VFD's and 3phase Induction motors FAQ corner / pls. Read all first

Hello FarmBoy,
Wow, what an overview - thank you for all of us here.

FIRST, let me say this. You all have heard me say the "heavy weights" will disseminate the info come up with an assessment of "it won't work, it may work, let's explore further".
I use this "phrase" to let those (and they know whom I'm referencing) folks know that I may understand the concept, But I cannot reply with "anything" factual as I do NOT have that expertise, PERIOD.

YES - or these VFD's that are "sensorless" - http://www.automationdirect.com/adc...S3_Drive_Units_(230_-z-_460_VAC_SLV)/GS3-21P0

http://driveswarehouse.com/p-2525-wj200-007sf.aspx

I know these have been used "before" in other DIYer's builds. Most recommend the "sensorless" VFD's.

I KNOW what they DO, WHY we must use them for AC Motor Control, that the Pots to boards to VFD's give the instructions - 230V Single Phase in - 230V Three Phase out - to my 3 phase motors.

NOW, that's ALL I know. I'm at the mercy of the "heavy-weights" here to help me wire it correctly.

Thanks my friend to being another integral part of this fine team........as "Larry - the Cable Guy" here in the US says: Let's get er done!

Have a blessed one folks!

Tom

Hi Tom,

My understanding of VSD are mostly within the pump industry. I have some experience in other fields, but not any when it comes to quickly changing direction of an induction motor. What i am trying to say is that there are plenty of things that i do not know or may not understand correctly (like vector-flux - i read somewhere that it is important but i do not understand it; perhaps the correct term i was looking for was sensorless vsd or somewhere else i read about space vector, there seems to be so many algorithms available), i actually hope to be proven wrong so that i may learn something.

This second drive you have shown me looks pretty good. Especially if the analog input is 16bit, i think your previous vsd might have an 8 bit analog input.

Some modbus commands are meant to be written to EEPROM while the VSD is idling, other allocated registers like PID feedback may be written to while the VSD is active. I couldn't find this split in documentation of this second drive, but i am sure it is there somewhere.

Cheers

Hi FarmBoy,
I'm almost sure I'm going with the G series, Sensorless VFD....glad you confirmed my choice - and due to my minimal understanding, GLAD you pointed out 16 bit vs. 8 bit - which I do highly understand.
I was going "more" for the Sensorless Version, as I've read it's so much better.

Again, your understanding is far more than I have acquired at this point.

HOPEFULLY the "other" VFD members with expertise will capitalize on this part of our VFD conversations and move further along?

Thanks,

Tom

Hi @FarmBoy ... My friend, Thanks for the really valuable info's ... Now I have someone to talk to technically ... It seems you have some good experience in Modbus programming ... Don't worry about the manual I've Tons of them and it is easy to set any parameter on it for every one ...
Here is what you've just added to me and I'm clearly understand ...

There is a stander Modbus protocol already used (SW developers good news ) that could be common for all kind of VFDs which already not SimTools compatible ... So when comes a day, SimTools sending data to Serial/Network ports there will be a later update for these devices ... perfect .

Forgive my following questions but I had a long time earlier searching for that you may have there answers ... I saved them " forced " for someone can really support and enough interested ... thanks for your patience and understanding .

Where is that motor feed back connection exactly in any VFD/VSD drive types ? what type of feed back sensor in sensor-less drive can be used ?...
Even IF, you managed to get Feedback to the VFD/VSD drive by any means ... and comparing the feedback data with the new data dump received from PC ...
Who will calculate the new set point ? ... " Of course NOT the SimTools on the PC !!! " ... With how much speed of data processing / Enough PID parameters control ? ...

Look at the big picture " extreme position control " a " Robot Arm " not just " Speed Control "...

FYI my friend stopping and reversing the motor with VFD is quit very safe and easy for 3ph motors at any industrial application that is done in two fast smart steps "stopping the motor - starting the motor reversed " ... for Motor breaks it is the application requirements is what makes you need a motor break ... And motor breaks are only for direct shaft driving mechanisms without gearbox ... with heavy torque loads where motors can't stop the load exactly in time due to it's big inertia or at the stop position the load can rotate the motor shaft loosing its position like Elevators for example .

P.S. the front/external 270 degree potentiometer is not a position feed back even if it is programmed stop in it's center and F/R when pot rotated in both directions ... simply the motor will not stop when it reaches the new position ... you have to exactly center the pot again to stop the motor .

I'll keep my "Hope in the better future " open just for you .
Best regards My friend

One more thing my good friend ... there are many other types of motors driven VFDs ... Stepper motors ... Servo motors ... BLDC motors .
Dig in my friend I want to give you a good start ...

next link is my 1HP VFD ... in case you want me to test something for you .
http://www.delta.com.tw/product/em/drive/ac_motor/ac_motor_product.asp?pid=1&cid=1&itid=10

and see on the lift the " motion & control " section ... many amazing devices plus this PCI controller ... For expensive high end racing simulators and CNC machines which I guess @value1 will be interested with too .

http://www.delta.com.tw/product/em/motion/dmc/dmc_main.asp

For analyzing the following simulator ... You need a good game engine and manager program, motion controller like above for stepper/servo motors , ball screw with end bearings support, fixable joint , stepper/servo motors, linear sliding bearings, ..., ..., ... , plus some good metal frame work ... you'll get an Empty wallet but a lot of fun .



@cthiggin ... I want you to download the full manual for your VFD and read it line by line ... start asking questions about its setup ... Do your home work please

Hello,

It is exciting! Perhaps someone have been working with a VSD + Arduino combo for a while now, if CRC16 or other protocols functionality is added to simtools then your only expense will be RJ45 connectors and some wiring. This combo is like having a V8 available for every wheel but you decide to install a 4 cylinder to power the whole car.

Sensorless VSD:
At first i thought they only meant the type of algorithm they use to calculate the position of the motor, but going back and through the documentation i noticed that they only provide definition of your process variable PV under PID control. If this is the case and they are not using any sensors to detect the position of the shaft, then that is firstly amazing!
If this is the case then I think the magic happens between C052 + C053 and (Perhaps) P072 + P073, i say perhaps because it says pulses and because i cant find another place to send the PID reference on that drive. This is the type of drive you setup and play with over a day or two before you will understand everything about it.

I looked at the delta drive. This is for sure a type of VSD parameter setting that i am more familiar with, so i will likely also go for something like this. What i don't like about it is that the analog input can only be scaled between 0-100%, this is not good enough for me.

With both drives, i dont know how you will reverse the motor very easily. It is possible and essential to me not to send a reverse signal to drive, it should just know that to get to that position i have to reverse now.
Say my pot limits between 3volts and 8 volts because of my mechanical structure. I want to set the scale between 30% and 80%, but then i want to set the range between -100% to 100% betweeen those voltages. This way the modbus string i will send can be short and i can therefore send more. I also dont have to worry about FWD or REV and leave that to my drive.

In simtools we have multiple interfaces available right? So say you put every drive on an separate interface and they can all squirt data to their drives in parallel simultaneously and say your drive can handle a baud rate of 19200. Provided that your VSD is not expected to reply to every command send and your VSD supports a range between -100% and 100% without expecting any fwd or rev signals. Now a modbus signal like this (01 05 0C 00 00 02 4b 37) is 64 bits. 19200/64=300 modbus signals per second max. That is 1000ms/300 = 3ms. One modbus signal every 3ms. Now if the crystal used on your VSD is also quick enough then it will be able to process at this speed. Your PC is not doing the PID calculation, so its processor can focus on the game and extracting data and each drive can take care of its own PID. This just sounds rights, doesn't it?

Thanks for the vid! That looks very awesome! I will enjoy building something like that someday!

Cheers

I forgot about this:

You will connect your feedback sensor to the terminals of the drive. One will typically find a 5V, 10V or 24V DC supply and a GND and a Analog Voltage Input on the terminals. Connect feedback there and under PID parameter settings one will select "terminal" for PID feedback. Make sure to tell the drive it is volts and not amps, very often only one analog input terminal is supplied.
Each one of your feedback sensors will have a dedicated DC power supply that is not shared with 1-5 other sensors.

Does that answer your question?

Hello FarmBoy,
I have read your latest information at least 10 times and while I basically understand your concept - IT's like me learning a new, complicated, language - which I am still learning English after 66 years.

Let me see if I am "even close"....
Digital Feedback Pots (5v) would attach "directly" to the VFD (VSD - assume VSD means Sensorless Drive?), bypassing the need of going though pcb/opto-isolators/pwm's???????

You're saying "this method" is like having an 8 cylinder engine ON EACH TIRE - which I gather would be super FAST - by doing this, we are linking directly to VFD/VSD??

I understand?? that one would need a dedicated DC 5v Transformer to supply the 5v Digital Pots??

What could be eliminated in the electronics set-up - board/opto's/pwm's ???????

Your method would allow for faster pc processing speed of extracting game data??


The attachment is from my friend, @yobuddy, who was kind enough to send me an initial tentative component/wiring diagram when we first started talking about my needs in building my AC Platform.

@speedy: I've downloaded the Operators Manual several months ago on my VFD/VSD that I will purchase shortly - I'm as far in the manual to the Wiring Diagram for input/output of the
2 phase to 3 phase - and I do "see" where you would hook up from your electronics board - I AM trying to "get my head around" and correctly understand some of the explanations and
nomenclature - WELL, please understand, it's like teaching a person to weld or do accounting that has NO clue or idea of what I'm talking about - it's a hard sell, and sometime an exercise
in futility - One thing is FOR SURE - I will NOT let someone do my job without my intervention...it's the only way I know how to work....and FINALLY, I am mature enough to know when
I'm in "over-my-head" that could put me into a dangerous situation.

THANKS gentlemen and ALL.
We have gone from a "hands-off - too slow - too expensive" AC dormant subject - to - an explosive hands-on positive approach on the AC operation for a motion sim. THIS is a GOOD thing
for the forum, and all folks present and future for information.

Have a blessed one,

Tom

Hi Tom,

Thank you for appreciating my comments. I am glad you find it helpful.

From your attached image and the way it seems to be done now you have the following manner of connecting things:
PC-Controller Board-VSD-Motor
PC: Extracts data from game and sends the new setpoint to Controller Board
Controller Board - First needs logic to determine what new setpoint belongs to what motor, then it does an error check with the setpoint by comparing it to the sensor and it then calculates an analog output voltage that is connected to the VSD. It also needs to tell each VSD when it should go FWD or REV. This controller board has allot of calculation to do.
VSD: Simply goes where it should according to its analog input value received from the control board.

What i am suggesting is this:
PC-VSD-Motor
PC: The PC will still do the same job it did before. After reading what i said about the PC not calculating PID i can see why it is confusing. I meant it as a statement to show the flow of things. Unless your PC was used to calculate your PID from before then only will you save processing time on the PC by letting the drive do take over the PID control.

Please excuse if this is too basic, but i will try to start here.
So a VSD takes AC input, it goes through something called a rectifier and filters to change it into DC and then transistors are used to change it back into the correct AC for the motor to turn in a certain way. All the algorithms like vector-flux control and sensorless control and spaced vector control have something in common and it is the way/manner in which the AC signal is send to the motor.
Most of them have voltage available on the terminals to power up your sensors. It is more sophisticated than you might think at the moment.

When i work with pumps i only use two analog inputs on the terminals of the drive: 1 for feedback and the other for a setpoint. So i have a pressure transducer that reads the pressure in the pipeline and a pot to adjust my setpoint. If i want a pressure of 3 bar at all times at my pump then i will change the setpoint (pot) value to give me 3bar. If the flow in the pipeline increase because somebody opened a valve, then the pressure will decrease and the speed of the motor will go up. This is a basic closed loop system.

The same principles will apply with a simulator. We will need a transducer type of sensor, but this time it is a sensor to measure the position of the gearbox cantilever. This sensor is powered by and connected to the drive. The second analog input does not have to be connected to the drive if one is able to change the setpoint from the PC using comms.

So many of these VSD's are being made. I dont think the price should be too expensive. That sensorless VSD you showed me is great, but perhaps too great and that is why they charge so much. Also why is everybody so stressed about the sensorless VSD? I understand that you want your ac motor to react like a dc motor, but there are other factors like those i mentioned above that are also important. If you can still decide what VSD to use, then firstly decide if you want to use an external controller (with simtools support that is available now) or a comms protocols like modbus (that will hopefully be available someday) and then secondly buy your VSD based on the option you have chosen. For comms option i suggest starting with a drive that gives you good analog input calibration.

Cheers.

Farmboy, many many people who do not regularly post on this topic appreciate your, and everyone's, effort to help people understand this technology.

Thank you

FarmBoy,
WOWWWWWWWWWWWWWWWWWWWWWWW & Geezeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee,

I "again" basically understood your overview - and it Sounds GREAT!

@FarmBoy says: Please excuse if this is too basic, but i will try to start here - PLEASE STAY BASIC - even "I" can mostly understand the methodology!

I assimilate this as "meaning" - We DON'T NEED the PCB's / Opto-Isolators/PWM's/ - JUST straight from PC - to VSD - to Motors? SOUNDS TOO SIMPLE!!!

Questions????
1. I/We would still use the Contactless 5v Pots to feed info to VSD for crankarm position - but they would connect Directly to VSD? (IS this the same as the "transducer type of sensor")
2. The 5v Contactless Pots would get their power from the VSD??
3. SimTools Software CAN be used same way - just as IF I were going to use the PCB, Opto's, PWM ?????
4. FYI - I know that @eaorobbie mentioned awhile back that specific Profiles for the simulator action (Ex. 1 for Flight Sim would be pretty docile - Ex2 - Racing would be a more radical Profile)
Could be written into firmware and sent to the SimAxe board - then the operator could flip a dip switch for the profile he wanted - IS this doable with your proposed method??
These settings (from my understanding) are "outside" of what the SimTools software would do.

I'm very very "thirsty" for the "Latest & Greatest" way of getting the best that I (We, the team has kept me out of trouble) - can......just want it work be the very best it can be!!!!!!

YOU'VE SOLD ME!!!!! - but I'm EASY - due to my lack of knowledge and expertise in the electronics arena.
@yobuddy / @speedy / BlazinH / the "Heavies" that KNOW everything your talking about - need to get into this as a Team and RESEARCH it further / bring up any negatives / weight it out / modify it /
whatever - so it can be published as a done deal!

Thanks FarmBoy for becoming a team member here at the forum. You are bringing a most helpful expertise level with you.

Have a blessed day friend,

Tom

Hello ...
you miss the basics of motion control again and again ... and again ... I'm sure there is tons of it written here in this Motion simulator website some where but I can't find it ...

OK ... Proceeding in a very basic words in English as I can to make it clear for each point for all of us ... so we can build a solid base line by line .

Enjoy the " data flow "all of you ... keep watching, posting, asking ... please understand every thing .

step1 ... Please don't mix between Speed control and Position control ... Honest mistake .

Speed control is for a something that is running and running "lets say " forward [ pump -fan - escalator - train - ... ] and needs automatically adjustments in its RPM up or down a little bit to keep things going normally ...
which by a good resolution transducer can give an analog voltage deference that represents the new required variation in the output side " no more than this for the transducer "...

This voltage deference can't be used like what you think it is ... Many Factories will be closed if this is real not just the Arduino .​

Then comes the VFD PID control system roll that is taking that new voltage converting it to the new frequency sent to motor's windings to run on that new speed ...
Very good PID ... solved the error decreased or raised the motor speed to the new set point effectively without speed over shooting "pumps " and in time ...
But ... this PID system didn't tell the motor to stop exactly here ... or a little further ... because you're talking about the motor speed not the motor shaft angle " like the stepper motor or servo motors " ...

You'll say its the traducer responsibility ... and you didn't use the proper transducer what ever its kind " potentiometer - Hall effect - ... ".​

I will say No as for ... Position Control is for a distance from point A to point B ... let's say this distance is divided into 512 set points ... you have to provide an actual 512 stop points that is electrically presented by your cheap 10k pot to your connected Motion Controller so when it powers ON/start ... it already knows where you are now ? where is the new set point ? are you off that point or exactly at it ? ... so it provide you with the proper motion command and you're in stop start ...

Then comes a new set point it "the controller" calculate the error [ PID algorithms here takes place ] gives a corresponding [ F/R + fast/slow ] command to the motor to move ... depending on the feed back readings from the pot the error will decrease ... again and again then comes to zero ... then a motor exactly stops where the new point is ...
Very good PID ... solved the error in position effectively without position over shooting "pumps " and in time too ...

that makes this Motion Controller do double the above system ... it increases the motor speed in certain direction then decrease it again to the required calculated stop point ... then keep doing that over and over ... maintaining a good follow up of motion .


based on the above ... if every one confirms OK ... first base line in argument is set in position .
I'll move to step 2 ... the VFD is a big H-Bridge ... concept of it's operation .

thanks for interesting

SPEEDY

No ... that is for connecting the 0-10v analog signal

Hi Speedy,
I'm going to read and read again your "rebuttal" to @FarmBoys' overview -
Boy, you "heavies" are really doing a good job of instruction for us that are not "heavies" yet.

I'm going to watch the pro's at work here. I have to quit reacting like a "little boy in love" when I read something that is so exciting, and making "un-justified" replies....as I am not in a position to do that
yet - I will only ask questions - make NO statements until the VFD/VSD "Program" is put to bed - IMHO, taking this position, I will help keep the thread cleaner without injecting meaningless chatter.

IF I understand what you're saying Speedy, in your overview and rebuttal to @FarmBoy - IS -
FarmBoys' methodology is great for another type of industry - BUT - for motion simulation, his method will not Work Well???

Thank you gentlemen for engaging and exchanging very very valid points...

Have a blessed one to ALL of the gentlemen here.

Tom

Hi,
Thank you for speedy for separating speed control and positional control.

Let me give you a case study of a VSD that i think has the potential of being a good VSD to be used for motion control. If there are plenty of questions then it will help to reference it to something.

So this is a local drive available here in Cape Town that is I think has the ability to do motion control:
http://www.gator.co.za/senlan_sb200.html
I found the datasheet for this VSD under downloads.
Also, this is a more expensive drive. But we need a proper starting point here.

So lets start with the analog inputs, found under parameter F6-00 to F6-03
F6-00 and F6-01 can be set used to set the boundaries of your feedback.
Lets freeze here for a moment.
So one might find yourself having a cheap pot that can do more than a full 180 degree turn (which we wont need with a "normal" motion simulator), so we open the pot; do some grinding until we've created the range of motion and use some silver or other compound to fill the grinding area. This is the mechanical way of calibrating a POT. Now i believe that there is a software approach to this too:
F6-02 and F6-03 can be used to set the new range of motion.

Example:
Sensor: Cheap 10k Pot, say you provide 10VDC
Problem: If you take a multimeter and measure the volts when your crankarm is pointing down one might get 2V if you measure the volts when the crankarm is pointing up then one might get 8V. So not only do we loose 4V worth of range of motion but have greater problem; we need to let the drive know that when we provide a new setpoint that is closer to say 8V then it needs to rotate the motor FWD and if we provide a voltage of say 2V then it needs to rotate in REV. This is the right way to differentiate speed control from positional control in my opinion, not activating FWD or REV every time you want a change in motion.
Solution:
F6-00 = 20% (2volts is 20% of the full 10volts)
F6-01 = 80%
F6-02 = -100% (linked with F6-00)
F6-03 = 100% (linked with F6-01)
Now the sensor is basically calibrated.
I do not believe (and i might be wrong, but please prove it) that if one can setup an analog input like this then you wont have to tell the drive when it needs to REV or go FWD you just send it a new setpoint or point of reference to reach and it will react according to your PID setting.

This VSD provides 10V supply at its terminals. If your sensor requires a 5V supply then create a voltage divider with two resistors.

Absolutely. You really want to pay that amount of money for a VSD and only use a small part of it. VSD's provide allot of resources that one might as well make use of.

I believe so. There should be little change to they way things look, if there will be any. Perhaps a way to use the interfaces in parallel will be great (i dont know if the software can do that already). The rest should be pretty straight forward.

You will be able to that with this drive. Under the PID settings of F7-05 to F7-10, they provide two sets of PID. Now you can call the second PID in a similar fashion to the way your are reversing forwarding your drives from an external controller now. Say you are connecting your button to discreet input X1 then F4-00 = 36
Now you can set integral time F7-06 lower for a quick reaction and F7-09 higher for a slower reaction (Obviously one will have to play around with the PID here).

Cheers

Hello FarmBoy,
Thanks for answering my questions...

I'm so glad you and @speedy are debating the VSD pro/cons for us who are using / or / will be using AC and VSD's....

I'm anxious for @speedy's rebuttal.... You guys really know VSD's well... (and that's where I will fit into the equation - the learning and as an end-user)

Have a blessed one folks,

Tom

As stated previously I don't know much at all about VFD/VSDs etc. But in doing some research I found this information that I thought some may find useful.

This information is from Gozuk, a Chinese VFD manufacturer.

And a little more information I got from wikipidia for VFD's.

However, the above comparison of may be unfair when comparing servo motors to asynchronous motors as wikipidia also states:

ps If you don't agree with the information then don't use it! I'm just passing it along.

This is really Good ...lots of appreciations .
well done my friend

I've been writing a detailed reply to @FarmBoy then came @BlazinH post ... I couldn't resist the impression of the will stated description of the concept ... I deleted my 3/4houre writings and posted hem his prize .

Guys ... Is every thing must be written here ? ... So, excuse my lake of English as it is my second language ... or memory short for my internet readings history "Can't link something new that I didn't read myself first " so I have to read it at a glance and again to be sure ... that doesn't affect any thing of what I may know and did by my self .

That's why you may notice I'm editing my posts and it still have lots of defects or doesn't express the meaning exactly .

Don't worry guys I'll do my best .

Thanks Gentlemen,

Continuing to Learn "more & more" with each of your thread replies.

Have a blessed one,

Tom