Jump to content

How does propullsion system work on the NTT's?


NYtransit

Recommended Posts

Im currently taking a electrical class and my professor said we need to explain why something works in a particular way in the electrical field. I gave him 3 topics and he liked the idea of me explaining why the NTTs make the sounds since hes never really met any one who would be willing to do this (Go figure) So I was wondering if some one knows where I can find an in depth explainiation of what happens when your gearing up the train and why the sound comes in, how power is tranferred and what not. Any thing would be helpful, Thank you!

Link to comment
Share on other sites


If I'm remembering this correctly, the NTTs are all using three-phase AC motors instead of DC motors, and they're presumably being driven with sinusoidal or field-oriented control. I'm currently going to school for electrical engineering, and if you're interested in this, I can give you a rundown of how three-phase AC motors work, how they're driven, and how (I think) propulsion and braking work, including why you hear the sounds you do.

Link to comment
Share on other sites

If I'm remembering this correctly, the NTTs are all using three-phase AC motors instead of DC motors, and they're presumably being driven with sinusoidal or field-oriented control. I'm currently going to school for electrical engineering, and if you're interested in this, I can give you a rundown of how three-phase AC motors work, how they're driven, and how (I think) propulsion and braking work, including why you hear the sounds you do.

That would be of great help, Let me know! Thank you

Link to comment
Share on other sites

No problem. Basically, a motor consists of a set of wire coils placed in proximity to an electromagnet; in a brushless inrunner (which is an AC motor where the outside stays still and power is transmitted through the shaft), you have alternating coils of wire wrapped around spindles on the inside of the motor can, and then a set of magnets with alternating polarity mounted to the shaft, as shown in this diagram:

 

Inrunner-Anatomy-SM.png

 

When each pair of windings is energized, the current through the wire produces a magnetic field that pulls the magnets on the rotor toward that winding, causing the shaft to rotate.

 

Now, driving a motor like this is fairly interesting; to achieve clean commutation of the motor you need to be able to turn different windings on and off quite precisely. The goal is to turn each winding completely on right as the rotor passes the previous winding and to turn each winding completely off right as the rotor passes it, thus keeping the rotor stepping around the inside of the motor continuously. If done properly this produces a smooth rotating motion on the shaft.

 

Now, driving a three-phase AC motor from a DC supply (like the 600VDC third rail that the NYC subway uses), requires a separate controller circuit that can switch power through the phases as required. The drive phase of the circuit (that actually handles the power) is comprised of a minimum of six power transistors (either MOSFETs or IGBTs), arranged like this:
 

STDiode1.jpg

Each of the six switches (labeled Q1 through Q6) is some form of semiconductor switch; in the diagram above they're all MOSFETs (metal oxide semiconductor field effect transistors), while in applications with very high voltages and currents (such as subway car propulsion systems) manufacturers will tend to use IGBTs (insulated-gate bipolar transistors). Choosing one over the other is a fairly complex tradeoff based on cost, expected voltages/currents, temperatures, and so on, but as far as I'm aware most heavy (and some light) rail applications use IGBTs.

Link to comment
Share on other sites

Now, when it comes to driving the motor, you typically have three different options; you can choose to use trapezoidal (also called brushless DC) commutation, sinusoidal control, or field-oriented control. Trapezoidal control is the simplest; you just toggle the transistors one at a time, creating a path for current to flow from power to ground through two of the legs of the motor. For instance, you could energize L1 and L2 by turning on Q1 and Q4 (leaving the other four MOSFETs off). From there, you can "step through" all of the possible phase combinations like this:

 

Start with Q1 and Q4 on; L1 is conducting from power to central node (high side) and L2 is conducting from the central node to ground (low side), while L3 is off.

 

Switch Q4 off, Q6 on; L1 is still conducting on the high side, while L2 is off and L3 is conducting on the low side.

 

Switch Q1 off, Q3 on; now L1 is off, L2 is conducting on the high side and L3 is still conducting on the low side.

 

Switch Q6 off, Q2 on; now L1 is conducting on the low side and L2 is conducting on the high side, with L3 off (the inverse of the first state).

 

Switch Q3 off, Q5 on; L1 is still conducting on the low side, but L2 is off and L3 is conducting on the high side.

 

Switch Q2 off, Q4 on; L1 is now off, L2 is now conducting on the low side, and L3 is still conducting on the high side.

 

Finally, switch Q5 off and Q1 on; you're back at the starting state and you've made a full electrical revolution.

 

Why does this work?

 

Let's take the simplest case (a motor with three windings spaced 120 degrees apart and a bar magnet in the center). Each winding can be in three states; current flowing into the central node (high side conduction), current flowing out of the central node (low side conduction), or no current flowing at all (off). Assuming that you must have one winding in each of the three states at any given time (one on the high side, one on the low side, and one off) to generate torque, then the motor itself can occupy any one of the six states above. Each state corresponds to a unique torque vector, and the spacing of the windings means the vectors must also be evenly spaced. This then creates a set of six torque vectors, spaced sixty degrees apart, that the stator can pull on the rotor with. If you energize the states in the proper order, then you create a rotating torque vector that will pull the rotor around continuously.

 

This form of commutation is the simplest; you can implement the control electronics with a few $1 chips and three $0.50-$1 Hall effect sensors to tell you where the rotor is at any given time; it's also fairly efficient at high speeds. However, at low speeds you start running into serious problems with torque ripple; each time there's a state change the rotor goes from feeling almost no torque (because it's right on top of the energized winding) to feeling the full torque exerted by the next winding as it comes online. Instead of exerting a steady torque and moving at a roughly constant angular velocity, the rotor is effectively flinging itself from one torque vector to the next. At high rpm you don't feel this much at all because the torque pulses are coming on and off too fast for the physical system to react to each one individually (and so all you see is the average torque, which is fairly steady). At low rpm (especially on a small vehicle like an electric scooter) you can feel the torque ripple quite well, and it can be quite disruptive.

Link to comment
Share on other sites

The sound comes from various sources including the motor and the gearbox.  Very rarely you have a direct traction motor to an axle as the speed is too great and initial torque is too low..  What is done in most locomotives or traincars is we gear it for what we are using it for similar to a rear axle in a car where the higher number rear gives power and a lower one can give speed.  Remember you are starting a 35 ton railcar  and direct drive puts current at a high amount but also produces heat in the windings at excessive amounts. which shortens the life of the motor.  The old BMT cars were geared for 42 mph on the standards and 45 on the Triplexes.  I have no idea what they are geared for now but it has to close to those numbers or otherwise it would come off the track on turns.as the system originally was designed that way..  You also have air compressors for the brakes which are audible when running.  Noise is also generated in the wheel design  I can go on and on.  Best way is go to a trainstation with a good recorder and record the sounds as it comes in, stops and starts again.  Stops gives you the ambient sounds which always occurs, slowing down gives the braking and start up gives the sounds of the motor and gearbox.  There use to be on Cnet free programs you could download to see the sound in waveform and clean out the ambient sound to hear only the motor sounds.

Link to comment
Share on other sites

The sound comes from various sources including the motor and the gearbox.  Very rarely you have a direct traction motor to an axle as the speed is too great and initial torque is too low..  What is done in most locomotives or traincars is we gear it for what we are using it for similar to a rear axle in a car where the higher number rear gives power and a lower one can give speed.  Remember you are starting a 35 ton railcar  and direct drive puts current at a high amount but also produces heat in the windings at excessive amounts. which shortens the life of the motor.  The old BMT cars were geared for 42 mph on the standards and 45 on the Triplexes.  I have no idea what they are geared for now but it has to close to those numbers or otherwise it would come off the track on turns.as the system originally was designed that way..  You also have air compressors for the brakes which are audible when running.  Noise is also generated in the wheel design  I can go on and on.  Best way is go to a trainstation with a good recorder and record the sounds as it comes in, stops and starts again.  Stops gives you the ambient sounds which always occurs, slowing down gives the braking and start up gives the sounds of the motor and gearbox.  There use to be on Cnet free programs you could download to see the sound in waveform and clean out the ambient sound to hear only the motor sounds.

The frequency display programs are pretty good; I ran one on several videos of R160B Siemens trains pulling in and out, and the initial acceleration sound is a linear increase from 800 Hz to about 1500Hz, which fits with the sound of several AC motors under sinusoidal or field-oriented control being spooled up to a set speed.

Link to comment
Share on other sites

The frequency display programs are pretty good; I ran one on several videos of R160B Siemens trains pulling in and out, and the initial acceleration sound is a linear increase from 800 Hz to about 1500Hz, which fits with the sound of several AC motors under sinusoidal or field-oriented control being spooled up to a set speed.

 

The sound of the Siemens motors/inverters makes sense to me - but it's the multi-pitch melody that the Alstom ONIX traction that I have a hard time wrapping my head around. My assumption had always been that rather than a slow steady increase in AC frequency, they had found a way to solve low-rpm torque problems by hitting different frequencies in sequence and this would line the rotors up with the coils properly at different RPMS. 

 

I'm not explaining this very well. In other words the frequency of the AC doesn't have to match the RPM it just has to sort of "harmonize" with it. 

 

Anyway - obviously I have no idea what I'm talking about but if any of you fellas know, I'm curious!

Link to comment
Share on other sites

The sound of the Siemens motors/inverters makes sense to me - but it's the multi-pitch melody that the Alstom ONIX traction that I have a hard time wrapping my head around. My assumption had always been that rather than a slow steady increase in AC frequency, they had found a way to solve low-rpm torque problems by hitting different frequencies in sequence and this would line the rotors up with the coils properly at different RPMS. 

 

I'm not explaining this very well. In other words the frequency of the AC doesn't have to match the RPM it just has to sort of "harmonize" with it. 

 

Anyway - obviously I have no idea what I'm talking about but if any of you fellas know, I'm curious!

 

That's most likely the case; ONIX is most likely an asynchronous AC system; the frequency of the AC doesn't have to match the mechanical frequency of the rotor and in fact will naturally be faster. In a synchronous system, the magnetic field created by the current rotates at a given speed (say 60 Hz for a household appliance running on standard 120VAC home power); the rotor will try to match that speed (3600 RPM) but be unable to when it's loaded down. The difference between the angular velocity of the rotor (and its magnetic field) and the angular velocity of the magnetic field from the motor windings leads the two fields to interact in a way that produces torque.

 

It sounds like ONIX is a notched asynchronous system, and thus each notch corresponds to a particular angular velocity that the traction motors will try to (but not quite) match. The reason for a notched setup is twofold; first, it allows the train to hit a number of speeds (each notch on a synchronous system corresponds to a single steady-state speed), and second it makes the system's current demands much more reasonable. Since slip is proportional to torque and torque is proportional to current, trying to start up an asynchronous system at a frequency on par with ~40-50mph operation would demand an insane amount of current and could quite possibly damage the power electronics inside. On the other hand, stepping up to 5mph, then 10, then 15, etc. is a lot more reasonable.

 

By contrast, it would seem that Bombardier (MITRAC, on the R142A, R143, and R188) and Siemens (SITRAC, on the R160Bs) are synchronous systems.

Link to comment
Share on other sites

The frequency display programs are pretty good; I ran one on several videos of R160B Siemens trains pulling in and out, and the initial acceleration sound is a linear increase from 800 Hz to about 1500Hz, which fits with the sound of several AC motors under sinusoidal or field-oriented control being spooled up to a set speed.

Could you explain the increase for the MITRAC?(R142A,R143,R188).Your explanations are fascinating.Thank you for sharing them.

Link to comment
Share on other sites

MITRAC sounds interesting; there's an initial 1100-1200 Hz tone, followed by a fairly smooth ramp from 2-300Hz back up to around 700Hz and then another ramp from about 2-300Hz up as high as 12-1300Hz if the T/O is pushing the equipment fairly hard. If I had to guess I would say that it's a two-speed system with some sort of setup tone being played through the controller at the start (although there's no evidence in the MITRAC 1000 brochure that the system uses a two-speed transmission) I'm honestly not sure why MITRAC emits that particular sound pattern. 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.