Cconventional DC control using the TIU's AC output variable channel

So for the record, this is the other thread.  The thread started meandering so I suggested a separate thread specific to the matter at hand.

At the risk of providing too-much-information, I'll ramble a bit to collect scattered thoughts and issues that have come up in the past in regards to this application.

To recap, there are 2 fundamental issues. 1) converting TIU AC output to DC, and 2) reverse DC polarity for bi-directional operation.

AC-to-DC is relatively easy.  A 10 Amp bridge rectifier is less than $1.  Two wire in, two wires out.  But no control over polarity so engine goes in one direction.  There are many applications where single-direction is good enough considering you have wireless remote control over voltage and hence speed.

Adding DC direction control via polarity reversal can be done with a DPDT switch using the classic cross-connect configuration:

For remote control, this would be a DPDT relay with a basic on/off signal that activates the relay.  For example for about $2 here's one with 10 Amp contacts and has a socket so no-soldering required. 

There are electronic equivalents of this typically using 4 transistors in a so-called H-bridge configuration.  In this case a basic on/off signal turns on 2 of the 4 transistors which effects the DC polarity reversal at the motor.  In the diagram, the switch symbol is just transistor.

So a real circuit might look something like this, and there are chips/modules which integrate all this into one component.

The problem is how to integrate the AC-to-DC conversion AND the polarity reversal under DCS control via WiFi tablet controller.

Since the OP has an AIU (per other thread), one idea is to use a relay in the AIU to control a DPDT relay.  So the TIU variable channel would alter the AC voltage, an external bridge-rectifier would convert the variable AC to variable DC (of one polarity), and an external DPDT relay would flip the DC polairty under DCS control (via AIU) for direction control.  Gets the job done but can be awkward to change directions if having to flip through tablet screens/menus between voltage control and AIU relay control.  Maybe not so bad for an occasional direction change but clearly a non-starter for yard switching operations.

The most common method for AC direction control of DC motors is the electronic E-unit which has the bridge rectifier and solid-state H-bridge polarity reversal circuit.   And it integrates the direction control using O-gauge AC standard of momentarily removing AC track power (for say, 1 second) to sequence through 4 states, namely forward-neutral-reverse-neutral.   There have been slight variants but it is still sold today...$16.50 MSRP.

Two wires in (AC track voltage), two wires out (DC motor voltage).  A forum member reverse-engineered this design and published an impressive how this electronic E-unit works. 

There are more state-of-the-art electronic E-units that can handle more current, have some additional features such as directional light outputs, more compact by using surface-mount components, etc., but in the end it's 2-wires in, 2 wires out with a direction change method that is part-and-parcel of the O-gauge conventional control method so the tablet interface of the TIU variable channel will have combined speed and direction.  I figure there are thousands of E-units sitting in drawers having been removed from conventional-only AC engines that have been converted to command-control.

Here's an example of a modern design using surface-mount components.  2 wires in, 2 wires out.

So the idea would be to place an E-unit at the output of a TIU AC variable channel, and then the DC output of the E-unit feeds the HO DC track.  No modification is required to the conventional HO DC engines.  Needless to say, an O-gauge E-unit would not even fit in an HO chassis!

My next post will be issues and workarounds for low-voltage operation using electronic E-units.

Have to say, I am impressed.  Even parts numbers and prices.  I am looking forward to your next post STAN2004 

There is a low-voltage issue specific to using O-gauge electronic E-units with the MTH TIU and HO conventional DC engines.

Many (all?) O-gauge AC transformers have a minimum output voltage of around 5V AC or so.  There's something called the TREOS document floating around that discusses in exquisite detail the whats, whys, and so on of the voltage levels for 3-rail AC operation - I didn't quickly find it on the web.  I believe it stands for Three Rail Electrical Operating Spec (or something like that).   E-units require some minimum AC track voltage to operate properly including have enough voltage to control the transistors and to maintain memory of which state (Fwd, Neutral, Rev) it is in.  5V AC is plenty to keep the solid-state electronics functioning.  And most (all?) O-gauge engines in conventional will stall/stop when the track voltage is lowered to 5V.  So you can have an engine on the track that is stopped yet have power to keep the lights on so to speak.

The low-voltage issue is then with HO DC conventional engines which can be as simple as a DC-motor directly connected to the 2 rails.  I think I've seen engines that move with only 1V DC on the rail.  So, if you use an E-unit that requires, say, 5V AC to operate...the minimum DC voltage it presents to a low-voltage DC engine might be far too high to operate at low speed or to slowly decelerate/accelerate to/from stop.

Additionally, I suppose to emulate the minimum output voltage of transformer throttles, the TIU Variable Channel has a minimum voltage setting that insures the AC voltage does not drop below, say, 5V AC.  If memory serves, the minimum "allowed" voltage is a software controlled parameter that has changed over the years in various TIU software releases.  I suppose it was a trade-off of insuring the minimum AC voltage to keep E-units functioning against guys who wanted to control AC voltage down to 2V, 1V, whatever - presumably for this very application of driving a bridge rectifier for low-voltage DC operation under DCS remote control.  I don't know where this stands today in the current TIU software.  And I haven't looked into what the tablet controller interface looks like for Variable Output operation.  So this is one of those read-the-fine-print considerations for the E-unit workaround.

But let's say you have an E-unit that requires 5V AC for proper operation but you want to be able to bring the motor voltage down to 2V, 1V, whatever.  I'd say the easiest way to do this is with one or more bridge-rectifiers added to the OUTPUT of the E-unit. 

If coming from the HO DC world, this is probably seems alien, but it's arguably the simplest way to lower AC voltage in small steps.   A pair of diodes wired back-to-back in opposition will drop an AC voltage by about 3/4 Volt.  So cascading several bridge-rectifiers (4 diodes or 2 pairs each) can drop several volts AC for about $1 in parts.  Seems complicated but relatively basic wiring...a daisy-chain of 2-wires in, 2-wires out, 2-wires in, 2-wires out, etc.

So in this scheme, the TIU AC variable channel is configured to drop to whatever minimum AC voltage is required to keep it operating.  Then one or more bridge-rectifiers is used the drop the E-unit output by, say, 3V, 4V, whatever so that the DC track voltage goes down to ~0V when the TIU AC output drops to the minimum E-unit voltage .  Note that the back-to-back diode method is needed since the polarity of the E-unit output (to the track) is unknown or can be either polarity depending on direction.

But wait, there's more!

An astute observer will note that a reversing protocol that uses some interval of 0 track voltage to change state must define what 0 Volts really means.  That is, you can't have a variable AC output that can go to 1V, 0.5V, whatever as you slow down an engine without hitting the 0 threshold which would, practically speaking, be some voltage near 0 but not 0 itself.  This would be in the TREOS document - I have it in paper form somewhere but haven't looked at in years.

So where I'm going with this is a roll-your-own reversing mechanism using an electro-mechanical DPDT relay as the polarity reversing element.  The point is a relay with 10 Amp contacts leapfrogs any issues of whether the E-unit can handle 2 Amps, 4 Amps, etc.

I haven't worked out the details but being the cheapskate I am, I'm imagining something using low-cost eBay relay modules.  Specifically something like a bi-stable circuit (aka flip-flop) which alternates on,off,on,off with each trigger.  For example, a bi-stable relay for ~$2. 

Then, every time the TIU variable output goes to 0 Volts (when DIRECTION button pressed on tablet), the voltage drop-out toggles the bi-stable circuit.  The bi-stable then controls the DPDT relay which reverses the polarity of the bridge-rectified variable DC voltage.

So this is an alternative to the E-unit with the advantage of 10 Amps (whatever the DPDT relay can handle).

Many of the low-cost eBay relay modules operate on 12V DC so some source of 12V DC must be provided.  Probably the simplest is just a 12V DC wall-wart with an adapter plug to provide screw-terminal connections so as not to require stripping the wiring to the plug.

Again, this is a half-baked back-of-envelope musing on alternatives to the problem at hand.

Back from today's train show with a Lionel E-unit 101-1. Looks like this.There are two soldering posts (you can see them on the left in the above image).  The TIU can be connected to these, but where is the output? I am not sure where to wire the E-unit to the bridge rectifiers. 

Stan2004, any advice?

Take a look at this thread: https://ogrforum.com/...ith-bridge-rectifier

I have not used a mechanical E-unit (like your 101-1) in the manner described (with external bridge-rectifier) to generate DC+ and DC-.  Hopefully one the the guys who apparently have done this will chime in.   I can see how this approach could have better current-handling capability than an electronic E-unit where "delicate" transistors switch current vs. arguably more "robust" mechanical contacts.

Looks like it is missing finger plate with 4 wires.  Either way, that device will not work.  That is a AC reverse unit swapping phasing polarity of an armature coil voltage with respect to a series field coil voltage with constant ground.  Swapping AC input phasing to a bridge rectifier will not changed the DC output polarity?

Or using it to swap a DC output doesn't really work because it has a single input wire and does not swap the second wire.

You really need an electronic DC reverse unit like the Williams Unit which is rated at 8 amps.  G

Sounds like an Arduino with a relay connected might be useful.  A drop to 0 volts could trigger the relay to latch/unlatch.  Lots of cheap 2-relay boards out there.... probably would need to latch/unlatch both relays because most of theses boards use a SPDT relay and polarity reversal is done with a DPDT.