Skip to main content

Replies sorted oldest to newest

Interesting question. A while ago, as a test,  I tried running a Lionel open frame universal motor from the earlier BlueRail board which outputs PWM and it worked OK. Not sure this would be true across the board for all PWM sources. Also, if you power universal motors via DC, there is a widely used circuit which provides rotation control with DC polarity and you do not need an E-unit.

"Also, if you power universal motors via DC, there is a widely used circuit which provides rotation control with DC polarity and you do not need an E-unit."

The "circuit" consists simply of a full-wave bridge rectifier in series with the field. The controller is a DPDT switch. When the switch reverses the track polarity, the commutator polarity changes, but the field polarity does not. Therefore the motor will reverse.

Hi Bob,  So with my efforts, Arduino PWM at 12000 Hz, I get good slow-speed control but worry about potential over-heating.  I give my stuff frequent breaks, however.  With a smoothing cap on the output, the MOSFET seems to heat up more and the slow-speed advantage is gone. Here's my video with the cap on the output:

I may have time to shoot something with straight PWM.  The motors sound ever so slightly different on the straight PWM, hence my concern.

Take care, Joe.

Joe Rampolla asked--

Does anyone have information on how PWM (pulse-width modulated) DC affects the Lionel universal motor?

@Joe:   This is a question related to supplying a Pullmor motor from the MTH 4000 transformer.  I had to measure this effect on a two motor engine at a wired distance of 250 feet out plus return (reasonably closely spaced wiring).  The output is not a sine wave as is commonly said, but rather a fixed 28v sine wave cut into 5 pulses... from the patent I could not deduce this number, but in literature circa 1992 I found a discussion of how to get the form factor of such a wave of 5 cuts.  I cannot recall if this was per half cycle or full cycle.  On reflection, I think it must be 5 per full cycle or perhaps 6.  This is because of the similarity to 400 cycle circuit practice, where we use aluminum wire (to get increased surface area) at a cost of 2 wire sizes for same conductivity, plus another 2 wire sizes because of the increased form factor.   Please note that at the end of 4/15 (midnight) I corrected some wooly-headed statements  in the paragraph which follows immediately below.  The bottom line is not affected because my line of thought hurtled onward, these few errant wheels regaining the rail as if they had never left it (...a rare railroading thing).

My measurements on the layout indicated that the increased form factor would require about not 3*, but one and a half wire sizes just to overcome the output form factor of the 4000.  That I had both calculated *mentally; and measured on the layout as about 1.35.  *One and a half wires sizes is not *~1.33 (it is *not actually cube root of three, *but rather that cube root =1.260; *then plus the square root of said cube root = 1.122.  *Thus the total =1.382, bit larger than my guess at ~1.33).   So this would have made the #16 wires into *not #13, but *#14 wires (nearest larger available size in limited quantity is #12, and 75-C temp insulation would be required, and usually is *available).  The distance is yet a problem; in my case the 4000s were back at the refreshments.   The MTH test layout (in DC, not owned by MTH) carried 120v to each TIU location, installed by an electrician.  A wire pair carried the low voltage to each 6 foot section of track.

An interesting comparison:  The postwar Lionel ZW had the low voltage coil under its rollers wound from #14 wire, but square wire and thus equal in size to #13 wire.  (In the Rochester layouts, Lionel used #10 wire for distribution to the track.)

A note on the 4000s track output breakers-- These appear to be commercial thermal 12-amp (so marked}.  In American practice, these would carry about 15 amps for 3 hours.  The ABA dropped the screw holding motor to truck one day, which skated atop the 3rd rail.  I estimate a current of about 25 amps, and made the 15-yard dash from the TIU location to the 4000, and threw the handle to closed before the breaker could respond (patent says 24 seconds to open, no other detail; requirement at time was 60 seconds).  I don't recall if this TUI had the 20-amp automotive fuse protection or not (I lifted the track feed to another channel, lacking an MTH engine for test).

After I wrote this, I realized there is a lot of useful information here for builders of large layouts.  It is worth noting that Lionel, in their TMCC system, generally recommend that remote control engines be operated with full voltage on the track if possible.  This reduces the form factor problem at high loads very dramatically.  I would note that 2-motor can motored (DC converted on engine) units seemed to operate satisfactorily on the mentioned club layout, both MTH and later-built Lionel.

Hope this helps.  --Frank

*Aatarics mark added or changed words or phrases; "...*not... is used to show first posted value now noted as incorrect.  Time 0250.

Last edited by F Maguire

Way back when the dinosaurs roamed the earth, maybe 1992, I tried putting a diode and electrolytic capacitor upstream of an E-unit to stop the buzzing. And it worked, but the E-Unit gets crazy hot! The coil of an e-unit is like an inductor, and resists the 60Hz, when the signal is just DC and a little ripple, just get more current, more heat.

There may be a way to keep the coil of the E-unit running on AC, but the motor running on a bridge rectifier, just keep in mind, it isn't going to be smooth DC like a battery. The trick would be at the fingers and drum of the E-unit, would have to find my K-Line book, don't have any more e-units. Most likely you will not want to have any wire of the motor running to chassis ground / outer rails, its just a weird thing with bridge rectifiers, if you study the resulting schematic with the wire ground, you would end up shorting out one of the diodes, and people have not been using bridge rectifiers with four diodes for what maybe 70 years and not notice the fourth diode is not needed, no way. Reductio ad absurdum, reduced to the absurd.

Add Reply

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
Link copied to your clipboard.