VFDs on the load motor side

[garciaj]

See attachment for details.

Background MTE products using a variaty of VFDs are currently tested in our system at only 60Hz at full load. How we currently do it Starting the drive @ 60Hz will spin the drive motor at 60Hz, then the softstarter is used to bring the load motor @ 60Hz The slip of the system is used, and small increments in the drive motor above 60Hz (less than 62Hz usually) are dialed to load our filters at the desired load. Purpose To be able to test MTE products at different frequencies (10Hz > 60Hz) at full load using any VFDs on the output of the system Can Variable Frequency Drives be connected on the output of our system? To drive the load motor? Mots likely the softstarter needs to be removed since it is rated at 60Hz With a VFD for load and Motor I will need to handle the regen energy somehow, any ideas? Which other options can be made? Alternatives 1. Purchasing a dynamometer for each system? 2. Purchasing a single dynamometer that will accommodate all systems? Note: A dynamometer can be very costly, MTE is looking for different solutions including using VFDs, other types of loads, etc

 

[Jraef]

Typically in a simple dynomometer system like you show there, the load motor is run with a 4 quadrant DC drive and motor system, because it's a cheap way to get the line regen. There are line regen AC drives now, but they are a lot more expensive. Their advantage is not having a DC motor and the associated maintenance issues for it, but as a dyno tester, the maintenance of a DC motor is not going to be that big of a deal and the cost difference is substantial.

That said, you ALREADY have the AC motor for the load side, so after the fact, buying a new motor + the DC drive might be more than buying just an AC Line Regen drive and using the motor you already have.

But the other question is, why? Your filters are going to be the most stressed at full load, which in a VFD means full speed. I'm not convinced on the real need to test them at lower frequencies / loads.

 

[mike_kilroy]

Why regeneration at all? Replace soft start with identical vfd and tie dc bus together. Then you regeneration back to driving vfd to power it. Only need to bring in a few amps from source to run 100-300 amp motors.

Sent from my SM-G900V using Tapatalk

 

[Ingenieur]

Why regeneration at all? Replace soft start with identical vfd and tie dc bus together. Then you regeneration back to driving vfd to power it. Only need to bring in a few amps from source to run 100-300 amp motors.

Sent from my SM-G900V using Tapatalk

is he really producing much power?
the way I understand it is the load motor is producing negative torque (in oppostion to the drive motor) soley to impose a load, ie, acting with braking action
so both require power to produce torque?

 

[mike_kilroy]

Op said "full load" and listed upto 450hp motors.... if his drive motor outputs 450hp, then his load motor is generating 450hp too. Only source power he needs to add from poco is to make up for the inefficiencies. We have built test stands this way for people for years.

Sent from my SM-G900V using Tapatalk

 

[mike_kilroy]

Why regeneration at all? Replace soft start with identical vfd and tie dc bus together. Then you regeneration back to driving vfd to power it. Only need to bring in a few amps from source to run 100-300 amp motors.

Sent from my SM-G900V using Tapatalk

OOPS... Jraef pointed out he is testing his filters; only the sine filter would see the real currents for a real test...

The input noise filter would only see about 5% of the motor current so the test may be tainted - until they tested the concept and compared tying the dc bus vs not. Since the input noise filter's job is to reduce the PWM noise going back out to the line from the vfd, it MAY still provide a valid test of those noise filters even though the incoming current is only 5% of the total... Probably worth testing because it would reduce their power and complexity enormously...

Easy enough to test since they no doubt have many vfds to cover the 5-450hp range...

 

[Besoeker]

Typically in a simple dynomometer system like you show there, the load motor is run with a 4 quadrant DC drive and motor system, because it's a cheap way to get the line regen. There are line regen AC drives now, but they are a lot more expensive. Their advantage is not having a DC motor and the associated maintenance issues for it, but as a dyno tester, the maintenance of a DC motor is not going to be that big of a deal and the cost difference is substantial.


That said, you ALREADY have the AC motor for the load side, so after the fact, buying a new motor + the DC drive might be more than buying just an AC Line Regen drive and using the motor you already have.

But the other question is, why? Your filters are going to be the most stressed at full load, which in a VFD means full speed. I'm not convinced on the real need to test them at lower frequencies / loads.

Agree with that.

One of the things that struck me is the range of filters to be tested - from 5HP to 450HP. I don't think a single test rig could accommodate that.

 

[Besoeker]

is he really producing much power?
the way I understand it is the load motor is producing negative torque (in oppostion to the drive motor) soley to impose a load, ie, acting with braking action
so both require power to produce torque?

The idea of this sort of dynamic testing is to consume just losses. Keeps the costs of the energy lower even though testing kit rated at quite high powers.

 

[Ingenieur]

Op said "full load" and listed upto 450hp motors.... if his drive motor outputs 450hp, then his load motor is generating 450hp too. Only source power he needs to add from poco is to make up for the inefficiencies. We have built test stands this way for people for years.

Sent from my SM-G900V using Tapatalk

Let me get this straight
drive motor is producing 450 HP
load motor is absorbing the 450 and regenerating 450 (less system losses) back into to the bus
and the vfd only requires 10% of 450 ~ 45 hp to sustain the system?
assuming resistance, frictional, etc losses are ~ 10% or 45 hp

 

[Jraef]

Let me get this straight
drive motor is producing 450 HP
load motor is absorbing the 450 and regenerating 450 (less losses) back into to the bus
and the vfd only requires 10% of 450 ~ 55 hp to sustain the system?
assuming resistance, frictional, etc losses are ~ 10% or 45 hp

For the VFD connected to the load side motor to function, it has to be connected to the line because it is going to supply the flux current to the load motor, then it would be given a speed command that is slightly below the speed the load motor motor is being driven by the line side motor / VFD. The load side motor cannot power up the load side VFD (unless there is some significant residual magnetism, but that can't be counted on). So you will still have the line losses of both VFDs, assuming 3% each, then the losses in those filters and of course both motors. I think 10% total is likely a low estimate, unless that was not counting the losses in the motors, because those exist no matter what.

That makes me think though; now with PMAC motors, I believe that COULD actually be done without any line connection of the load side VFD. In other words the load side motor, as a PMAC motor with Internal or Surface Mount Permanent magnets, COULD in fact act as a generator from the outset, so the VFD connected to it could be powered BY it, then the VFD would give it the lower speed command to make it a load. All you would have would be the losses in the line side VFD then. Kind of a moot point in this case though, nobody I know of is making a 450HP PMAC motor yet.

 

[mike_kilroy]

Close. Not vfd. Ok, maybe you mean Both VFDs are putting out or taking in the full 450hp less efficiency. The Poco, Aka power source will be pulling in the 45hp of losses to the motoring vfd. So guess yes, you got it.

Sent from my SM-G900V using Tapatalk

 

[Ingenieur]

For the VFD connected to the load side motor to function, it has to be connected to the line because it is going to supply the flux current to the load motor, then it would be given a speed command that is slightly below the speed the load motor motor is being driven by the line side motor / VFD. The load side motor cannot power up the load side VFD (unless there is some significant residual magnetism, but that can't be counted on). So you will still have the line losses of both VFDs, assuming 3% each, then the losses in those filters and of course both motors. I think 10% total is likely a low estimate, unless that was not counting the losses in the motors, because those exist no matter what.

That makes me think though; now with PMAC motors, I believe that COULD actually be done without any line connection of the load side VFD. In other words the load side motor, as a PMAC motor with Internal or Surface Mount Permanent magnets, COULD in fact act as a generator from the outset, so the VFD connected to it could be powered BY it, then the VFD would give it the lower speed command to make it a load. All you would have would be the losses in the line side VFD then. Kind of a moot point in this case though, nobody I know of is making a 450HP PMAC motor yet.

Many wind turbines use a pmig
in the 1-3 MW range
so I'm sure a pm 0.5 motor could be made, by why?

I am not convinced (assuming even 20% losses)
the the drive motor makes 450 HP
the load absorbs/regens 450 less losses
and only 90 hp is required from the mains

 

[mike_kilroy]

My reply before jraef, so sounds weird... jraef, the VFDs we use for dyno test stands like this have their logic supplied off dc-dc converter off dc bus like most vfds. So we do NOT hook the load side vfd to the line. In fact we sell one ac input vfd (converter) and one dc only (inverter) vfd. Save the cost of the ac input circuitry on the load vfd. I think the 10% total losses is actually higher than what we normally see, so using 10% seems very reasonable to me. Of course different VFDs may require different schemes as you showed.


Sent from my SM-G900V using Tapatalk

 

[Ingenieur]

Close. Not vfd. Ok, maybe you mean Both VFDs are putting out or taking in the full 450hp less efficiency. The Poco, Aka power source will be pulling in the 45hp of losses to the motoring vfd. So guess yes, you got it.

Sent from my SM-G900V using Tapatalk

I don't get it
I do not believe the extrrnal power supplied to closed system is 45, 90, whatever the losses are

I believe BOTH are acting as motors
not a motor generator set

 

[Jraef]

My reply before jraef, so sounds weird... jraef, the VFDs we use for dyno test stands like this have their logic supplied off dc-dc converter off dc bus like most vfds. So we do NOT hook the load side vfd to the line. In fact we sell one ac input vfd (converter) and one dc only (inverter) vfd. Save the cost of the ac input circuitry on the load vfd. I think the 10% total losses is actually higher than what we normally see, so using 10% seems very reasonable to me. Of course different VFDs may require different schemes as you showed.


Sent from my SM-G900V using Tapatalk

Oh, right, you are tying the DC busses together. Duh, you even said that...

 

[mike_kilroy]

Many wind turbines use a pmig
in the 1-3 MW range
so I'm sure a pm 0.5 motor could be made, by why?

I am not convinced (assuming even 20% losses)
the the drive motor makes 450 HP
the load absorbs/regens 450 less losses
and only 90 hp is required from the mains

OK. Just for fun, here is a scope pix of a dual motor 'dyno' test we did at Nasa in CA on a proposed motor for the space shuttle hydraulics:

The purple is RPM, the RED is actual current in the driven motor and the blue is the actual current in the load (generating) motor. We used fiber optic link between the two 200kw water cooled drives to send instantaneous parameters back and forth. We dialed in a small offset in current (or slip or speed however you want to say it) to make them fight each other at the load we desired.

The only power we drew from the wall was the difference between the two currents.

One motor makes the HP the other absorbs it exactly like a windmill and exactly like a windmill feeds it back to the common bus. Only losses supplied from the wall.

 

[Besoeker]

OK. Just for fun, here is a scope pix of a dual motor 'dyno' test we did at Nasa in CA on a proposed motor for the space shuttle hydraulics:

The purple is RPM, the RED is actual current in the driven motor and the blue is the actual current in the load (generating) motor. We used fiber optic link between the two 200kw water cooled drives to send instantaneous parameters back and forth. We dialed in a small offset in current (or slip or speed however you want to say it) to make them fight each other at the load we desired.

The only power we drew from the wall was the difference between the two currents.

One motor makes the HP the other absorbs it exactly like a windmill and exactly like a windmill feeds it back to the common bus. Only losses supplied from the wall.

Nicely demonstrated if I may say so.

 

[mike_kilroy]

Thank you.

 

[Ingenieur]

OK. Just for fun, here is a scope pix of a dual motor 'dyno' test we did at Nasa in CA on a proposed motor for the space shuttle hydraulics:

The purple is RPM, the RED is actual current in the driven motor and the blue is the actual current in the load (generating) motor. We used fiber optic link between the two 200kw water cooled drives to send instantaneous parameters back and forth. We dialed in a small offset in current (or slip or speed however you want to say it) to make them fight each other at the load we desired.

The only power we drew from the wall was the difference between the two currents.

One motor makes the HP the other absorbs it exactly like a windmill and exactly like a windmill feeds it back to the common bus. Only losses supplied from the wall.

not sure what trace is what
if it were an externally excited synchronous machine I would agree
not 2 induction machines