frequency inverter

More
29 Sep 2010 20:16 #4421 by occesar_0
frequency inverter was created by occesar_0
Well, this is not exactly an emc2 question, but hopefully somebody can help me:

I want to control the siemens- frequency inverter-micromaster 420, ( I removed of of a machine that no longer operates).

So, I was expecting to control it by a pwm signal, and then control the drill velocity through emc2.

Then I started to read the manual, but I didn't find how to do it.

The question is: Does somebody knows about the digital operate of this kind of drive? I only found how to control it by a potentiometer, and that is not useful for me.

Thanks

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 10:17 #4430 by andypugh
Replied by andypugh on topic Re:frequency inverter
occesar_0 wrote:

The question is: Does somebody knows about the digital operate of this kind of drive? I only found how to control it by a potentiometer, and that is not useful for me.


Generally you need to convert the PWM to an analogue voltage (which can be connected in place of the potentiometer, ignoring the +5 or +10V terminal.

This can be as simple as the attached circuit with a resistor and capacitor (which could be directly on the inverter inputs)

Resistor should be about 1k so that the P-Port current is 5mA, and then choose a capacitor so that the time constant is about right, 10Hz would be 100uF.

That circuit will only supply a voltage as high as the p-port voltage, but you can allow for that in the drive.

It would be better to opto-isolate the signal and use the VFD +5 (or +10) V and I have a circuit to do that (using two optos in a push-pull arrangement) , but I have not actually tested it yet.
Attachments:

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 10:42 #4432 by dab77
Replied by dab77 on topic Re:frequency inverter
that's an interesting thread. is it a three-phase inverter the one you want to test?
this is a point that i wanted to reach, too.
andy, you're sayng that in this simple way you can control an inverter directly with the step pin on the par-port, or i've misunderstood?

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 10:46 #4433 by dab77
Replied by dab77 on topic Re:frequency inverter
ops..pwm is not step/dir.. sorry.
ok so the question is: which is the best way to control three-phase inverters from emc2? :)

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 11:41 #4435 by andypugh
Replied by andypugh on topic Re:frequency inverter
dab77 wrote:

so the question is: which is the best way to control three-phase inverters from emc2? :)


What sort of three-phase inverter?

If you just mean a Variable Frequency Drive (as in the original question) then the setup above will work. it will run a 3-phase induction motor at a speed controlled by EMC2. Most EMC users are doing exactly that, I think, even if the PWM-voltage conversion is done on their break-out boards.

If you want to control a 3-phase inverter bridge (ie a set of 6 MOSFETs and a DC supply) _directly_ then you can do that with the Mesa tppwm function available on their FPGA cards (such as the 7i43 or 5i23). That is how the Mesa servo drives work.

Alternatively, I have written Arduino code that can generate the 6 drive signals required to create 3-phase from DC. (which is what a VFD does)

See the 3-phase section here: en.wikipedia.org/wiki/Inverter_(electrical)

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 14:05 #4439 by dab77
Replied by dab77 on topic Re:frequency inverter
the point where i want to go (through some clever steps..) is control 4 motors with emc2.
-the motors must be powerful and fast, as long as they should move something like 50kg at 2-3mt/s
so i think i cannot use servos, but i have to use induction 3-phase motors. (?)

-to control this motors i need VF inverters , maybe Vectorial Inverters (i really don't know the practical use differences) (?)

-to drive the VFDs do i need a drive card between emc2-pc and VFDs, or there are VFDs connectable direcly with parallel port or PCI?

-after that i'll need also encoders, limit switches, etc... so the need to plug them.
thanks for any suggestion

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 16:14 - 30 Sep 2010 16:18 #4440 by andypugh
Replied by andypugh on topic Re:frequency inverter
dab77 wrote:

the point where i want to go (through some clever steps..) is control 4 motors with emc2.
-the motors must be powerful and fast, as long as they should move something like 50kg at 2-3mt/s
so i think i cannot use servos, but i have to use induction 3-phase motors. (?)


It rather depends on if you want to move very slowly too. 3-phase induction motors are very bad at slow movements, and will not lock stopped.

You _can_ use a 3-phase motor as a servo, as I showed here:


And having experimented with a flux-vector VFD (keep an eye out on eBay for the Altivar11 for example) I am sure that one of those would work a great deal better (I was rotating my milling spindle at <10rpm yesterday with an Altivar11)

However, the ideal motor for what you describe would be a large brushless DC servo motor. To confuse matters these are often called "AC servo motors" . The same name is also given to an old-fashioned sort of motor which is a true induction motor but with a different style of winding that works well, but needs a very special sort of drive. The trick is to always check the data online before bidding on eBay.

All the high-end very fast CNC machines use brushless DC motors.

You might be able to use model helicopter motors, but you would probably need to add hall-sensors.
This thread is very interesting: endless-sphere.com/forums/viewtopic.php?...e405c63a9a80b86d603f

EMC has a number of ways to drive servo motors.

You need to calculate your forces and speeds to get an estimate of required motor torque and power, then look at the spec sheets (power = torque x 2 x pi x revs-per-second if you work in Nm for toque and watts for power. Note that is revs-per-second, not per minute.
Last edit: 30 Sep 2010 16:18 by andypugh.

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 18:56 #4447 by occesar_0
Replied by occesar_0 on topic Re:frequency inverter
Dab77

I agree with andy and I will test the proposed circuit and add an optoisolator, so after do that, I will came back here to show you what I got and get feedback.

Please Log in or Create an account to join the conversation.

More
30 Sep 2010 19:14 - 30 Sep 2010 19:15 #4448 by andypugh
Replied by andypugh on topic Re:frequency inverter
occesar_0 wrote:

Dab77

I agree with andy and I will test the proposed circuit and add an optoisolator, so after do that, I will came back here to show you what I got and get feedback.


It isn't quite as simple as adding an opto-isolator. The circuit I showed works both ways, current flows out of the capacitor when the PWM is low, and in when it is high.
With an opto-isolator you get current flow only when the PWM is high.
You can work round this with a parallel "leak" resistor across the capacitor, but that makes the circuit non-linear (which might not actually matter)
The circuit I intend to try is this:
Attachments:
Last edit: 30 Sep 2010 19:15 by andypugh.

Please Log in or Create an account to join the conversation.

More
01 Oct 2010 14:56 #4471 by dab77
Replied by dab77 on topic Re:frequency inverter
andypugh wrote:

dab77 wrote:

the point where i want to go (through some clever steps..) is control 4 motors with emc2.
-the motors must be powerful and fast, as long as they should move something like 50kg at 2-3mt/s
so i think i cannot use servos, but i have to use induction 3-phase motors. (?)


It rather depends on if you want to move very slowly too. 3-phase induction motors are very bad at slow movements, and will not lock stopped.

I have to add a brake anyway, so when the motor starts to move, the brake will open (think using the enable out, or something similar..)
that's because the load is suspended (as for the volleyball, if the ball or whatelse weights 50 kg, i have to be sure it doesn't fall to ground!!)

And having experimented with a flux-vector VFD (keep an eye out on eBay for the Altivar11 for example) I am sure that one of those would work a great deal better (I was rotating my milling spindle at <10rpm yesterday with an Altivar11)

nice! yes, i will not need too fast moves, so i will demultiplicate the drum rpm. so that i'll not need the motor to rotate less than 40-60rpm

However, the ideal motor for what you describe would be a large brushless DC servo motor. To confuse matters these are often called "AC servo motors" . The same name is also given to an old-fashioned sort of motor which is a true induction motor but with a different style of winding that works well, but needs a very special sort of drive. The trick is to always check the data online before bidding on eBay.

All the high-end very fast CNC machines use brushless DC motors.

You might be able to use model helicopter motors, but you would probably need to add hall-sensors.
This thread is very interesting: endless-sphere.com/forums/viewtopic.php?...e405c63a9a80b86d603f

EMC has a number of ways to drive servo motors.

You need to calculate your forces and speeds to get an estimate of required motor torque and power, then look at the spec sheets (power = torque x 2 x pi x revs-per-second if you work in Nm for toque and watts for power. Note that is revs-per-second, not per minute.

thanks for the info.. unfortunately i'm still confused on the difference between motors..
i use to work with programmable (very limited programming) motors for the shows, and the motors are 3-phase, brushless, think something around 1.2kW.. driven by inverters, but don't know what kind of. till now i've been only a motion programmer, not engeniering... :)

Please Log in or Create an account to join the conversation.

Time to create page: 0.494 seconds
Powered by Kunena Forum