Rotary encoder best option here?

JZerb

Member
Join Date
Oct 2016
Location
Here
Posts
421
so im trying to figure out a little setup in my head for a machine and im looking for some advice. im using a micrologix 1400 so i figured i could use the HSC if i were to go the encoder route. i want to have positioning feedback so know where the table is within a 360 degree rotation, it will never go more then 270 usually.

with that said, i have a motor, that has a gear ratio of 1003:1 which has a gear attached to it that meshes with another gear which in turn rotates a table. the gear ratio from the gear attached to said motor output shaft to the gear on the bottom of the table is 350(ish):1. i say ish because the exact ratio will just change the math used in the logic and not much else i believe.

would an incremental encoder (quadrature) attached to the top of the motor be the best option for this setup? or is this a situation where i should just setup some prox sensors and flags at each position that i need the table to stop at in its rotation and be done with it? any help/advice is appreciated. 🍻
 
1003 x 350 = 351,050 rotations to rotate table 360 degrees?
Typical encoders are 100 - 2500 Pulses Per Revolution - PPR.
A 1000 PPR encoder would give you 351 million counts. You would do this only if you were trying to position within a millionth of a degree. Highly precision, with zero backlash gears, etc.
I think it's too many counts to use an encoder on the motor. Not a problem if the motor isn't too fast, and the HSC can handle the speed. Counts will bounce around a bit when unit is stationary, but not your display if you don't go too many decimal points.

I would try to put it on the output of the motor/input of the 350 ish gear. A 1000 PPR would give you positioning around a thousandth of a degree, 100 PPR a hundredth.

Here's a simple page that covers the basics without getting too deep.
 
so im trying to figure out a little setup in my head for a machine and im looking for some advice. im using a micrologix 1400 so i figured i could use the HSC if i were to go the encoder route. i want to have positioning feedback so know where the table is within a 360 degree rotation, it will never go more then 270 usually.

with that said, i have a motor, that has a gear ratio of 1003:1 which has a gear attached to it that meshes with another gear which in turn rotates a table. the gear ratio from the gear attached to said motor output shaft to the gear on the bottom of the table is 350(ish):1. i say ish because the exact ratio will just change the math used in the logic and not much else i believe.

would an incremental encoder (quadrature) attached to the top of the motor be the best option for this setup? or is this a situation where i should just setup some prox sensors and flags at each position that i need the table to stop at in its rotation and be done with it? any help/advice is appreciated. 🍻


What's the RPM of the motor?
How many motor or gear_1 or gear_2 rotation = one table rotation?
What is the maximum frequency of an HSC input on the micrologix 1400?
 
What's the RPM of the motor?
How many motor or gear_1 or gear_2 rotation = one table rotation?
What is the maximum frequency of an HSC input on the micrologix 1400?

motor rpm is 1,750
i would have to figure out how many rotations of the gear thats attached to the motor would have to go through in order for the table/gear to make a full rotation.

1003 x 350 = 351,050 rotations to rotate table 360 degrees?
Typical encoders are 100 - 2500 Pulses Per Revolution - PPR.
A 1000 PPR encoder would give you 351 million counts. You would do this only if you were trying to position within a millionth of a degree. Highly precision, with zero backlash gears, etc.
I think it's too many counts to use an encoder on the motor. Not a problem if the motor isn't too fast, and the HSC can handle the speed. Counts will bounce around a bit when unit is stationary, but not your display if you don't go too many decimal points.

I would try to put it on the output of the motor/input of the 350 ish gear. A 1000 PPR would give you positioning around a thousandth of a degree, 100 PPR a hundredth.

Here's a simple page that covers the basics without getting too deep.

the positioning im looking for in no way needs to be this accurate to be honest. thats why i wasnt even sure if the way to go here was to use an encoder, or just basic prox sensors and flags to get the positioning that i desire. the motor will not be moving quickly. as i said above about 1,750 rpm, say 40-60hz, so it moves pretty slowly.

How accurate do you need it to be? Are you going to try to position the table using the motor and encoder feedback?

not highly accurate. yes i would be using it for positioning of the table
 
Nobody mentioned yet, that you will need either an absolute encoder or some procedure of zero seek at least after the PLC power up, since an incremental encoder HSC reading is valid only after valid preset.
Absolute encoder is unlike with MicroLogix1400.
Does your application allow rotating the table from unknown position for zero seek purposes? If yes, the PLC has to know the zero seek direction. This means, that the zero sensor target cannot indicate only the zero position. It must cover the entire single side from the zero position, with no interruptions.
 
Good point. Put the encoder directly on the table.
Use the third output (Z - Index) for your home position. You need to be slow enough to catch it repeatably. If not, then you need to seek, find, go past, then slow in reverse until you find it again.
 
I deliberately did not mention the Z signal from the encoder, even if it is attached directly to the table (which may be not easy task itself).
As I said, the PLC has to know the zero seek direction. So, even with Z signal in use, an additional sensor is required for the search direction.
 
JZerb, are you planning to have some end switches for the table rotation?
If yes, they may be used for the first time position validation.
Of course, the HSC preset values will be different for each sensor found.
And the application has to allow the sensor seek move, even with non-empty table.
 
Does your application allow rotating the table from unknown position for zero seek purposes? If yes, the PLC has to know the zero seek direction. This means, that the zero sensor target cannot indicate only the zero position. It must cover the entire single side from the zero position, with no interruptions.

Here in lies what I think may be my biggest issue with using an encoder. Unfortunately I do not have the luxury of being able to seek a zero position upon power up if the machine were to go into a power loss state.
 
Probably the best will be an absolute encoder with parallel output interface.
Attach it to the table so that it never exceed a single turn, and connect the required number of its higher bit outputs to the PLC inputs.
This is not a high-speed count, but neither do I think you need it, taking into account slow response of AC motor and low resolution required.
 
motor rpm is 1,750


A typical encoder and a very economic one (I often buy Sick optical encoders)with 2048 PPR connected for four fold counting will still be under the 100Khz limit of the micrologix 1400 and will give you high resolution. Combined with the gear ratio you mentioned you will achieve excellent positioning precision.

1750RPM / 60 = 29.167 RPS
2048PPR x 29.167 = 59734.016 PPS or about 60Khz
2048PPR x 4 (Four fold) = 8192
8192 x 350 (Gear ratio) = 2867200
You can see where this is going,after you've figured out how many motor revolutions per one table revolution you will have very high count to divide over your 360 degrees.


As for your original question regarding setting up a prox sensor, I consider that to be too amateurish and a complete waste of money as I pickup an optical encoder for $177.00 Canadian dollars.


Edit: Someone commented once that he would never mount and encoder to a motor and was quite adamant about it but I have no idea why. Unfortunately I cant find that thread.
 
Last edited:
I’m sure others will disagree and probably have a bigger bag of tricks than I do, but positioning within a certain degree of accuracy with the micrologix and a Vfd isn’t going to yield the results I think you might be after or it didn’t for me when I last had to do one.
You can roll your own motion control but it won’t be as accurate and the amount of code needed to handle different states can get tedious.
If you can or have the say so, I would suggest just purchasing a servo system. All of the motion would be done in the drive and all of the functions we are discussing are already included for you.
Not to say it can’t be done but just because your encoder is highly accurate doesn’t mean you can accurately position. Enters the motion planner...
Just my few dirty pennies...
 
Yes, this would be the most natural and straightforward solution.
Not an answer to the original question, though.
And interfacing a servomotor absolute encoder with ML1400 is not that simple, if at all possible with the specific servodrive model.
AC motor may be used for positioning, keeping in mind that the positioning will be fast or precise, but never both.
 

Similar Topics

Hello Guys, In order to measure the length of a product using a photo-eye and a Timer .. It's not a reliable operation even if you got it right...
Replies
32
Views
10,275
Our store has few DIGITAL DEVICES Rotary Encoder which is original used on obsoleted USA weighing scale for speed feedback. But we has no idea how...
Replies
2
Views
2,197
Hello, all. We have one 840Dsl milling machine and a (absolute rotary angle) encoder alarms show up. We took it out for checking, then put...
Replies
1
Views
1,465
Hi, Everyone I am interfacing rotary encoder(incremental encoder) with PLC. This encoder using for speed measurement (RPM). The encoder is giving...
Replies
6
Views
4,515
I am needing a rotary encoder switch. Preferably, it would have screw terminals for connections. I'm looking for it to have a digital pulse as...
Replies
7
Views
3,374
Back
Top Bottom