I am interested to know if anyone has any experience using a Milltronics Compu-M belt scale (dual load cells) to input analog 4-20mA to a PLC5? According to the Milltronics specs, the 4-20mA output is proportional to the "flowrate". Since the Compu-M is an integrator for load, speed, flowrate, and total flow, I am wondering if the analog output can provide more than just flowrate? According to Siemens, belt scale LOAD is equal to Flowrate over Speed (of the belt). One would think that once analog data is input to the PLC it could be used at least to calculate the other quantities.?. This system is using a MD-36 speed sensor for belt speed (36 ppr).

Anyone have a bit of exerience using a belt scale input to a PLC5 (or any other PLC analog input) to trend and/or manipulate data? Thanks for any information.

It's a one dimensional analog signal. Let's get rid of the terminology here for a moment and consider what you are trying to do. X = load cell reading, Y = speed, and Z = output (tons/hour). So, Z = X * Y.

Now, you are asking to recover Y from Z. Okay, this is easy enough. Y = Z / X. If X is a constant, then you can easily recover Y, except if X happens to be zero (0). In which case Y is indeterminate. Or rephrasing it the original way, you can't determine Y if X is 0, because 0 = 0 * anything. L'Hopitale's rule won't help here either (derivatives).

So as long as X is a constant other than 0, then, you can recover Y. BUT from a practical sense there's no way to actually hold X (the load) constant, so there's no way to recover Y.

Algebraically, you have two unknowns so obviously you need two different equations to get there, and you only have one.

Electro Sensors makes a very good tachometer that's just as rugged as the Milltronics stuff and outputs 4-20mA on a redirected output with roughly the same resolution as what you have now (typically 10-30 ppr). They are expensive but you don't have to worry about any sort of external encoder reading device. If you just want to know motion (whether the belt is running or not), their motion detectors can go right on the tail pulley and are amazingly rugged devices.

Otherwise, you can go with the outright rotary encoder system which gives you both higher accuracy and speed. In this case, take a look at something from Dynapar, encoder.com, or BEI. Just get an incremental encoder that matches your mechanical system. I usually go for hollow shaft since I can usually find something appropriate and it eliminates couplings. If you have a NEMA C face motor, that's great because you can mount the encoder directly between the motor and the gear box but with conveyors, that's usually not the case. Pick an appropriate electrical spec. but avoid SSI for incremental encoders (best for absolute encoders). The best encoder output is differential with a line driver, BUT with the recommendation of interface cards I'm going to make, this may not be useful to you anyways.

Since you are using a conveyor belt I have two suggestions as to mounting location. Either put the encoder on the tail pulley, on an idler wheel that rides on the belt, or directly on the motor. If you put it on the tail pulley or on an idler wheel, if the belt breaks or the drive shuts down, you know immediately because you directly measure the belt starting/stopping. Belts run slow usually with a large reduction ratio from the motor to the head pulley. So if you don't measure actual belt speed, putting it on the input side of the reduction ratio gives you a higher resolution or higher response speed since the shaft is moving much faster. Putting it on the head pulley may be the most obvious place but it is the worst place.

Put in a 1794-ASB (for Remote IO unless you are using some other network), a 1794-TB3, and put a 1794-IJ on it. It is a little confusing but this particular IO card gives you direct (RPM) readings for up to 2 inputs. With the other high speed counter cards, you have to do some math and timing to get that number.

Thank you very much, Paul, for your reply. This Milltronics scale and tail spool encoder are already installed, but have not been used for quite some time. At this time wood chips are not selling for a very good margin so we are using them for fuel to a biogen power plant. We are needing to primarily know how much product we are sending to the fuel storage area... thus, the belt scale and encoder. I have one analog channel available on an A/B 1771-IFE in the remote I/O rack that is located within 500 feet of the Compu-M. Accordingly, the 4-20 output will be sent to the PLC5 to allow us to trend/log data to a file (MS Excel or RSView or ??) The Milltronics product/programmer lit really does not go into much detail about the composition (dimensioned data) in its output. We are most interested in product flowrate.

So, I am thinking that the output can be sent directly to the A/B 1771-IFE and from there we can do whatever calcs we want for the unknowns?? I am not sure how the Milltronics performs the integration of load, speed, and total flow, but if the flowrate is in fact output on a 0-15mA or 4-20mA scaling, all I probably need to do is simply program the Params on the Compu-M and send the output to the PLC, do some ladder logic for data arrays and be done with it.

I appreciate your input and suggestions.

Assuming that you have the belt scale calibrated properly and cleaned out at least 2 idlers upstream and downstream of the weigh bridge, that's what you will get.



Yes, that's all you have to do. You will get an instantaneous output (as in either tons/hour or lbs/minute or something like that). Then, since you probably want to totalize it, just put it into an ADD (or CPT) statement inside an STI ladder. I can vouch for how well this works because we used it to totalize huge (500,000 gallon) batches of clay slurry using the exact same hardware when I worked in Georgia.

The only other things we had were a motion switch to indicate when the belt was stopped unexpectedly, and a FIFO to deal with the difference between when material discharged and reasonably fast response speeds at the feeder end of things (getting chemical dosages for dispersion correct with clays is critical to not turning the entire works into a giant mud ball).