AC also alternates polarity.

http://www.imperialelectric.com/product ... motors.php
Most seem to be made to run at 240v but they do state they can make other custom voltage motors.

Remember that part about NEMA large-frame three-phase motors? Those require F to run in the first place, so you can figure something has to be done with the DC to allow this.

The VFD actually synthesizes the "F" and yes, it can do that from DC ... you may recall that most diesel-electrics with AC drives actually first rectify and filter the output of the traction alternator to a good approximation of DC (this is called the "DC-link" in some descriptions) and then invert this to get the phase, amplitude, and waveform going to the traction motors. It is relatively easy to adapt the DC-link connection so its input is 600VDC with appropriate voltage clamping or conversion; in fact I have seen proposals to do this for dual-power third-rail locomotives with voltage-to-voltage conversion from 750V or whatever. (And yes, they remember the data from 1915 about rapid changes in the DC reference voltage when other trains cause or shed substantial load on the line...)

The electronic control approach for DC motors would more likely be a 'chopper' of some sort, which acts like a very rapid switch cutting pulses at a fairly high frequency and variable 'off-width out of the DC fed to a traction motor. The output when filtered approximates a 'resultant' lower voltage and not quick cycling of power with associated reactance effects through components generating magnetic fields. This method whether involving frequency or pulse-width modulation (PWM) is a common way to control multiple-horsepower permanent-magnet DC motors, as well as being familiar on a number of transit systems for a half-century or so.

I find it hard to believe that someone couldn't figure out how to circulate coolant through a VFD in an appropriately-sealed enclosure for whatever degree of component cooling (no pun intended) would be involved, even at high continuous demanded horsepower and improperly adjusted power factor. That sort of design hasn't been rocket science for quite some time; neither have heat pipes for designers who don't want to have to use pumps or circulation.

Come to Dallas and see for yourself!

I'm not electrically inclined, so I don't know if this will make a difference, but the VFD was to control AC motors. It was fed 600v DC to an unknown (to me) output of AC. I can't tell you more than that.

The motors MATA put on 754 are AC.

Let me take a stab at this... since I am NOT an electrical engineer, maybe my explanation will make sense to the layman.

If I were going this route of solid state control, I would WANT to use 3 phase AC motors, If I could satisfy myself that the bearings were up to the service. Why? Three phase motors have no brushes, therefore essentially never need service. In addition, AC induction motors are way more common than DC motors, giving a larger selection to choose from, and hopefully find surplus at reduced prices.

One thing to consider is AC induction motors are usually designed as constant speed machines. That means the internal fan (or external fan on TEFC, Totally Enclosed Fan Cooled motors) is not going to do the job whn the motor is running at reduced speed. I think I'd want to design around TENV (Totally Enclosed No Ventilation) motors. These have big heavy cases and are designed to cool strictly through convection.

Now to the VFD (Variable Frequency Drive). The way these normally work is they take in 3 phase power and rectify it, yielding DC. The "DC bus" voltage of a 480VAC drive is 660VDC, because the 480 is the RMS voltage of the AC line (Root Mean Squared, in layman's terms an average of the value of the sine wave over time). When you rectify AC, you see a voltage that is related to the peak voltage, appx. 660. The drive uses big capacitors to eliminate the remaining ripple on the DC bus. Since some industrial applications use a separate DC power supply for multiple VFD's, it is not unheard of to order large VFD's without the rectifier/capacitor front end, since we already have 600VDC present.

The logic of the VFD then uses large IGBTs, transistors, to build a sine wave on each of the three output lines through a process similar to pulse width modulation. The motor sees these series of pulses as the rising and falling voltage of the normal AC sine wave, BUT, we can easily control the frequency of this pseudo sine wave. VFD's can typically run an induction motor from 10% to 200% of rated speed.

Don't know what went wrong with the McKinny Ave. experiment, possibly the VFD was undersize for the application. I agree that force ventilating a closed case with enough airflow (from a BIG external blower) should have done the trick.

Of course, then your reproduction streetcar sounds like one of the New Haven "jets".

In my plant we have many old school open ventilated 3 phase AC motors running retrofitted with VFD control at lower than spec speeds with no overheating issues - and they run constantly in pumping and air handling duty, not intermittently with off time to cool down like streetcar motors would. Which does also open up the idea of AC powered air compressors available off the shelf...... and retrofitting air conditioning to the cars used in transit service like McKinney Avenue's are.

Dave,
The biggest problem with what you just said is that with a VFD drive, the motors would not see intermittent duty. Since, unlike a K controller, you can power the motors continuously at any speed (and the speed control could just be a rheostat knob, like a train set) the operators will quickly forget the proper way to run a streetcar, and just "dial in" the speed they need, much like driving an automobile. The motor fans need to run a reasonable speed to cool the motors; I doubt your air handlers spend much time running at 10% of rated speed, but I could see a streetcar doing so.

The VFD motors I have operated had a 20% minimum speed for fan cooling. 30 years ago it was 40 to 50% min. If you buy a new motor, get one that is VFD rated with better insolation. I would look into AC diesel locomotive technology to get some ideas. They can hold a train stopped on a grade while the train line is pumped up. It might be overkill for a streetcar application. VFD fans makes an awful lot of noise as well as do the motors.

Tom Hamilton

So I think the question is: to make "new" traction motors....to what end?

Is it just "make it go" or an authentic museum restoration?

if the former, it is probablly easier in this day and age to use off-the-shelf 3-phase
induction motors with VFD, derate sufficiently for extended low-speed operation,
high ambient and restricted ventiliation. Figure out how to mount the motor and
couple it to the axle.

Or attempt to use a modern AC traction motor meant for LRV duty, if it fits.

To actually manufacture a new WH506 or GE264 to function as a DC traction motor,
with axle bearings and caps, pinion and bull gears and all the trimmings, I would think
is going to be at least $200K per motor in small quantities. But the other choices won't
be cheap either.

If a museum back shop has sufficient capabilities including access to a foundry that can
cast steel, and a good machine shop, it is certainly theoretically possible to make the DC
motor from scratch. There is really no "unobtainium" in the motor.

One also has to be clear about the type of service it is going to see. Will it make a trip or
two once in a while in a light-duty museum sertting, or is it going to be running for many hard
hours at a time in heritage service, like McKinney?

I don't think the cost difference between a replica for occasional duty and for intensive duty would be more than marginal - so gear up for the more robust. If it lasts forever in a museum, it is worth the small additional investment.

On another thread about controllers somebody or two posted an opinion that new replica controllers could be replicated inexpensively - I have looked into that and found that (apart from perhaps different ideas about what defines "expensive") it would not be inexpensive. So, I challenge them to demonstrate what the real cost would be for , say, a new K35 with specifications and quotes.