Danfoss VLT2800 output unplugged

Hi,

Setup:
S7-400 connected to Danfoss VLT2800 through profibus. (PPO2)
Drive is controlled through the PLC.

Issue:
Simple enough setup.
But I've noticed something. If you unplug the output connector, thus removing the power to the motor, and then start the motor, nothing happens.
Nothing telling me the motor is not connected anyways. The Danfoss ramps up and gets to its target speed.
Also telling the PLC it's driving the motor and everything is a-ok.

Question:
Is there any way to make the Danfoss tell the PLC it's not actually driving the motor, that it's output is not connected/no load on output/etc.?

Removing one phase will trip the drive, but removing them all and it's like the drive doesn't even notice.

Tried tinkering with low amps warning etc, but so far haven't been able to get this particular situation known in the PLC.
Due to the mechanical setup, it's not possible to add an external rotation sensor.


Thnx,
Jeebs

I would have tried the minimum amps idea as well. Are you saying you can't get an amp reading using the Profibus?

I have seen the same issue. Don't think you can use the low amp idea because the inverter showing that the motor drawing amp. even that is not connected
Banker

Banker said:

inverter showing that the motor drawing amp. even that is not connected

Click to expand...

In that case I would be on the phone to Danfoss griping up a storm. That is just silly.

Every drive Ive seen using v/hz mode does this
Generally you get a warning with something like "no load detected" etc
but they run fine
Open loop/closed loop vector- it will fault

Just monitor the current and create your own fault

How can you detect a fault by monitor the current when the inverter indicate a current even the motor is disconected.

It seems a few features one might like/use as a controls engineer aren't present on the 2800 series. The higher/newer series do offer some/most of the features.

Output disconnected is not detected.
Output current is not measured, it's calculated.
The current warning bit present is only over-current, isn't used for low current.

Guess I'll be looking for another solution then. There's always another solution

Sounds like they are running out of corners to cut.

I simply can't believe they are not measuring the current at all. I have never heard of a VFD that did not. Sounds like the best solution is to use another drive.

We used to have Hitachi drives.
Not bad, but they tend to stop working while telling you all is well. (sounds familiar?)
Didn't have much in the way of feedback, so it needed changing. Suggested using what we already have inhouse with not that bad feedback and profibus controlled, seemed the 2800's would fit the bill nicely. Not too expensive, not too hard to implement. Main focus was profibus control, speed setpoint and monitoring.
Telling management mow that the 38 drives we bought are not the right ones.
I'd rather not. (not after 30+ of em have been isntalled)

They do measure the current, but not on the output. They calculate an approximation for the output, which is a bit off it seems.


Got a few idea's I wanna test before giving up on this though.
But for now, back to switching out some really old Movidyns for spanking new Movidrives.
Will probably revive this thread once I found a solution or completely gave up.

A few "out of the box" ideas.

Possibly altering the starting voltage to see if that creates a big enough difference in the calculated output current.

I assume you have already tried param 223 (low current warning)

Has it been tuned to the motor? For example P108 and P109 for the winding resistance and reactance. If these are not correct then the calculated output will also not be correct.

See how changing the current minimum value P140 affects the calculation.

Maybe one or a combination of these things is the ticket?

D

The VLT2800 is calculating the motor current from the motor data (best to do the AMA Par. 1-07), but it is calculated using a meassurement of the DC link and this is why it can still show a current drawn.

I have no idea of the level, but you should be able to map the DC Link Voltage to your PCD Configuration in Par. 9-16.x.
Perhaps you will be able to se a difference from motor connected to motor not connected (and running).

Just out of curiosity, how much current does it show?

Michal_dk

/target Old Thread
/cast Revive

Figured this one out.

There's no way to get a current reading from the device.
Oh sure, there's a reading on the screen, can get it through the parameter communication and even using an analog signal.
Problem is, the reading is a calculated value. NOT an actual measurement.
Which means it doesn't even flinch when you disconnect the motor.

This means all them warnings and settings and outputs you can supposedly use to tell you something is wrong with the current, all useless.

So how can one get a signal from the drive telling you the motor has been disconnected?
Power reading.
There are no warnings or digital outputs that can be linked to the actual power usage, but there is a reading and an analog signal.
This reading/analog signal is accurate and can in fact be used to get a signal to a PLC to tell if the motor is running or not.

Good thing we have spare Analog inputs, thus the only cost is the cable (and my time).

Hi Jeebs!

To say it is NOT a calculated value is a truth with modifications! The DC link current IS meassured! But the motor current is then calculated using the motor data the user has to put into the system. The VLT2800 will only meassure the value of the stator resistance, not the reactance and as this is usually not readily available from the motor manufacturer - well then the calculation will be wrong.

I talked to a colleague who has more experience using the VLT2800 than me, and he told me that he has seen systems with this exact behaviour. Using a VLT5000 (then, FC302 today) they were able to get the right motor data, transfer them to the VLT2800 and now the current was calculated correct. BUT at low and high frequencies other things will be inaccurate and you will again experience the problem.

That aside, if I remember correctly P = U x I x Cos fi. So how is it that the power is more accurate than the current?

I haven't got access to a profibus model, but if you set Par. 9-16.1 to 12 wouldn't you see the power (assuming you have default settings for display - par. 0-12)?

Best regards,

Michal_dk

Question 1:

Why is power reading more accurate than current reading?

When you run the drive, the current reading when connected = disconnected. (It should drop to 0 or damn close when it's disconnected)
This is not only the reading, but also the analog output you can configure to provide the current, relevant to Imax.

When you run the drive, the power reading when connected != zero, but it drops to zero if you disconnect the motor.

Given a certain permissible margin of error, I'd say power reading is more accurate...


Question 2:

Yes it's possible to read out the power using the parameters you've mentioned, but that would involve changing the PPO and standard blocks used, which then again is not 'accepted'.



We have close to 200 of those drives at the plant, used in different machines by different machine builders. I personally checked the behavior on ~70 of them, 2 different drives for each individual application. The rest are copies of the ones I tested.
All of them behave exactly the same.

They all drive 'visible' motors however, so if it's not running, it's noticed by man or machine. Murphy just happened to have it happen with the only motor which is not visible, won't be noticed by man or machine and is vital.


Was fun trying to find a solution for this one though, learned a lot.

I didnt know drives calculate theoretical current. Do the new modern drives still do this. I use to trust the drive current reading more than my fluke clamp.Not any more.