Encoder signals on a rotating trolley crane 1

[sunrays]

I intend to install a cage motor and a vector drive operating in closed loop mode for the main hoist of a rotating trolley crane. But the issue is of getting the encoder signals back to the drive.

The platform on which the main hoist motor along with the gearbox and rope drum are mounted rotates on its axis on the rotating trolley crane. The control panel is mounted on a stationary platform on the crane. The power to the main hoist motor and the electromagnetic brake is fed through sliprings.

While this is good enough for the power in the new setup of cage motor and vector drive, but the problem arises when I need to take out the encoder signals is the same manner. As the brushes on the sliprings could jump due to vibrations on the crane, momentary loss of encoder signals will lead to problems in the performance of the closed loop vector drive system.

Does any body have a solution to this problem? How do I bypass the sliprings?

Thanks.

 

[nbucska]

The best solution would be a rotating transformer consisting of two ferrite E cores. One is stationary, the other one rotates around the axis of symmetry.


Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[Skogsgurra]

No. Not even pot cores will work. It is about digital signals and they do not travel reliably across an inductive path since they will contain DC signals when the movement stops.

Gunnar Englund

http://www.gke.org

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

 

[nbucska]

Gunnar:
The DC level of the digital encoder is not a signal since
it doesn't contain information. Of course selecting the encoding is part of the design. Some kind of encoding
is necessary for the encoders usually are two-phase
systems with some kind of index signal so you have
to transmit at least three signals.

Since the required data rate is low, it shouldn't be a
difficult design.



Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[ScottyUK]

Bring power in on sliprings and mount the drive and encoder on the rotating section? On the evidence presented it seems you've designed in a problem which didn't need to be there and are struggling to find a solution. Is there a reason why the drive can't be mounted on the rotating section?

----------------------------------

Sometimes I only open my mouth to swap feet...

 

[sunrays]

This change is being done on an existing crane. The rotating section does not have the additional space to mount a drive panel.

The present setup is a slip-ring motor with rotor resistance control.

 

[waross]

Is it feasible to consider optical or wireless transmission of the encoder signals?
respectfully

 

[nbucska]

OP:
Unless the rotating section is embedded in a block of concrete I respectfully disbelieve that there is no space
to mount a drive board.

Could you pse post a picture of the crane or a link to it?



Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[sunrays]

waross:

You just read my mind. I had searched the web for a wireless transmission of the encoder signals, but could not get any significant lead on this.

Another point to ponder is that the encoder could be out of the line of sight of the panel as the platform on which the motor is mounted rotates on its axis.

nbucska:

I did not get your point of view on the "block of concrete".

I will try and get a picture of the crane posted in a day or two.

 

[cswilson]

I have seen "optical fiber slip rings" suitable for 2 channels, and I know they have been used successfully in these types of applications, but they have to be mounted exactly at the center of rotation. Is that feasible?

See for example this webpage (not an endorsement):

http://www.polysci.com/FiberopticsandSecurity/forj.html

Curt Wilson
Delta Tau Data Systems

 

[sreid]

Perhaps the motor could have a resolver instead of an encoder for shaft position feedback. Then the excitation and sin cos signals could be sent through slip rings or rotory transformers. Drives may be available to use the resolver signals directly or an R to D converter could be used if encoder format signals are required.

 

[cbarn24050]

A sensorless vector drive would seem to be the solution. If you are worried about slip rings jumping on encoder signals you should be more worried about sliprings jumping on the drive outputs.

 

[sunrays]

Curt Wilson, thanks for the post. This was exactly the kind of solution i was looking for. I was looking at Radio Remote signals.

cbarn24050: Sensorless vector is good enough upto 2-3 Hz., but i need to deliver full torque at 0 Hz. for which the encoder is essential. The idea is to use this feature of the drive to apply the electro magnetic brake at near zero speed. The jumping of the power sliprings do not matter as the drive selected has the "mains dip ride through" feature.

 

[Skogsgurra]

Hello sunrays,

I have seen a pretty convincing demonstration of 100 % torque at 0 RPM with no sensors at all.

Have a look at Invertek and their Optidrive 3rd Generation. They are limited to 160 kW, though. Will that suffice in your lifting application?

Regarding "jumping power slip-rings" and ride-through: The drive may ride through - but what will the slip-rings look like after a few bounces under full load?

nbucska: Yes, there is as much information in the DC zero level as there is in the V+ level. Trying to transfer them across a transformer, be it E-core or pot core or any core shape, is not possible. Your scheme might work if the drive were never stopped or running slowly. But cranes run slowly and they definitely do stop.

Gunnar Englund

http://www.gke.org

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

 

[cbarn24050]

Hi Sunray, you seem to have changed your setup. If your drive is not on the same part as the motor the you have to run the outputs through the sliprings not the mains inputs and if it is then you can wire the encoder direct.

 

[nbucska]

Gunnar:
Exactly -- V+ has as much inf as GND -- zilch

After 30+ years solving similar but ususally more difficult
problems let me show you just one way :

Use absolut encoder with serial output ( or generate the
equivalent signal from the cheaper two phase+index encoder)

The power in provided in one direction in the form
of a square wave of F frequency. This is rectified and filtered for V+

Shorter pulses of the same frequency carry the encoder
information superimposed on the power square wave
in the other direction.(e.g. by shorting the power line)

Let's use one pulse of different length to indicate
the beginning of the word e.g. the MSB. If we need
10 bits and 100 samples /sec. it is 1 kHz pulse frequency

An improvement: use phase modulation for "0" and "1"





Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[Skogsgurra]

I am sure that's the practical and reliable solution sunrays was looking for. ;-)

Gunnar Englund

http://www.gke.org

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

 

[Drivesrock]

Hi all!

I understand that a closed loop vector drive basically controls the slip between the rotor and the stator to achieve the required speed/torque. This can be done by indirect means - calculation with measurements of the drive actual output volts and current or by direct means - normally an encoder.
When using indirect means at very low speeds - something like <5% nominal (your 2 / 3 Hz) and the drives V & I input circuits resolution becomes important and this is why the drive has trouble in really maintaining accurate speed although I believe full torque is not a problem for good drives.

My concern with transmitting digital encoder signals via wireless or other circuits is the introduced delay and possible random failure of the signal in part or in full due to interference/contact. The drive will not be able to correctly control the motor or may trip regularly on encoder failure.

My suggestion is to keep the solution simple by either mounting the drive after the sliprings somewhere - don't forget to install an input line reactor to protect the drive I/P (filter eventual sparking) and wire the encoder directly OR run open-loop vector control and program the drive to operate outside of +/-3Hz by opening / closing the brake at/above this frequency. You may have a 'minimum frequency' parameter in the drive. Perhaps the gearbox reduction means that the motor normally works above the equivalent rpm speed? I would expect the existing maximum rotor external resistance to achieve more than 3Hz equivalent rpm. Oh, and don't forget to fit an output line reactor to help protect the drive O/P as mentioned before.

Select your drive supplier wisely - a product that has a brake control application or functions that can be programmed to do this. I believe I've even seen a drives manufacturer or two say they now control speed well at 0Hz without an encoder. Perhaps get a manufacturer to do a trial? I didn't see your motor power rating but ensure you adequately rate your drive for the starting/acelerating overload - 150% over nominal is usual but you may need more. Don't forget you will need dynamic braking unit and resistor bank (or mains regen).
I have on many occaisions re-used the existing slip-ring motor and had the slip-rings short-circuited but always used an O/P reactor.

I hope this helps.

Drivesrock