Stopping a revolving liquid tank agitator in the same location electrically

Status
Not open for further replies.
Trying to dream up a method to stop a mixing machine agitator in the same location every time. The mixer is a 300 gallon round tank in a horizontal position. The agitator revolves horizontally and is most likely going to be driven by a 7.5 hp 480 3ph motor. Controls would include a VFD and will have a PLC for operator controls. My thought was to use a rotary encoder for positional feedback and brake the agitator with the vfd. Would like to avoid any mechanical brakes for simplicity. Has anyone done a similar system? Only in the planning stage right now so I don't have all the specs for the system. Thank you

Sent from my SM-G950U using Tapatalk
 

sii

Senior Member
Location
Nebraska
Having a hard time picturing your application but if there's any part of the main shaft exposed could you put a "dog" on the shaft and use a proximity sensor to trigger the deceleration ramp? Also no idea your required level of precision.
 
Think a big meat mixer. Access door is closed on the front during the mixing process. (It is actually a cleaning process.). I have not seen the agitator yet but apparently it will only allow removal of the cleaned items from one stationary position.

Sent from my SM-G950U using Tapatalk
 

sii

Senior Member
Location
Nebraska
I should have been more thorough. We have an application in our plant similar to what I described above. At the end of a cycle I use the dog to trigger a slow down from run speed to about 15%. Then the next time the sensor triggers I ramp to stop from the slower speed. I does provide a very repeatable stop.
 

Besoeker

Senior Member
Location
UK
Different application - machine tools industry.
We used motor shaft resolvers. From 10,000 rpm to zero in half a second to within +/- 0.5 degrees of position accuracy.
Not saying the OP needs anything like that, but it shows what can be done.
 

kwired

Electron manager
Location
NE Nebraska
simple from programming perspective may be a manual "jog" at a low speed and somehow indicating what position to stop at?
 

Ingenieur

Senior Member
Location
Earth
just a prox switch
when a stop is commanded do not initiate until the prox is triggered
measure the coast down distance several times
verify consistent and use the avg to position the prox
allow for some adjustment of the switch and target
 
I should have been more thorough. We have an application in our plant similar to what I described above. At the end of a cycle I use the dog to trigger a slow down from run speed to about 15%. Then the next time the sensor triggers I ramp to stop from the slower speed. I does provide a very repeatable stop.
We have a couple
Cutler hammer E59-m18c prox switches available. They should do the job. Does you application have paddles on the interior? I'm just wondering if the weight inside the drum would throw off the stop position. Depending on how far out of balance the load is.



Sent from my SM-G950U using Tapatalk
 

kwired

Electron manager
Location
NE Nebraska
We have a couple
Cutler hammer E59-m18c prox switches available. They should do the job. Does you application have paddles on the interior? I'm just wondering if the weight inside the drum would throw off the stop position. Depending on how far out of balance the load is.



Sent from my SM-G950U using Tapatalk
I would think if driven load isn't the same weight or speed every time you initiate a stop, simple timing methods wouldn't stop it in same position very consistently.

How much mechanical speed reduction there is between the motor shaft and driven load will make a big difference in how easily you may be able to accomplish this also.
 

Besoeker

Senior Member
Location
UK
just a prox switch
when a stop is commanded do not initiate until the prox is triggered
measure the coast down distance several times
verify consistent and use the avg to position the prox
allow for some adjustment of the switch and target
Probably not of coasting given the application.
A motor shaft mounted brake might be a solution.
 
As I am thinking about it I will probably need some type of mechanical brake. I'm envisioning the machine operator reaching into barrel to remove cleaned items. Don't want anyone losing body parts in this thing.

Sent from my SM-G950U using Tapatalk
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
VFD on the motor and a Proximity sensor or photo eye looking at something on the shaft of the rotating element. When a Stop command is given, the VFD is put into a “Creep Mode”, meaning slow speed / low Hz. That then enables the sensor and when the target passes, the VFD goes into Braking mode to stop the agitator. Depending on the mass in the drum and the duty cycle involved, you may need a VFD with Dynamic Braking using external resistors. If the duty cycle is very low, ie once or twice per hour or something like that, you could just use DC Injection Braking which is built-in on most VFDs. The difference is that DCIB dissipates the kinetic energy of the rotating mass into the motor itself, so it adds thermal stress to it, whereas Dynamic Braking dumps that energy off into resistors as heat. If you can, you could use two sensors, one to put it into creep mode and use Dynamic Braking, the second one initiating the hard stop as you get closer to the final desired position by triggering DCIB.

You may need to play with it a bit in that if the stop command is initiated when the trigger point is too close, it may not be able to stop it in time. So what you do is to observe that time it takes to stop and put a timer in that sensor circuit so that if the sensor is triggered before there is enough time to stop, it lets the agitator go around one more time in creep mode.

You COULD do this with an encoder or resolver, but honestly, that’s going to be more complicated than you might think. But we don’t know your level of programming savvy, so it’s up to you to decide. Using an encoder/resolver is more of a “motion control” application which means constantly tracking the position at all times and divining a deceleration profile to get the load to stop at an EXACT spot. That then requires additional intelligence in the control system, which means higher costs etc. It didn’t sound as though you need that level of resolution.
 
Last edited:
VFD on the motor and a Proximity sensor or photo eye looking at something on the shaft of the rotating element. When a Stop command is given, the VFD is put into a “Creep Mode”, meaning slow speed / low Hz. That then enables the sensor and when the target passes, the VFD goes into Braking mode to stop the agitator. Depending on the mass in the drum and the duty cycle involved, you may need a VFD with Dynamic Braking using external resistors. If the duty cycle is very low, ie once or twice per hour or something like that, you could just use DC Injection Braking which is built-in on most VFDs. The difference is that DCIB dissipates the kinetic energy of the rotating mass into the motor itself, so it adds thermal stress to it, whereas Dynamic Braking dumps that energy off into resistors as heat.

You may need to play with it a bit in that if the stop command is initiated when the trigger point is too close, it may not be able to stop it in time. So what you do is to observe that time it takes to stop and put a timer in that sensor circuit so that if the sensor is triggered before there is enough time to stop, it lets the agitator go around one more time in creep mode.

You COULD do this with an encoder or resolver, but honestly, that’s going to be more complicated than you might think. But we don’t know your level of programming savvy, so it’s up to you to decide. Using an encoder/resolver is more of a “motion control” application which means constantly tracking the position at all times and divining a deceleration profile to get the load to stop at an EXACT spot. That then requires additional intelligence in the control system, which means higher costs etc. it didn’t sound as though you need that level of resolution b
That certainly sounds like a solution. My initial thoughts were to use an encoder for picking up multiple positions. I have never used an encoder for this purpose and dont fully understand them. Would the encoder work to locate position at 45 degrees from stop location and also locate full stop position? I can't imagine it is very complicated.

Sent from my SM-G950U using Tapatalk
 
I will have more information in the coming week. Leaning towards a prox. Encoder seems like it would work but I'm no expert in programming for sure. Going through availability of washdown duty components sends item cost up in a hurry. Will need to determine final requirements for positional accuracy and safety interlocks as well.

Thanks everyone for the input. The wealth of knowledge on this forum is remarkable. Gonna read up on my PLC capabilities. I will keep thread updated as I get more information.

Sent from my SM-G950U using Tapatalk
 

Besoeker

Senior Member
Location
UK
As I am thinking about it I will probably need some type of mechanical brake. I'm envisioning the machine operator reaching into barrel to remove cleaned items. Don't want anyone losing body parts in this thing.
Quite right. Here (UK) The supply to the motor would have to be locked off before an operator was allowed near the moving parts. What you guys refer to as LOTO.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
An encoder doesn’t “do” anything, it just provides you with pulses as it revolves, ie 3600 PPR (pulses per revolution). You need something that tracks the rotation using these counts. Then you either need to “home” your counting every time you cycle power, or use what’s called an “absolute encoder” that knows where it was after powering back on. SOME very advanced drives are capable of being the intelligent device that does the tracking of the encoder pulses, those are not going to be the inexpensive ones.
 
Status
Not open for further replies.
Top