Pulse train from a VFD 2

[cokeguy]

We need to obtain a pulse per Hz signal from an A-B Powerflex 70 VFD in order to use it as a master synch pulse signal for a certain production line. The rest of the line is servomotor driven and is follower to this master pulse, no problem there. Right now we are getting this master synch pulse signal "mechanically" from an external synchronous motor/gear/encoder combination connected to the master drive along with the other motors (this single master VFD runs three small synch motors simultaneously). This arangement works OK but it is pretty cumbersome because of the additional motor required just for synch purposes, since we cannot couple encoders to the other synchronous motors.

If we could get this pulse/Hz signal directly from the drive, directly or through an option card or an external card/converter/whatever, we´d simply feed the pulses into a counter and generate the master pulse every N pulses.

Thanks a lot for your help.

 

[Skogsgurra]

The problem is the PWM waveshape. I have used an LP filter to sort out the fundamental frequency.

Like this:

Phase A --- resistor ------------- optocoupler anode
| |
capacitor diode cathode
| diode anode
| |
Phase B --- resistor -------------- optocoupler cathode

Make the cut-off frequency low. I used 1 Hz. The voltage across the capacitor will then stay constant as speed (frequency) increases. For a 400 V drive, I used two 5 kohm resistors and a 10 + 4.7 uF capacitor. That makes your cut-off frequency a little higher than 1 Hz, but it works.

Remember that the full PWM voltage is driving current through the resistors. I used 25 W resistors. Yes, 25 watts - not 2.5 watts.

The capacitors never see more than the V/Hz voltage. So you can use almost any non-polarized capacitor. I used old Siemens MKL 63 V capacitors that happened to be available. But you can use much lower voltage.

And, of course, there should also be a current limiting resistor in series with the optocoupler. Make it something like 330 - 1000 ohms. Depending on what coupler you use.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[DickDV]

skogs, I think you might have misunderstood the original poster. I seems to me that he is looking for a pulse train that represents speed for use by the follower drives. Similar to an analog output line speed signal but a pulse train instead.

It looks to me like your network is stripping out the PWM pulses and passing the fundamental sine wave on to the optocoupler.

Am I seeing this right?

 

[Skogsgurra]

Yes, it is a low pass filter. It sends the fundamental (usually between a few Hz and 50-100 Hz) to the optocoupler.

The optocoupler outputs a reasonably good square with same frequency as fundamental. And I think that is what the OP wanted. At least, that is what he gets from the electromechanical contraption he is using now.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[DickDV]

Ah, yes, I missed the part about the optocoupler squaring up the fundamental wave.

Clever!

 

[cokeguy]

Thanks guys, although I would have thought that an off-the-shelf solution (option card, converter, programmable drive digital output, etc..) was available for this type of applications, but now I suppose it is probably not the case, or is it?
The truth is that, for reasons more political than technical, I´d prefer to install something already made and proven rather than having it build, any other suggestions?

 

[fangas]

Well, I'm not too familiar with the A-B 700. But, if it was me, and I was looking for an off the shelf solution. I'd see if that VFD had an analog output that can track the output frequency. Then, I'd get a dc voltage to frequency signal conditioner. (

http://www.api-usa.com/,

comes to mind. specifically,

http://www.apcs.com/index.htm).

Hope that helps

Ed

 

[CJCPE]

cokeguy:

The pulse train output that you want used to be a fairly common option, but I don't think it was used much and manufacturers don't often have it now. I would check first with your local drive supplier if you have not already done that. There may be an accessory supplier that makes something like Skogsgurra described. I am thinking of one or two companies that might do that and will try to remember the names.

 

[cokeguy]

Thanks fangas, unfortunately the analog output converted to pulses won´t give me the required resolution, since I do need the exact pulse per hz signal. And thank you CJCPE, if you do remember something please let me know. Just like you mentioned, a few years ago we used Yaskawa drives with an option card that did just that.

 

[CJCPE]

Bonitron makes a number of VFD accessories, but I didn't see what you want at

http://www.bonitron.com.

The input and output reactor people,

http://www.mirusinternational.com/,

http://www.transcoil.com/

and

http://www.mtecorp.com/

have not branched out that far from their core products. There must be some other sources to check.

 

[ozmosis]

I think there is only Yaskawa that can receive and generate a pulse train used as a setpoint reference. A few other companies can receive a pulse train (Danfoss FC300 range can receive up to a 5kHz open-collector or 100kHz push-pull pulse train and use this as their setpoint) but I know of only Yaskawa who can both receive (as a setpoint/follower) and generate a pulse train.
If your Powerflex70 is to be configured simply to receive the pulse train from the encoder you already have, it might be worth looking at their encoder option to see if this can be configured as a SP follower rather than simply a closed-loop encoder feedback for speed set-point.

 

[cokeguy]

Thanks sed2developer, but the Powerflex will actually be the master, so it needs to generate the pulse train, not receive it.

 

[controlsdude]

I thought powerflex had this option for pulse train master and slave drive where they use an ethernet port on the drive on. The master sends the sync signal and the rest will follow the commanded of speed. The only problem with this I think is its ethernet and then you run into the IT people demanding that that all ethernet is home run to a switch and all that crap political bull that goes along with it. So I guess dont they them and just creat your own network internal to the drives.

I do know that yaskawa has the pulse train where one drive sends out the pulse train of say speed and the rest follow. This would be a digital signal not an analog signal on most drives. Using the analog signal to send out this signal would give you slightly different results in speed on each drive since it is an analog signal. This would be the reason for the digital pulse train above which would be better.

In the MHE world this would be the difference in a smooth ramp up with no jams where multiple drives have to come up together at same speed.

 

[cokeguy]

Thanks controlsdude, but it has to be a digital output and not go through the ethernet. In our case, we do set the reference speed via ethernet on the master drive, but the rest of the equipment (servo-driven gear form other manufacturers) has to follow pulse from a digital outut to keep in sync... thanks... I guess it would have been very simple to a drive manufacturer to include this function in its programmable outputs, but well, sometimes we can´t have it to easy...