Poor man's encoder

[dickieus]

I have just been given a situation where I need to synchronize 2 belt speeds. The original system used a set of 8 Prox switches on each belt looking at a wheel with one slot feeding to siemens logic blocks.

What I was given was the Prox inputs to digital inputs on a micrologix 1500. they are taking the word on 8 bits and srnding it to an interger file. Problem is the numbers can never be the same.

To match speeds they are trying to use an NEQ instruction and then if not equal add or subtract 1. This they are sending to an analog output.

I'm open to any ideas.


P.S. the master belt has a "jerk". The thinking is the prox system would be better than using encoders and HSC.

P.P.S. They also are adding an analog tacho-generator to the mix.

 

[kamenges]

First of all, do they need to be just speed synchronised (proxes on both conveyors change at the same rate) or do they need to be phase synchronised (the same prox on both conveyors changes at the same time).

Second, did this system ever work to the end user's satisfaction and, if so, why are you changing it?

Third, does the slave conveyor have a speed baseline based on the master commanded speed and the proves just provide a speed tweak or are you trying to develop the whole speed command based on the proxes?

Keith

 

[dickieus]

Hi Keith,Thanks for the reply.

I just need to match the speed of the frst belt to the Second.

They are replacing the old logic blocks with a single PLC.
the whole line includes 10 belts, plus gates,shuttles and solenoid valves. The other belts are controlled by a tacho-gen and tweaked with pots.(DC motors).

The belt I need to control by the prox system has a VFD drive.

The tacho-gen is on the machine sending product to the line and is the main speed reference.

Thanx,
Richard.

 

[PeterW]

Why don't they feed both belts from the same inverter? Just have to ensure the gearing is matching.

 

[Jiri Toman]

P.S. the master belt has a "jerk".

That is common on belt driven conveyors. If you just have a set speed motor (no VFD) you would normally add a soft start to take care of the jerk. I am not sure in your case why you would have a jerk if your motor has a drive on it? Slow down the accel rate.

If you need to tweak the speed of your follower motor, I suggest that in addition to bringing in inputs from proxes to obtain the speed reference, wire in a multiturn pot to the PLC. This will require an analog input on the PLC. Then you can scale the analog output to the follower motor with this pot.

In as far as the calculation of the speed by using NEQ instruction etc. I am not sure what that is supposed to do. Normally you would create a scheme within your PLC logic to emulate a rate meter. There have been numerous discussions on this forum on how to do that. I would just count pulses during 1 sec intervals and then average over 10 seconds. Use shift register to slide the data fwd.
Assign a scaling multiplier constant to the tweak pot analog value and multiply the rate meter value with it. Then move the result to the analog output.

 

[rsdoran]

Not trying to be funny or mean but gee whiz who designed this deal, a kindergartner?

Many of the Micrologix can take high speed inputs, 20khz, without the use of an HSC. In your situation you are using a few hundred dollars worth of prox's and 8 inputs where you could have used one $100 encoder and 1-3 inputs.

The other aspect of this is that with VFDs and encoders if both belts have VFDs then you do not really need a PLC, the drives can be used in master/slave setup. Now that I think about it I believe you could use an encoder on the master belt as the speed reference to the second belts VFD, havent done this but I think it is possible.

As mentioned if the master drive is on a VFD you can remove the "jerk" by increasing ramp time.

It is possible that the "master" motor/belt is fixed and you can not touch it, just have to match the next belt with it. In that case you would not care about the "jerk", just allow it to run 10s or so to reach operational speed then start second motor and match it.

Another thought is that if the belt speeds are fixed then their is no need to use encoders, proxs, etc, just use the VFD and/or gearing to match their run speeds. As mentioned, one method would be to use one VFD to drive both conveyors. If they have separate controllers then just use an rpm/fpm meter (http://www.multimetercenter.com/tahcometer.htm) to get the belt speeds matched by setting the drive parameters.

 

[dickieus]

Time warp

Yes this place is quite the time machine. We are "upgrading" some older equipment. The goal is to replace the old logic blocks with a single PLC. The master is on a moulding line and the second belt is on the packaging line. The jerk appears to be mechanical. What I'm after is positioning for the packaging equipment. I'm going to try Jiri's rate meter idea. It looks like I will also have to include an adjustment for drift.

 

[rsdoran]

I am curious, how are the prox's setup i.e. can they count at the same time or does it just count 8 pulses per revolution?

 

[dickieus]

proxes count 8 times per rev of wheel. this is feed to a seperate input card for each set. they take the input as 1 word through a MVM instruction to get a value.

 

[kamenges]

So you have a circular array of 8 proxes around a single wheel and the wheel has a single slot in it. Not oly is it a poor man's encoder (although, as Ron correctly pointed out this is more expensive than a true encoder) it is a poor man's ABSOLUTE encoder.

One thing that people do too often when upgrading systems is simply try to directly replace what is there with a newer version of what is there. Resist that urge. You want to really look at the application as if you are designing it from the ground up. Use the technology that makes sense, not a new version of a technology that was used in the past just becasue nothing else was available.

If you really are going to stick with the prox setup and you aren't, concerned with phase then Jiri's suggestion is the way I would approach it. Just keep in mind that the previous system would most likely have settled in to a particular phase relationship. Make sure you really don't need the phase relationship before going on.

Keith

 

[Dave_in_Oz]

I have engineered a few Caterpullerrs/hauloff upgrades for the cable and plastics industry, where the top and bottom belts are seperatly driven and therefore need to be speed matched to avoid damage to the product.
All I used were two 690 SSD inverters, one set up as master
and the second as a slave with the encoder that was mounted on master motor fed to the slave drive to act as a speed reference.
Can't see a need for a PLC on this job in my opinion.