Homebrew Transformer Speed Control

The problem with that circuit is that it's a "Phase Limiter" that controls the voltage by limiting how much of the sine wave gets through (controlled by the switching on/off of the Triac). It's more commonly known as "chopped sine wave". It will work with the old Lionel Universal Motors aka "PullMor", but if you have locomotives with sound boards like QSI/Proto-1, it will drive the boards nuts (I'm not sure of any damage that might happen). Used an identical circuit over 40 years ago to control heat guns. It's better for 120VAC motor control.

The most effective way to control the output voltage and get a pure sine wave is to use an "Autotransformer" on the primary side. An autotransformer is a variable step-down transformer that typically operates from 0-100% of line voltage or 0-140 volts AC. The problem is they're pretty expensive. The Right-of-Way power supply used two of them, plus two 120-to-24VAC step-downs. They were great, heavy, and expensive.

Thanks for the reply I knew there had to be more to it. Is there a way to mimic the circuitry of the old analog transformers so as not to affect the sine wave? My ProtoSound locomotives run great on a postwar Lionel KW. I am guessing "Variable Secondary Winding" is what I am after?

Well, a lot of "modern" transformers are using the standard chopper circuit, including the MTH Z-500, Z-750, and Z-1000.  Also in the pack is the famous (or infamous) CW-80.  I really don't see why he can't use that, it's no worse than these examples.

Here's the Z-1000 output at half throttle

For a nice contrast, here's the CW-80 at half-throttle, check out that nice waveform!

Thank you for the reply, very interesting to see it on the scope. Now for a new question, most stock motor control circuits are rated for 110 - 220V AC, could these units be run on 18V from the transformer? I am guessing not because the TRIAC and DIAC would not have the correct trigger values.

You would probably have to pick components based on the voltages involved, but the same basic circuit should work.

Thanks for the encouragement I will start looking for the components. It seems the that 18V AC is much lower input voltage then most regulators are designed to handle. Can anyone explain how modern transformers like Z-1000 are able to regulate the lower voltage?

You don't need a high voltage to trigger a triac itself. I think most of the designs you're referring to use a microcontroller to monitor the zero crossing of the AC wave, and then trigger the triac with a logic pulse at the desired time point in the AC cycle.

In your circuit the diac has a high breakover voltage and therefore by itself triggers the triac at a defined point in the AC cycle.  I'm not sure what you can use to achieve the same effect at low AC voltages.

@Gsyverstad posted:

Thanks for the encouragement I will start looking for the components. It seems the that 18V AC is much lower input voltage then most regulators are designed to handle.

If you are just using the transformer core and are looking for just speed control, you can use a motor or fan speed control on the 120/primary/input side of the transformer to get 0-18 volts on the secondary.  I have done this many times and actually built a 10 amp controller for my brother-in-law w/ whistle & bell controls about 20 years ago that is still working well.

Back in the Pre-Z4000 days of overheating ZWs when trying to run show layouts, the $$$ ROW transformer was a commercial option.  A lot of folks come up with cheaper options.  There was a lot of discussion on boards (remember Compuserve?) about using 120 VAC Triacs and an 18v step-down isolation transformer.   There were a number of limitations to the solution including a requirement to have a load resistor (lightbulb) on the output side (can't remember the technical reasons).  So I decided to build power supplies using Hammond 18V stepdown transformers with a variable voltage autotransformer (AKA Variac)  on the input side  - the same circuitry as the ROWs.  A successful, robust solution at half the cost of the ROW capable of running 15 amp loads without breaking a sweat.  Some assembly required but isn't that why we enjoy the hobby?  I now use the autotransformer to trim the output voltage to 18v and TPC400s to control track voltage.  An evolving challenge has been circuit protection and considering the current cost of instant trip magnetic or electronic overcurrent breakers, I don't recommend a home-brew solution these days.  Off-the-shelf Z4000, PowerHouse and PowerMaster gear is good value with superior circuit protection.

The autotransformer models I purchased 25 years ago are still available at about the same price (~ $75) so that is an option.  A search of the net will turn up lots of links to Triac and motor controller solutions.  They work, but YMMV.

The best solution for defective post-war transformers is retirement.  There's a current thread on a Type R and its internal wiring - I use one on my test bed - but even that needs good output protection added on and after reading of some of the Type R issues, I'll replace it soon with new gear.  The days of thermal breakers with seconds before trip time is OK are long gone.

The "load" for triacs is a requirement today.  The triac needs a brief shot of resistance to current flow to get the ball rolling.  The go-to solution nowadays is a small capacitor across the output.  That gives the triac a shot of current to trigger, but doesn't impose a constant load on the output.  For TMCC LED upgrades, we add a .01uf cap across the lighting outputs to allow reliable operation.  Without the cap, the LED doesn't draw any current until it's close to it's rated operating voltage, to the triac never "sees" it and doesn't trigger.

Thanks for all the help everyone, I did not realize these transformers could be a senior design project for electrical engineers! Anyway, I will be ordering some components soon and hopefully after the holiday I will be able to report my findings. Thanks again for the help.

@Professor Chaos posted:

You don't need a high voltage to trigger a triac itself. I think most of the designs you're referring to use a microcontroller to monitor the zero crossing of the AC wave, and then trigger the triac with a logic pulse at the desired time point in the AC cycle...

The CW-80 schematic has been posted several times on OGR...for example here.  Presumably you aren't planning to use a microcontroller... but rather a potentiometer to choose where/when in the line cycle you turn on the triac to chop down the voltage.  I don't recall seeing the schematics published for the MTH Z-1000 (and its cousins) controller schematics...but that would be an example of a potentiometer controlling the chopping time.  I'm assuming part of your objective is to learn and experiment as part of the hobby experience.

The MTH Z-4000 patent is decades old, but it's a good reference to learn/study how to synthesize a more "pure" sinewave (vs. a triac chopped sinewave).  If done today, I'd look at more modern techniques such as used in home-theater "digital" amplifiers ...where in amazingly compact packages you get hundreds of Watt of audio power.  Audio amps are, say, 20-20,000 Hz...which of course includes 60 Hz.  So just put in a low-level 60 Hz sinewave signal into a modern audio amplifier and voila!

Well after some tinkering the results are in, and it works pretty good!

My homebrew transformer can be seen here powering trains on the Christmas Layout. The motor controller was wired in ahead of the transformer core and the diodes add a whistle and bell control. Thanks for the help everyone!