Railbus Controller
Nelson Riedel, Nelson@NelsonsLocomotive.com
3/20/2017, last updated
As discussed in the overview, the old controller went up in smoke last fall,  RIP.   I couldn't get the interest up over the winter to rebuild it.  Last fall I saw the controller that Laura and Dominik of Integral Circuits made for the small Critter Locomotives  Dick McCloy and Mike Price were making.   It left a good impression, especially the physical packaging.  I looked up the specs on their website (www.integralcircuits.com -the specific controller is Model CC 50A).  It does 50 amps continuous for 1080 watts at 24 volts and my motors are 500 watts each.   Called Laura and she said she could send me one in a week and it arrived in a week.   Rather than describing the system I just copied stuff from their website: 

There is a neat wiring diagram on the Integral Circuits website (link to diagram) that shows all the connections including the optional accessories.  To simplify this description  I removed all the wiring for the accessories in the drawing above and kept only what is necessary to get it up and running.  I used a 90 amp 28 volt breaker instead of the main fuse.  I used a 70 amp 28 volt breaker in the positive lead to both motors instead of the individual 30 amp fuses.    The main breaker is also the main power switch.  The motor breaker is normally left on but is useful to disconnect the motors when towing.   

The controller arrived before I finished making a number of repairs and the drive unit was still off the frame.   Great, I could install the speed sensor as described on page 26 of the manual and repeated on the right.  The sensor is a magnetic reed switch that is encapsulated in a ~3/8" diameter plastic bolt and held in position by a pair of plastic nuts.   The magnet is 5/16" diameter and about 3/16" thick and really strong.  The end of the sensor is positioned about 1/8" from the magnet.

I would have preferred to mount the sensor above the motor mount platform but there is no room so I had to put it underneath as shown in photo.    A concern was that I'd wipe off the sensor the first time I derailed so I made the pictured mount from a piece of easy strut and 1/2" angle.   The wire feeds up through a hole in the mount platform and is bundled with the motor wires to get to the controller.   The controller was later set up for the correct wheel size.     

This shows the controller installed on the plastic drive cover.   The controller is attached to a 1/2" aluminum angle with Nylock nuts on two of the studs protruding from the case.  The 24 volt power enters via the terminal block on the left.   The motor and speed sensor wires come up through the channel behind the mount cover.  The motor wires connect to the terminal block on the right and the speed sensor  cable plugs into the controller (small red & white wires).   That is the motor circuit breaker mounted on the back of the drive cover. The heavy (I used 12 gauge stranded) wires connect to the controller via provided barrel connectors soldered to the ends of the wire.  These connectors are really neat.   I questioned the reliability but was assured that they are reliable --- and quite expensive.  After everything was connected I found the motors ran backward  so I reversed the motor wires in a couple seconds.   That brass hex rod sticking up is a really tall head for one of the two 6-32 screws that secure the drive cover. The cover must be removed to reinstall and/or adjust the chain so quick and easy removal is a must. 

This is looking down on the controller after the body was installed.  Laura suggested the neat fuse block available from MCM Electronics  (link).   The power connection to the block is via a bare stud and bar.  I covered the stud with heat shrink and the bar with silicone caulk.   I don't find low voltages such as 24 volts intimidating.    However, I am intimidated by the high current capability of the batteries that can instantly turn a dropped tool into a welder.  Hence, I try to keep all exposed terminals covered.

That is the radio unit setting on the controller.  It comes with ~3' cable that is probably useful if the controller is buried in the bowels of a steel bodied locomotive.  The railbus body and roof is mostly wood so that is a good place to put it.   There are no attachment tabs so Duct Tape will be used.  (Every project needs Duct Tape someplace.)

So --- how did it run.   Ran great in the shop --- once I released the brakes.  Once the wheel size was entered via the setup menu the top speed registered at just less than 10 mph matching the calculations in the Drive section.

It was off to the Mill Creek Central track for a real test.  Dick McCloy jointed me for a really real test (it only groaned a little).   Ran perfectly.   Did a lot of testing of the throttle controlled slow down on the steep grades and it worked as advertised and far better than I expected.  The  E-STOP feature  brought the locomotive to a nearly complete stop in 20 or 30 feet.  This is as fast as I could stop by locking up the brakes since I have to reach for the brake handle which eats up 10 or 20 feet --- if I find it quickly. 

Did a lot of observing of the battery voltage to see if it appeared charge was going back into the battery when going down grade -- and  it appeared to be working.   Need to install an ammeter to get a better evaluation.

As expected I confirmed that I need to keep the air brakes for at least the park brake function.   Going to consider using one of Integral Circuit's Brake Controllers.
I did about ten miles total on the first run and then got bored.  Brought it home to finish dealing with the accessories and install an ammeter.

At this point the controller is exceeding expectations.   Will update this after I get some current readings.   The accessories are described in the accessories section.    

More Data (4/22/2017):

Today I did another test run at Mill Creek Central.   The addition of brakes, light and horn via the handset are great!    These features are discussed in the Accessories Page.  The big deal at this point is the ammeter.

The analog ammeter is zero center and plus minus 100 amps full scale.    A momentary switch changes the scale to plus & minus 20 amps full scale.

I did some runs at half power and the speed was just around 4.8 mph on a flat track.   On a 3 % up grade grade the speed dropped to about 4 mph and it went up to about 6 mph on a 3% down grade.  These are approximate.  I was alone on the railbus.

The current readings were very interesting.  On the level track  at half throttle the current was about 10 amps and dropped to near zero on a slight down grade.  The current  went up significantly for steeper grades reaching 30 to 40 amps on a 3% grade.    At one point on the 6% grade of the logging line it went to current limit of ~ 70 amps but kept pulling.   

On the steeper down grades significant current went back in the batteries from the regenerative process.  This  was as high as 5 to 10 amps on the 2%-3% down grades to 20 amps on the 6% down grade on the logging line.     The regenerative action did limit the increase in speed to maybe 25%-30%.   (I'm now a believer.)

One thing I observed was that the current didn't change much  as the throttle was increased (there was a spike initially but then it settled down to about the same current).    This makes sense  --- current is proportional to the required torque and torque is a function of the load.   Speed is related to the voltage.  As the voltage increases the speed increases.   Power is speed X torque and voltage X current.  Makes sense.

The bus seemed to run much smoother at the higher power settings.   I also verified that the top speed of nearly 10 mph is really too fast.    It looks like I can replace the 60 tooth axel sprocket with a 70 tooth sprocket.   That would reduce the top speed 16% and also increase the torque 16%.    That is worth doing!  

More Data (5/27/17):
The drive is now equipped with a 70 tooth sprocket.  Bench testing in the shop showed the no load top speed at 8 mph, about as expected.  The no load current for both motors in parallel is about 6 amps.  That is the no-load starting current listed on the motor performance graph for one motor.  These motors should be a bit worn and loose.

Tests at the track showed the anticipated increased in torque/reduced current.   The key performance indicator was that the wheels slipped before the controller went to current limit.   That was with one rider.   The controller seems to be a really good match to the motors.

At this point the controller exceeds all expectations.     

Additional design information is at the following links:

Return to Railbus Overview
Railbus Frame
Railbus Drive
Railbus Body
Railbus Chargers
Railbus Accessories


NLW Home