSwitch_0&algo_expid=d9582442-3857-47f2-88d3-13bf7f1e2932-0&algo_pvid=d9582442-3857-47f2-88d3-13bf7f1e2932

4kW for AU$230 and AU$79 shipping
https://www.aliexpress.com/item/Sako-4KW-VFD-Input-220V-1ph-to-Output-380V-3ph-Variable-Frequency-Inverter-for-Motor-Speed/32966135325.html?spm=2114.search0104.3.15.7ed47cb2clgSVb&ws_ab_test=searchweb0_0,searchweb201602_3_10065_10130_10068_10547_319_317_10548_10696_453_10084_454_10083_10618_10307_537_536_10131_10132_10133_10059_10884_10887_321_322_10103,searchweb201603_51,ppcSwitch_0&algo_expid=d9582442-3857-47f2-88d3-13bf7f1e2932-2&algo_pvid=d9582442-3857-47f2-88d3-13bf7f1e2932

Full range here
https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20190509032533&SearchText=SKI670+SAKO+VFD

[OxxAndBert]

Trouble with the "parameter names" in the manual is that they are a bit open to interpretation.

eg F0.02 Frequency command resolution ==> options are 1:0.1Hz, 2:0.01Hz ==> default is 2.

So I think hmmm - that sounds like it would just change the coarseness of the frequency adjustment via the input panel pot.
WRONG!!

It changes the resolution of the frequency that is already set, so if your frequency was set to the default 50.00Hz and you change the resolution setting to "1" you end up with 500.00Hz :no::no:

Things got a bit exciting there for a few seconds....so a smidgen gun shy now...

Trying to set a maximum frequency of 75Hz (already set in F0.19 MAXIMUM OUTPUT FREQ)
And be able to control it via the pot on the control panel (think this requires also setting F0.21 UPPER LIMIT FREQ ==> 75, and F0.20 UPPER LIMIT FREQ SOURCE ==> 3 (Panel encoder)
Does that sound right?

Steve

[OxxAndBert]

For reference, this is what I ended up with to set max freq to 75Hz and allow it to be controlled by the control panel pot while running:
F0.03 FREQ SOURCE MASTER ==> 4 (panel potentiometer)
F0.19 MAXIMUM OUTPUT FREQ ==>75
F0.20 UPPER LIMIT FREQ SOURCE ==> 3 (Panel encoder) **May not be 100% necessary**
F0.21 UPPER LIMIT FREQ ==> 75

Steve

[OxxAndBert]

I just stumbled across this quickstart guide to programming one of these Powtran PI500 VFD's.
It shows a bunch of different scenarios for how you might set up external control etc.

http://www.anlauf.klibo.de/fileadmin/klibo/redaktion/PDFs/bedienungsanleitungen/PI500%20Quickstart%20Guide%20Drive%20Operation%202017.pdf

Steve

[OxxAndBert]

Digging up this older thread as I'm currently in the process of finalising this install.

Regarding external speed control, the manual for this Powtran VFD says to use a 1Kohm 2W pot.
I've currently got it running fine on a 25K pot - not sure what its power rating is but the pot is around 20mm diameter. Its just one I had sitting in my stash so won't be anything special.

Should I bother sourcing a 1K 2W item, or just run with what's working now.
I'd prefer to do it properly from the beginning rather than have having issues in the future.

Steve

[OxxAndBert]

Just in case it helps anyone else, after doing an auto-tune for the motor in vector mode I found I was getting a 06 error (deceleration overvoltage).
Look like the auto-tune changes the deceleration time (F0.14) and my case it was now only 2 seconds.
Not sure what it was originally as I didn't change it and the manual says the default "depends on models", but I increased the value to 6s and the error no longer occurs.

Possibly its a byproduct of doing an auto-tune with the motor still attached to the mill, and with the spindle in gear. I know you shouldn't auto-tune with a load connected, but thought the empty spindle wouldn't be an issue. That might have been a bad call.

Steve

[racingtadpole]

That error is a warning that you have too much voltage on the bus caused by too fast a deceleration rate. A braking resistor would fix it, as does slowing the deceleration ramp as you have done.

[OxxAndBert]

That error is a warning that you have too much voltage on the bus caused by too fast a deceleration rate. A braking resistor would fix it, as does slowing the deceleration ramp as you have done.

Yes, thankfully the Powtran manual is pretty good with the error descriptions.

Supposedly doing an auto-tune is recommended, but there's nowhere that I can find that has details of what parameters it changes.

I expected it to be internal profile black magic stuff to work out the optimum way to drive the motor, but didn't expect it to be messing with accel/decel times. I thought those would be related to the application rather than the motor itself.
I guess the takeaway is don't be surprised if you get some errors and have to re-tune parameters after the auto-tune has messed with them!!

Steve