Can I run two motors off one VFD?

I have a Boyar Schultz grinder I am working on. It has a 3/4 Hp, 3450 RPM spindle motor and a 1/2 Hp 1725 RPM hydraulic pump. Can I run both off a 2 Hp VFD, or will they require 2 different VFDs.

Thanks

They'll always run at the same frequency, you can't start and stop them separately, and you'll need a pair of motor circuit breakers if you want to protect them individually from overcurrent/overload.


Whether that's OK is up to you.

Hardly seems worth the compromises and effort of running both motors off one VFD given the low prices of basic 3/4 and 1/2 hp VFD boxes these days.
If you don't plan to vary the speed the simple ones made to replace conventional contactor controls, such as the Eaton DE1 series, would seem the way to go. Compact, no display or controls on the box. You basically just switch them on and off, speed pot can be added if you want.

Running wo motors of a V/F type VFD is fine, given sufficient overload drive capability you can switch on and off individually too but the drive starts getting silly large. Vector drive types will probably be less than happy. Self tuning ones won't be able to tune and may well not work at all if the internal safety routines detect something thye consider a fault.

Clive

Yes, pretty much all drives need to be run in scalar (V/f) mode, so you don't have torque control and the like.

For the price of the 2 Hp VFD, you could buy a 1 Hp and 1/2 Hp VFD and put them both in the same enclosure or base. I assume you do not need speed control, you might look at the KBVF or KBDF for simple VFD drives. Tank type coolant pumps, depending on the type can often be easily switched out to single phase. This would be a much better solution at a nominal cost and protect both motors.

Thanks everyone for the advice. Obviously my original idea was not good. I will look up two simple VFDs since I don't need speed control for either motor.

bll230 said:

Thanks everyone for the advice. Obviously my original idea was not good. I will look up two simple VFDs since I don't need speed control for either motor.

Click to expand...

I think you could use a drum switch to select what motor to power using 1 VFD.
If you had a 3p drum switch laying around and didn’t mind only running 1 machine at a time then I think you’ll be fine.

If you had something like a coolant pump, much smaller than the spindle, then you could just switch it on and off, without bothering the VFD.

Those motors are too close in power to get away with it on 2 HP, but something in the 5 to 7.5 HP range would pretty much let you do whatever. Switch on/ off, etc.

That machine is what RPCS are for.... you don't want anything from the VFDs other than 3 phase, which is what an RPC does. A motor in the 3 to 5 HP range for an idler, and you would be doing fine. a 2HP would be pushing it.

The manual for Invertek VFDs says that an output choke is necessary when two motors are run from one inverter. They don't explain why.

Bill

9100 said:

The manual for Invertek VFDs says that an output choke is necessary when two motors are run from one inverter. They don't explain why.

Bill

Click to expand...

Same situation appeared in lots of earlier railroad traction inverters, when one inverter is running two traction motors (one on each axle of a bogey).

reason is mostly... that the inverter is measuring the current flow out to the load, and trying to adjust it's pulse train to best meet the propulsion demand... but having two motors means you have two different slip circumstances going on, and that means the load signal at the inverter's sensing point is really goofy. By adding the output choke, the inequities of the two motors is settled beween THEM, and the load seen by the inverter is 'less wierd'.

From a simple perspective, one would think that two axles on the same bogey of a locomotive would always turn the same speed, and 'find' a way to be in-phase... but they're usually not. Even with the wheel diameters cut just perfect, wheelslip occurs as a result of wheel/rail interface variations (improper profile cut, road damage, tagging a flange in a curve, etc).

The modern solution, since inverters are much less expensive than they used to be, and much smaller too... is just to dedicate one inverter for each traction motor. ;-)

Ot but I had a railway literally in my backyard (city backyard, like 100ft!) and I remember vividly the tops of the rails wer “smashed flat” and created a jagged sharp edge on both sides

Man was those tracks dilapidated!

DaveKamp said:

Same situation appeared in lots of earlier railroad traction inverters, when one inverter is running two traction motors (one on each axle of a bogey).

reason is mostly... that the inverter is measuring the current flow out to the load, and trying to adjust it's pulse train to best meet the propulsion demand... but having two motors means you have two different slip circumstances going on, and that means the load signal at the inverter's sensing point is really goofy. By adding the output choke, the inequities of the two motors is settled beween THEM, and the load seen by the inverter is 'less wierd'.

From a simple perspective, one would think that two axles on the same bogey of a locomotive would always turn the same speed, and 'find' a way to be in-phase... but they're usually not. Even with the wheel diameters cut just perfect, wheelslip occurs as a result of wheel/rail interface variations (improper profile cut, road damage, tagging a flange in a curve, etc).

The modern solution, since inverters are much less expensive than they used to be, and much smaller too... is just to dedicate one inverter for each traction motor. ;-)

Click to expand...

not a good enough explanation at all in my opinion.

in my opinion the only reason this could be a problem is that the current measurement in the vfd is not sufficiently averaged and the pwm frequency combined with the continuously oscillating motor inductance.. (due to the stator-rotor slot combination) creates a beat frequency that confuses the vfd.


last paper i read a while ago, explained why it takes 30 mips to calculate true rpm of an induction motor from the vfd's measurements of volts and amps.. utilizing the rotor-bar combination (typically 36 slots and 34 rotor slots) to measure the slip in real time.

as i understand it from the very beginning of vfds becoming cost effective for trains.. engineers were trying to use separate drives for each axle to push the traction limit as high as they could get.

if someone thought they could save money by paralleling motors... then this was an experiment i've not read about, and its no surprise to me they ran into problems especially if those vfds were trying to calculate the slip by reading the beat frequency of the stator-rotor bar combination.

johansen said:

not a good enough explanation at all in my opinion.

Click to expand...

My apologies- I was attempting to keep it at a simple level. Your description is correct, and you are correct that the result of paralleled motors was substantial enough to abandon the concept. It was attempted, though, because at that time, the componentry and technology of the inverter was not nearly 'small' enough in space and cost, to be capable of a one-inverter-per-traction-motor scheme. Realize that the inverters were running under basically 'analog' control... extremely little 'processing' capability. They were proven to 'work', and they performed very well under 'ideal' circumstances, but they didn'perform well under the common (marginalized) environment. . It was a clear case that there was performance enhancement potential, as long as the details could be worked out.

The 'cost effective' intention WAS to produce a much higher effective tractive effort... and this was by virtue of the inverter's ability to react to loss of adhesion MUCH faster than a gang'd DC drive. Rather than adhesion lost wildly, and an operator backing off the TE several notches to recover (but only after noticing sparks shooting out the sides of the wheels), the system sensed it within a half-rotation of the wheels, and dumped drive... and then reapplied it to some point LESS than it saw the loss. Basically,it was working on the same principle used by anti-lock braking of an automobile today. Early systems used a dedicated speed sensor (my earliest units used a small doppler radar under the belly, pointed at the rails), while later systems used a numeric algamation of wheel speed sensors in the gearboxes,and on the motor shafts, to look for wheel speed changes in excess of the 'average' of all other axles AND... an algorithm which identified the maximum change in velocity of the locomotive under 'best case' circumstances. If any ONE motor/axle identified a speed change FASTER than the entire vehicle's predetermined performance envelope could support, the axle was obviously spinning (or under braking, sliding).

But yes, the fact that the two motors were effectively running at different rotating frequencies meant that there was a definite amount of 'electrical slip' occurring between the incoming AC, and the rotor-to-polepiece relationship. The faster-spinning motor was drawing less current than the slower, and this resulted in a very 'lumpy' phase current and voltage signal on the VFD's output. The output reactor was a feeble attempt to quell this. As a technically-savvy electronics/electromagnetics tech would presume, the presence of the output reactor did help reduce high-frequency harmonic content, it did little to help solve the genuine problem.

Technology improved, granting inverters in smaller packages, at less princely costs, and the result was individually powered (and intimately controlled) axles.

Those issues would appear with more advanced algorithms. A straight V/Hz output would almost certainly run multiple motors. I have seen Invertek units do it just fine, in V/Hz.

How the software would figure out what is going on with a hash of both motor's signals coming back is a real puzzler, and probably there is no good way of doing it with present devices.

Theoretically, one could detect the apparent existence of two speed signals, and try to treat them separately. I think there might be a way to do that, but I cannot see there being a business case for spending the time to try. Just are not enough cases where that is needed, and once you had everything separated, you only have one output to control, so what you cannot do is to send each motor a separate drive. Inevitably what is sent to one will affect the other.

much better explanation, but doesn't explain why a modern inverter is recommending a reactor to drive two motors in parallel. i tend to think their recommendation for that has to do with the transmission line, and the motors creating a resonant circuit between the two motors which would amplify existing well known problems.

in the example of one inverter driving several paralleled motors each driving their own locomotive axle, if one motor slips then its amps drop and so does its power factor.. but not by much since i'm assuming the slope of the curve of the rail:slip vs torque curve isn't steep, and you only have a maximum velocity difference of 3% for a normally loaded motor.. maybe 10% for a traction motor.

if the slop of the slip vs torque curve is pretty steep and non linear, then i can understand that several inverters each driving several axles are going to fight one another and result in a lot of interesting confusion between several locomotives. seems similar to the matter of "traffic waves" when human drivers reach a limit of how far they can look ahead to integrate the velocities of the 2-5 vehicles in front of them. this matter has nothing to do with paralleling motors but rather the algorithms built into the inverter to manage rail slip.

for what its worth i got rear ended today on the highway. i had to slow down to about 40mph suddenly due to a sudden slow down in inclement weather, i considered changing lanes to give the guy behind me more room but thought it was too dangerous, guy behind me hit me at probably 47-50. i tried to give him as much time as i possibly could, and i did not hit the guy in front of me as a result of the impact.. which was also sort of calculated. as for my truck, no impact no idea lol, and i didn't notice any visible damage to his mid 2000's era.. either bmw or mercedes i think.. memory already fading lol. he didn't stop but rather passed me and ran off.