RMS Motor Current

A few years ago I had a problem with a motor running a hydraulic pump. We used a hand held data collector to get the motor current throughout the cycle. Then my hydraulic guy took the data dumped it into Excel and calculated the RMS current of the motor. The idea was the motor could go into overload though the cycle as long as the RMS current of the motor was less than the rating on the motor. So he broke the cycle down into chunks where the current was constant. Took the amps and the time and calculated the RMS current of the motor though the cycle. We found that we could remove spikes in the current lowering the RMS current and keeping the overload from tripping. All things were well.

Fast-forward 2 years and I have a similar problem on a different machine. I have installed a current sensor on the motor and took it into my plc. I already have a way to collect the data and store it so no problem. I took this data and did the same type of calculation my old hydraulic guy did. I would like to take this to the next level. I would like to do these calculations on each part doing an RMS check on the entire cycle and do some peak readings at different times. The peak readings are no problems but I don't know the best way to do the RMS. The overall cycle changes depending on the part they run so it has to be somewhat flexible. I was thinking about breaking the cycle down into the segments I want to get the peak in. Then doing an RMS of each segment then an overall.

Has anyone done this type of thing before? I don't see any reason I can't do this but I don't want to put time into something that could be done easier.

Comments and suggestions welcome

Charles, what I think you are asking about is RMS loading of the hydraulic pump motor. I suspect the hydraulic technician was measuring the current (which may have been with an RMS current meter) to determine the loading on the motor at various points in the process. Then he used that data to calculate the RMS load on the motor. If I am correct, then the RMS current and the RMS load are two different (but related) things at least so far as what I think you are wanting to figure out.

RMS load determination is commonly done in hydraulic pump motor sizing. In a hydraulic system electric motor power requirements are proprotional to the flow rate * the pressure/pump efficiency. If you are metric , then flow rate in m^3/sec * pressure in Pascals)/efficiency = watts required. If you are using US units then flow rate in GPM*Pressure in PSI/(1724*Efficiency) = motor HP needed. But this is a ball park figure, and it frequently results in an oversized motor.

One thing that happens often in industrial settings that that a pump is correctly set up to limit pressure and or flow to keep from overloading a motor. But down the road, Bubba tweaks one of the pump adjustments set screws and then Cletus tweaks one, and then Beaudreaux tweaks one and next thing you know the product of the pump flow and pressure output exceeds the horsepower input, and it starts tripping motor overloads all the time.

Many hydraulic system loads are brief in duration and occcur at predictable intervals. Therefore, sometimes you can get away with using a smaller electric motor by allowing the motor to run slightly overloaded for a few seconds out of a longer time period of a few minutes. Higher frequency and longer duration loads increase the RMS load on the motor and decrease the likelyhood that you can get away with a smaller motor. This is not a trival thing to do however, but sometimes the pay off can be worth it. I just completed an RMS loading study on a new system we are designing and found we could drop the HP requirement by 10HP. But Bubba can't go and start tweaking the system without messing things up big time unless he knows what he is doing.

Before we go any farther though, is this a new system? Or is it an existing system where the load requirements have recently and intentionally changed? Or is this an existing system that has recently begun having problems and you cannot assign the cause to a recent and known change in load requirements?

Meanwhile, here is some light bed time reading:
http://www.hydraulicspneumatics.com/200/Issue/Article/False/21602/Issue

Before we go any farther though, is this a new system? Or is it an existing system where the load requirements have recently and intentionally changed? Or is this an existing system that has recently begun having problems and you cannot assign the cause to a recent and known change in load requirements?

Click to expand...

This will be on new systems. My problem is the systems are designed with the RMS mode your talking about. However my running parameters and adjustments allow me to overload the motor. The end result that I am looking for is to give information and maybe sometype of warning.

good - do the RMS calculations on each part to aid me in adjusting the cycle of the machine so I don't overload the motor.

better - give indication back so when the machine is setup they can tell if they are overloading the motor.

best - give indication back that will lead them to the part of the cycle that needs to be adjusted to unload the motor

very best - have the machine take care of itself so it wont load the motor.

I guess I'm just fishing for ideas. I would hate to go to alot of work then find out that a simple average works just as well. When we start a system we have a pre-defined cycle that the system is designed to. This tells us pressure, flow-rates HP etc. These calculations work correctly 9 out of 10 times. Depending on the part being ran it can change drasticly. Up until now I have been just dealing with it when it comes. Sometimes I can make some tweeks here and there and lower the current in the cycle. Other times I just put in an amp meter and tell them if it goes over X amps for more than X % of the cycle slow it down or what ever to lower it. I just want to put more tools into the control to get this 9 out of 10 up to 19 out of 20.

Alaric said:

Meanwhile, here is some light bed time reading:
http://www.hydraulicspneumatics.com/200/Issue/Article/False/21602/Issue

Click to expand...

Tried to look at the referance got an error message flag.

Anyway I am curious. It seems to me I can interpret this two ways:
1. You are taking peak current readings and then doing RMS calc to get the RMS current.
2. You are taking current readings with an RMS meter and then doing another RMS calc.

Why cant you use a data logger and take readings every 5 seconds (which would be an average of the RMS for a 5 second period) for whatever time period you choose. Then dump the data in Excel and do following calcs:
Max Min AVerage, std dev, and also do a graph of each data point so that you get the visual picture of what went on when?

Seems to me that taking voltage readings also would be a good idea since you are putting all this time into it.

If you put a pressure transmitter on the discharge AND were able to sense bypass valve position, you would really get a good picture of what the pump (and the motor) are doing when and under what load conditions.

Dan Bentler

What kind of pump controls do you have?

Do you have electronic stroke and or pressure controls on the pump (not downstream)? Or is the pump a non-electronic pressure compensated pump? Does it have a horsepower limiter?

Last, do you have a system that you can compare to it?

When I'm designing a new system I determine the speed and load requirements for the actuators. I then use this to determine the maximum pump flow rates required and the pressure needed for the load requirments given the actuator sizes. I use the max flow rate requirments to determine the pump displacement. Then I calculate the HP requirements for each actuator and select the highest for my pump. If the largest HP requirment for an actuator happens to be a short duration load then I start looking at RMS loading of the motor.

I have attached a spreadsheet that shows an example of how I do it. For each step I first determined the HP requirements and the duration of the load. It outputs the RMS load calculation. In this example I can see that a 40HP motor will do ok, even though I have a very brief stage that reqirues 51HP (in this particular case, a large actuator, the full 51HP load really only occurs in the last couple of seconds - although to be conservative I considered it as occuring throughout the step)

Next a certain amount of judgement comes into play. For example, if I had a 75HP load and it lasted only five seconds every couple of minutes and the rest of the time the load was under 30HP, the spreadsheet could tell me I could get away with a 35 HP motor. But that peak load is more than 2 times the motor power rating. I wouldn't do it. I would increase the motor size or try minimize the pump load if possible with an accumulator or by changing the load profile. A generalized rule of thumb is about 110% for 30 to 60 seconds every couple minutes, 125% for 30 seconds every three to four minutes, 150% for no more than 15 seconds every five minutes. Don't go above 150%. Some of this is simply a judgement call based on experience and your knowledge of your process.

Don't forget to include parasitic loads and the amount of energy required to compress the oil itself, especially if the actuator volumes are large. See this thread for a discussion on that: http://www.plctalk.net/qanda/showthread.php?t=32866

And don't forget to appropriately size the motor contactors if you are going to do this.


edit to add: If you are using a fixed displacement pump (such as a gear pump) and a relief valve to control pressure (we call them heaters because they waste energy) then throw all this out the window. HP is GPM * PSI/(1724*efficiency) and you aren't going to gain anything by sizing to your actual load. This only applies to variable volume pumps with changing load conditions.

edited again to add some more: Re-reading your post, it seems to me that you could use some motor load sensing instrumentation. You may be able to feedd this back to the PLC to achieve the best or very best objective you indicated above. Check out http://www.loadcontrols.com These units will sense the actual load on the motor and give you an analog output as well as an adjustable relay alarm point. You can then log this value and/or respond to it with your control system.

It seems to me that, if the object is to measure true average overload, a suitable method would be to measure motor temperature with an analog transducer.

That way you would include in the evaluation ******t temperature, time in overload, and depth of overload. As a bonus, you could set limits to protect the motor from overheating which would be much more accurate than the average overload block.

What kind of pump controls do you have?

Do you have electronic stroke and or pressure controls on the pump (not downstream)? Or is the pump a non-electronic pressure compensated pump? Does it have a horsepower limiter?

Last, do you have a system that you can compare to it?

Click to expand...

It is a pressure comp pump with variable pressure control done electronicly.

Last, do you have a system that you can compare to it?

Click to expand...

I can't compare it to anything because we never have the same machine running the same part. The part determins the cycle.

I have attached a spreadsheet that shows an example of how I do it. For each step I first determined the HP requirements and the duration of the load. It outputs the RMS load calculation.

Click to expand...

We do similar calculations. But the calculations don't always work because the true cycle of the machine is determined by the part it is making. The part that it is making can change as many as 6-8 times a shift.

Here is some data that I have already. This is collected each time I make a part. My idea is to use the loop that collects this data to also do the calculations. The sample time on this data is 40ms. You can see the moves by the increase in current. I did like you were saying and took the peak of each section and did the RMC cal over the entire time at the highest amps. I think however that I can do an overall just based on the sample time and do each reading.

One of the things that really stands out on that graph is that during the first spike your pump pressure is relatively low but yet motor load goes way up to ~380 amps. I also see you are moving one axis. Is this a large piston or similar load? If so then I interpret this mean that most of your energy is going into moving the oil, not creating pressure. One thing that will decrease the amplitude of this spike is if you don't have to move quite so fast, therby limiting how much your pump comes on stroke which has the side effect of limiting your horsepower requirements. There is a similar spike at the end when you retract (assuming its a piston) that axis, now you are moving oil into a rod area which has a lower volume, so the power requirement is not quite as high. These are the two points of concern as far as motor sizing is concerned.

One of the things that really stands out on that graph is that during the first spike your pump pressure is relatively low but yet motor load goes way up to ~380 amps.

Click to expand...

Its really 38.0 amps this is a 25HP motor

I also see you are moving one axis.

Click to expand...

Again like I say the cycle changes depending on the part. Depending on the parts we can have both axis moving at the same time with 3000psi on the pump.

One thing that will decrease the amplitude of this spike is if you don't have to move quite so fast,

Click to expand...

True but you know that I can't always do that. It all depends on the part being made. I have thought in the past that I could do some math and try to limit the speed. Its very hard to come up with a max speed with all the different options.

When I first added the current XDCR I put the motor amps onto the screen. This morning I had an idea that seems to work well. I put an indicator beside the amp display. As the current goes up I change the colors from no color, to yellow to orange then up to red. This gives them something to look at during the cycle besides numbers. I tell them yellow is ok, orange is pushing it and red is overload. Its ok to see red for a split second but any more than a second's worth of red and you could trip the overload. I think this will help.

I started working on the RMS cals this afternoon. You can see in the cycle that I have steps in the program. I would like to do the RMS based on each step then overall. Then setup some limits to these steps. I should then be able to tell the setup person where in the cycle they need to slow down or adjust.

In a situation like this, where there are so many variables in the process that demand different ablilities, do you actually save anything by using calculations in an effort to use a smaller motor?

I am not sure, again, that I am asking the questions correct.

IF a system has high HP that it only runs every so often then I can see it being done.

BUT a system that has mutiple occasions with different variables that could cause it to run high HP, wouldn't it make more sense to use a motor that is closer to the MAX HP required?

The price different between say 25 and 50 HP is not that great, over 50HP the price range can increase immensely.

This is just my humble opinion but any system under 50HP I would rather oversize or match to high demand then try to save a couple dollars. This should eliminate any issues when someone like me or Beauregard make changes to the system.

This is what our friend "fluidpower" used to preach about, not enough people are properly trained concerning fluidpowered equipment so may make improper adjustments etc. The adjustments may make the system work that time but affect performance at other times.

CharlesM said:

Its really 38.0 amps this is a 25HP motor


Again like I say the cycle changes depending on the part. Depending on the parts we can have both axis moving at the same time with 3000psi on the pump.

Click to expand...

Then this is where the "judgement call" comes in that I was talking about earlier. Clearly there are some points in your process where the HP input requirements are greater than 25HP. (For some reason I was picturing a larger motor system) And from looking at the chart, these time periods can be fairly long and close together. RMS loading of the motor is all about managing energy and heat. The motor cannot be allowed to heat up excessively, so overloads have to be short and you have to have time in between for the motor to cool off - something you don't appear to have. The cost difference from 25 to 30 HP is virtually non-existant - yet for the profile you posted it would mean mean you don't overload anything and have the headroom and flexibility for more agressive profiles. If you are moivng both Axis full speed at 3000 PSI, then from what I see on that graph, a 25HP motor doesn't cut it even if an RMS load calculation says it can - the load duration, % of overload, and frequency is too high. As I already said, a certain amount of judgement is involved here as well. You are the one who best knows your machine and your process. The reasons for doing an RMS loading study is not to undersize your motor, its about not oversizing your motor.

Alaric said:

Then this is where the "judgement call" comes in that I was talking about earlier. Clearly there are some points in your process where the HP input requirements are greater than 25HP. (For some reason I was picturing a larger motor system) And from looking at the chart, these time periods can be fairly long and close together. RMS loading of the motor is all about managing energy and heat. The motor cannot be allowed to heat up excessively, so overloads have to be short and you have to have time in between for the motor to cool off - something you don't appear to have. The cost difference from 25 to 30 HP is virtually non-existant - yet for the profile you posted it would mean mean you don't overload anything and have the headroom and flexibility for more agressive profiles. If you are moivng both Axis full speed at 3000 PSI, then from what I see on that graph, a 25HP motor doesn't cut it even if an RMS load calculation says it can - the load duration, % of overload, and frequency is too high. As I already said, a certain amount of judgement is involved here as well. You are the one who best knows your machine and your process. The reasons for doing an RMS loading study is not to undersize your motor, its about not oversizing your motor.

Click to expand...

That was basically what I was trying to say, since it appeared that the objective in this case was trying to minimize motor size instead of load matching. IF I know that in the future the demands on the system can or will change I would rather oversize on purpose OR load match to the worst case scenario.

As mentioned the dollar difference between certain sizes is nominal, even from 25HP to 50HP it may not be that big of a difference.

I understand completely what you guys are saying. I have been saying the same thing to other people here. With that being said I think you can always say up the HP of the motor. I guess it all comes down to I need to be able to limit the machine more. I know that if I put a 30HP motor on I would have the same problem because they would up the speeds.

Thanks for all the good suggestions. At least I know I am on the right track for the calculations. Like I said before I don't know if I can make the problem go away completly but if I can get it to happen less and less then im improving the product overall.

I still think that motor heating is the real issue here. Why not measure the heat with an analog transmitter, set a load limit and warning signal at 95% of Insulation Class temperature, and trip out the starter at 100%.

That way you can be sure the maximum output is available adjusted for duty cycle and the motor is properly protected if the warnings are ignored.

This would be as close to optimum output that I can picture.

CharlesM said:

I know that if I put a 30HP motor on I would have the same problem because they would up the speeds.

Click to expand...

I don't know anything about your machine, but isn't this generally a good thing?