Incremental encoders vs absolute encoders

rejoe.koshy

Member
Join Date
Dec 2011
Location
kolkata
Posts
193
Hey guys,
I have been to trying understand the advantages of using absolute encoders over incremental ones & I am told that no homing process is required (when used for positioning) for absolute encoders on power resets, since they give absolute position.

This concept has not sunk in well, since the current position is always retained by the controller /CNCs etc & can start positioning post power resets irrespective of whether the feedback device is absolute or incremental.

Need a little help with this since I have to choose an encoder for a positioning application for my upcoming project.

Pls help!!
 
This concept has not sunk in well, since the current position is always retained by the controller /CNCs etc & can start positioning post power resets irrespective of whether the feedback device is absolute or incremental.

Well, here is a little thought experiment. Turn the machine power off. Move one of the motors by hand. Then turn power back on.

How does the controller know the new motor position in case of an incremental encoder?
 
Well, here is a little thought experiment. Turn the machine power off. Move one of the motors by hand. Then turn power back on.

How does the controller know the new motor position in case of an incremental encoder?

Some systems (robots in particular come to mind) use batteries to maintain the encoder value without power. This of course requires that the batteries be changed when necessary.

I've heard the arguments in favor of absolute encoders, frankly I'm not convinced, at least in the arena of automated equipment in a factory environment. Is it really that hard to just make the machine home itself when the power comes on? With absolute encoders you eliminate the repetitive homing but what if Bubba changes a motor gearhead at 3 a.m.? Or a motor overtorques and the coupling slips? Yep, position lost... There needs to be a manual homing operation for these situations, and more importantly, the operators and/or maintenance people need to know when and how to activate it. It's better to just make it automatic in my mind.
 
Well, here is a little thought experiment. Turn the machine power off. Move one of the motors by hand. Then turn power back on.

How does the controller know the new motor position in case of an incremental encoder?
Adding to that idea, imagine the controller dies. As in: the circuit board housing the processor and memory is completely, irrevocably, and forever after unusable. Like maybe somebody put a forklift through the control cabinet. And then the motor suubsquently drifted a bit. How will the new controller know the position?
 
I've heard the arguments in favor of absolute encoders, frankly I'm not convinced, at least in the arena of automated equipment in a factory environment. Is it really that hard to just make the machine home itself when the power comes on? With absolute encoders you eliminate the repetitive homing but what if Bubba changes a motor gearhead at 3 a.m.? Or a motor overtorques and the coupling slips? Yep, position lost... There needs to be a manual homing operation for these situations, and more importantly, the operators and/or maintenance people need to know when and how to activate it. It's better to just make it automatic in my mind.

Don't quote me on this, but I think that overall you get better precision with an absolute encoder than a pulse encoder.

The other "problem" with the absolute encoder is that it would fail much less than a pulse encoder (as shutting down a machine won't affect it) and Bubba won't know off the top of his head how to set it up at 3AM. Obviously, proper documentation is key... but we all know how that goes.

One particular case of using absolute encoders to their strengths is on coordinated movement. Imagine a column that is moved from side to side by two trolleys (one above and one below), if you had to re-home the machine frequently, you'd be stressing it more than it had to be, whilst with absolute encoders this is much less of a problem.
 
Hey guys,
I have been to trying understand the advantages of using absolute encoders over incremental ones & I am told that no homing process is required (when used for positioning) for absolute encoders on power resets, since they give absolute position.

This concept has not sunk in well, since the current position is always retained by the controller /CNCs etc & can start positioning post power resets irrespective of whether the feedback device is absolute or incremental.

Need a little help with this since I have to choose an encoder for a positioning application for my upcoming project.

Pls help!!
Hello,

To add to others' comments.

For our servo motors we use encoders (From SICK) that have both absolute feedback and incremental/sincos technologies in one.

On top of that the encoders can be singleturn or multiturn.
- Singleturn (the encoder measures an absolute value within one revolution on boot up and then continues to count incrementally from this point)

- Multiturn (the encoder measures an absolute value within multiple revolutions(4096 in this case) on boot up and then continues to count incrementally from this point.

Homing: Homing/referencing in positioning systems is necessary if running in absolute positioning mode and is not necessary if running in relative positioning mode.

- If using the above singleturn encoders on boot up the absolute position will be provided only within one revolution, but there is no information about how many revolutions like the multiturn. So, homing/referencing becomes necessary on every boot up (assuming that the motor has moved beyond one revolution during normal operation).

- If using the above multiturn encoder , then homing/referencing is necessary only once (unless 4096 revolutions have been surpassed during normal operation). After boot up the absolute position will be provided within 4096 revolutions.

So, if you need absolute position retained after power loss then look into singleturn or multiturn absolute encoders or the combination technology(abs/inc both in one), keeping in mind homing/referencing on boot up might be necessary in case of singleturn. If price is a concern then incremental encoders can be used with homing based on the encoder marker pulse etc.

I hope this helps.
Thanks.
 
Well, here is a little thought experiment. Turn the machine power off. Move one of the motors by hand. Then turn power back on.

How does the controller know the new motor position in case of an incremental encoder?

Ok...What if I rotate the shaft of the motor by 360 degrees from the present position?The controller still woudnt know the new position!!
 
So to conclude I only need to use absolute encoders to register any movement which might take place during power off condition......also to say that if I am sure my axis will not move one bit (example- servo motor with brake) I can use incremental encoders & that will suffice.
 
Hello,

To add to others' comments.

For our servo motors we use encoders (From SICK) that have both absolute feedback and incremental/sincos technologies in one.

On top of that the encoders can be singleturn or multiturn.
- Singleturn (the encoder measures an absolute value within one revolution on boot up and then continues to count incrementally from this point)

- Multiturn (the encoder measures an absolute value within multiple revolutions(4096 in this case) on boot up and then continues to count incrementally from this point.

Homing: Homing/referencing in positioning systems is necessary if running in absolute positioning mode and is not necessary if running in relative positioning mode.

- If using the above singleturn encoders on boot up the absolute position will be provided only within one revolution, but there is no information about how many revolutions like the multiturn. So, homing/referencing becomes necessary on every boot up (assuming that the motor has moved beyond one revolution during normal operation).

- If using the above multiturn encoder , then homing/referencing is necessary only once (unless 4096 revolutions have been surpassed during normal operation). After boot up the absolute position will be provided within 4096 revolutions.

So, if you need absolute position retained after power loss then look into singleturn or multiturn absolute encoders or the combination technology(abs/inc both in one), keeping in mind homing/referencing on boot up might be necessary in case of singleturn. If price is a concern then incremental encoders can be used with homing based on the encoder marker pulse etc.

I hope this helps.
Thanks.


Thanks for the wonderful explanation Mohua!!...is there any advantage when using absolute encoders when comes to reliability & why?
 
Ok...What if I rotate the shaft of the motor by 360 degrees from the present position?The controller still woudnt know the new position!!

That is what multi-turn absolute encoders are for :) As it was mentioned above, a small battery maintains power to the encoder electronics so the accurate motor position is available any time without the need to perform homing (origin search, reference seek etc.) sequence.
 
So to conclude I only need to use absolute encoders to register any movement which might take place during power off condition......also to say that if I am sure my axis will not move one bit (example- servo motor with brake) I can use incremental encoders & that will suffice.
That's just one of the many scenarios though. When using encoders for Vector Control of an AC motor on a hoisting application for instance, the purpose of the encoder is to provide a feedback into the Vector Control algorithm so that the motor develops the precise amount of torque producing current to hold the shaft in place when you release a mechanical brake. If you are wrong and the load BEGINS to drop, it takes a lot more torque to STOP it again than it did to hold it, and you might not have that much available. So also you cannot "home" that process, because that too means dropping the load! This has nothing to do with whether or not there is a power failure, it has to do with knowing an ABSOLUTE position at any given moment, including when it was not moving. So if power never failed, but you have an incremental encoder, how certain are you that NOTHING moved while the brake was set? Are you willing to bet someone's life on that?

What it really boils down to is whether or not YOUR process requires knowing where something is when it starts, without having to find a "home" position first. If you are OK with finding a home position, or with occasionally redetermining it to maintain a level of accuracy, then incremental is fine. LOTs of applications are fine with incremental encoders.

Accuracy is related to the PPR (Pulses Per Revolution) and bit conversion resolution, either of which can be whatever you need in both incremental and absolute encoders. It's not about that.
 

Similar Topics

i have an application that has AB 845H incremental encoders connected to VFDs (Powerflex 70ECs). i can get the "raw encoder feedback data" to my...
Replies
3
Views
2,396
i have documentation that speaks of using a high resolution "absolute" encoder with the Powerflex 700s for "absolute point to point positioning"...
Replies
10
Views
7,736
Hi all, I am using a B&R X20 unit and want to integrate an incremental encoder (OUT A, B, & Z). I have three BOOL digital input pins for my A,B...
Replies
3
Views
396
I have a question regarding the features of the so called incremental (velocity) form of the PID controller. As far as I understand this form in...
Replies
14
Views
2,385
Hi, I have some Baldor (Baldor is now ABB) servo motors that have revolvers for position output. But to my surprise the Baldor MicroFlex...
Replies
11
Views
3,707
Back
Top Bottom