Need for track voltage regulation with TMCC board and prewar motors

My solution to those issues is to have more power districts fed by 18 volt transformers to spread the load. 

First, What John said.  Simple, the more power feeds you have the less voltage drop you will see around your layout.  I use 14 gauge wire from my transformer to my power distribution blocks.  From the distribution blocks to the track I use 16 gauge wire with connections about every 5 feet or so.  I also don't rely on a switch or cross over to feed power to an adjoining section.  

I have a PW ZW that I use for my primary power, which is limited to around 18v. And my oldest PW engine runs a fairly consistent speed around the entire layout.  I also have an early TMCC engine with 3 poll motors that runs perfectly with this setup.

I'm going to have to assume that you are planning on routinely running the prewar trains alongside, and on the same power districts, as the ones with modern electronics, otherwise you could simply use one output set to 18V for the modern engines and adjust a second output as needed for the prewar ones.  

I can think of two solutions (well, one solution with two ways to get the desired result), neither of which is a ready made product and would require some component level electronics work.  I mention this because it might scare some folks off, where as others are perfectly comfortable with "complex" solutions.  

Basically what I would do is use a micro-controller to constantly read the voltage of the track in either option.  from there my 'simple' solution would be to have a large load bank made with large wire-wound resistors, or perhaps car head light bulbs, that would switch on if the voltage rose over a certain point to pull it back down .  The more complex option would be, rather than load down your transformer, to use a pair of triacs to regulate the output, much the same way the "chopped sine wave' is produced on most modern transformers.  This gets into some somewhat complex circuits but would be the better option I think if you are only occasionally running the prewar engines.  

I'd be happy to help with more specific details of how this would work if they are needed, but don't want to spend too much time on it if there is no need.  

JGL

JohnGaltLine posted:

I'm going to have to assume that you are planning on routinely running the prewar trains alongside, and on the same power districts, as the ones with modern electronics, otherwise you could simply use one output set to 18V for the modern engines and adjust a second output as needed for the prewar ones.  

I can think of two solutions (well, one solution with two ways to get the desired result), neither of which is a ready made product and would require some component level electronics work.  I mention this because it might scare some folks off, where as others are perfectly comfortable with "complex" solutions.  

Basically what I would do is use a micro-controller to constantly read the voltage of the track in either option.  from there my 'simple' solution would be to have a large load bank made with large wire-wound resistors, or perhaps car head light bulbs, that would switch on if the voltage rose over a certain point to pull it back down .  The more complex option would be, rather than load down your transformer, to use a pair of triacs to regulate the output, much the same way the "chopped sine wave' is produced on most modern transformers.  This gets into some somewhat complex circuits but would be the better option I think if you are only occasionally running the prewar engines.  

I'd be happy to help with more specific details of how this would work if they are needed, but don't want to spend too much time on it if there is no need.  

JGL

JGL: thanks for the reply. Complex doesn't scare me (or I wouldn't have started this particular project!) My first/best thought was how can I have something power a SPDT relay to switch the power through a bidirectional (for AC) diode drop with about 3 each 10 amp silicone diodes on each side when voltage exceeds 20 or 21, which would create about a 2v constant drop under load (22 to 20 volts), then when the input voltage gets down to 20v or less the relay goes back to the NC terminals and feeds 22v again.  It's what could I use to trip the relay coil that stumped me.  Does your idea provide a way?

Don

Tony_V posted:

First, What John said.  Simple, the more power feeds you have the less voltage drop you will see around your layout.  I use 14 gauge wire from my transformer to my power distribution blocks.  From the distribution blocks to the track I use 16 gauge wire with connections about every 5 feet or so.  I also don't rely on a switch or cross over to feed power to an adjoining section.  

I have a PW ZW that I use for my primary power, which is limited to around 18v. And my oldest PW engine runs a fairly consistent speed around the entire layout.  I also have an early TMCC engine with 3 poll motors that runs perfectly with this setup.

 

Thanks. I do have large wire (12 ga buss fed in center by 14 ga "star" feed for each block), and the voltage drop is with no load other than the COD train, drawing about 2-3 amps because lighting is LEDs. Surprises me too.  Since I've invested in MTH and Lionel power supplies for most of my districts, would prefer not to have to sell them off and go to some industrial transformer to lick this!  The voltage drop issue is internal to the Z4000, since the measurement is at the output terminals. See my query to JGL just above.  I can't readily explain the drop, but it's not related to my wiring or to any extreme draw by this train.  The problem is unique to these motors and their preferred higher voltage range.  I will have 3 motors to spread load when done, so each will not be asked to pull more than a reasonable number of cars in the train.  Amps then will still be <4a.

I like the bi-directional diode idea better than either of the ones that I came up with, actually, as far as dead simple goes.  Just keep in mind you are going to need some really big diodes.  The benefit of using the load bank would be that it could be made responsive to the change in voltage at the transformer to keep output at exactly 18V (or whatever) where as the diodes would drop in .6 volt steps, or in the simplest form a 2 volt or so jump.  

As for how to read the transformer voltage, I would use a knock-off arduino micro-controller as the brains of the operation.  I'll work out some details for you, if you like as I have time over the weekend, but the basic setup would be to rectify the voltage to DC, then use a voltage divider to bring the transformer's maximum power output down under 5 volts.  this will give you a zero to 5 VDC voltage from your 0-24VAC transformer output.   This signal voltage can then be read by the Arduino, which, in turn can be programed to turn on a relay at a set voltage to turn on the diodes, or turn on a varying, pulse-width modulated signal to a set of big power transistors on a load bank or triacs on the line to the track.  If I went with the diode method, I think I would at least use a few relays to let the system have a few steps, clicking on more and more diodes in series as the voltage rises.  

JGL

A easier method,, though it should have been done when upgraded is to go with modular TMCC upgrade.  You could have used a Engine Mother board with R2LC and an ACDR for the Pre War Motor.  You can let AC Track Power directly into the ACDR which is just a H Bridge Triac circuit for controlling the motor.

You could have than used a small device to take the High AC input and regulate it down to a constant 16-18VAC and send this to the R2lC mother board.  Since the ACDR and the Mother Board only have a PWM DC Connection between them this would keep Voltage in the right range for the TMCC components.  The ACDR can handle a Higher AC Voltage.   G

GGG, modifying the engine electronics was my first thought as well, but it would require that you do the modification to every locomotive you put on the tracks, wouldn't it? 

With the Z-4000 set to 20 volts at no load, how many amps does it take to pull it down to 16 volts?

PLCProf posted:

With the Z-4000 set to 20 volts at no load, how many amps does it take to pull it down to 16 volts?

As noted, the train only draws a couple of amps, so it's not a high load issue.

Don

GGG posted:

A easier method,, though it should have been done when upgraded is to go with modular TMCC upgrade.  You could have used a Engine Mother board with R2LC and an ACDR for the Pre War Motor.  You can let AC Track Power directly into the ACDR which is just a H Bridge Triac circuit for controlling the motor.

You could have than used a small device to take the High AC input and regulate it down to a constant 16-18VAC and send this to the R2lC mother board.  Since the ACDR and the Mother Board only have a PWM DC Connection between them this would keep Voltage in the right range for the TMCC components.  The ACDR can handle a Higher AC Voltage.   G

Thanks, George. If the Arduino circuit idea from JGL proves unsuccessful for any reason, your idea here would be the best bet.  The cost and work of swapping is not extreme. Don

JohnGaltLine posted:

I like the bi-directional diode idea better than either of the ones that I came up with, actually, as far as dead simple goes.  Just keep in mind you are going to need some really big diodes.  The benefit of using the load bank would be that it could be made responsive to the change in voltage at the transformer to keep output at exactly 18V (or whatever) where as the diodes would drop in .6 volt steps, or in the simplest form a 2 volt or so jump.  

As for how to read the transformer voltage, I would use a knock-off arduino micro-controller as the brains of the operation.  I'll work out some details for you, if you like as I have time over the weekend, but the basic setup would be to rectify the voltage to DC, then use a voltage divider to bring the transformer's maximum power output down under 5 volts.  this will give you a zero to 5 VDC voltage from your 0-24VAC transformer output.   This signal voltage can then be read by the Arduino, which, in turn can be programed to turn on a relay at a set voltage to turn on the diodes, or turn on a varying, pulse-width modulated signal to a set of big power transistors on a load bank or triacs on the line to the track.  If I went with the diode method, I think I would at least use a few relays to let the system have a few steps, clicking on more and more diodes in series as the voltage rises.  

JGL

I've been intrigued by the Arduino stuff at my electronics store for a couple of years, so it would be fun to use that as the method here.  Need to learn about those devices and associated software anyway.  Note I already have suitable 10 amp, 50 PIV diodes on hand so that's not a barrier.  Dividing the drop into 0.7v increments by bringing one silicon diode pair in at a time would certainly add precision, so output to track would always be either 19.3 or 20v unless Z4000 primary output gets below 19.3v.  That should be OK, because the train behaves well down to track voltage of 18v or so. That could solve some other higher load passenger trains' concerns when a max distance from feed, too.  I use a K-Line voltmeter boxcar to see what's what "way out there", and even with my 14 and 12 ga wiring the voltage at the train can drop to 14v or even a tad less with the Z4000 set at 18v.  If the Arduino-diodes-relays approach is tied to a differential in voltage from baseline (is it?) then it could let me make the "usual" voltage for all other trains 19v, and keep the output to track at 18.3-19v.  That would add a few volts at all times, especially the critical times when the train is most distant. 

BTW I had a large basement loop years ago using ZW power, and a heavy train like two 2343 "growler" F3s with 4 motors, plus 9 lighted aluminum passenger cars drawing 8-10 amps, would drop the most distant voltage down by 1-1.5v, and that was with 12 ga copper wire in a continuous bus, with soldered drops every 6 feet or so for both sides of the GarGraves rail circuit, so that current had 2 paths to the train 50 feet away, and still there was this much drop.  Even giant wire doesn't mean "no voltage drop!"  If  you calculate the resistance per foot for your wire from reference tables, then calculate total resistance using wire gauge's ohms/foot and distance, that is about how it comes out, as well.

Thanks for your help with this! Should be fun figuring out how to make it work.

Don

cnwdon posted:

PLCProf posted:

With the Z-4000 set to 20 volts at no load, how many amps does it take to pull it down to 16 volts?

As noted, the train only draws a couple of amps, so it's not a high load issue.

 

Don

I guess that is exactly my point, if a 10% load pulls the transformer voltage down 20% there may be more to the issue than meets the eye! 

In an attempt to understand this, I would put a fixed load of a few amps on the transformer, maybe some incandescent lights or something, reset the voltage to 20 and try to run the loco again with the fixed load still connected. I am guessing that the voltage soars at no-load, rather than dipping at light load. In that case running a small load continuously might be all you need to do.

I would also be sure I was measuring the voltage with a good voltmeter. Be a shame if you went through a lot of trouble and learned that it was a measurement anomaly.

PLC Prof: Thanks. Have used expensive electrician's meter to confirm track voltage. I like your idea re: high volts at no load being part of the issue.  Will try a modest incandescent bulb load as baseline and post results.

CNWDON, 

Before getting too far along, after seeing what some other folks have to say, I think it would be a good idea to clearly define the problem, and to insure your track wiring is up to snuff as well, to eliminate that as a problem.  George (GGG) also got me thinking with his post on modifying the locomotive and as I think on it, that may be a better way to go depending on exactly what we are dealing with.  So, in order to understand exactly what is going on, and what options are available I have some questions.  

1.  As others have suggested, have you had a chance yet to see if the high voltage reading is simply a 'no-load' condition that could be solved by simply placing an always-on lightbulb across the transformer's output.  

2.  For clarity, what you have now is ERR electronics inside the pre war locomotive to run it with TMCC?  

3. Do you only have the single pre war locomotive that is a problem, or are you running several that would need a 'fix'?  

4.  Are you running the pre war locomotive on the same track/power block with any other engines with modern electronics in them at any time?  

Once we know exactly what's going on, it will be easier to see if modifying the engine or adding something to the supply to the track is a better option.  

As for the Arduino stuff, it is a  lot of fun to tinker with.  As to how hard it is to learn, I have a hard time judging that for other folks.  I was taught electronics and computers from a very young age and often pick up new technologies relatively quickly.  That said, the Arduino project has a wealth of free tutorials online, an extremely helpful community forum,  and the platform is designed to be easy to use.  As it related to things someone might want to do on a model railroad, once you've gone through the examples in any starter kit and understand how they work, you should have no trouble doing whatever you need for your train layout as far as controlling lights, relays, and the like.  Arduino is also rather awesome, in my opinion in that it is an open source project.  The board designs are freely available for anyone to produce, which has led to such low costs that the off-brand clones are only a few dollars.  I do recommend purchasing at least one of the genuine boards at some point as the Arduino product is constantly improved and funded through sales of the genuine boards and donations.  Once you feel you've done your part to fund the developers, however, there is no difference in a genuine board and the $2-$5 versions.  

JGL

My Point was primarily about selecting the proper electronic pack based on what you are modifying.  Unique in this case to Lionel and command for AC motors.

Airwire as has a board for G with a jumper.  The board can't take more than 18VDC for the electronic side, but the motor driver can.  So if you only need 18V no problem, but for other model where you might want higher voltage for the motor drive, they can use the jumper.   For example 2 12V batteries in series give 24V for the motor drive, but you only send one battery 12VDC to the control section.

Pre war AC Standard gauge is obviously a power hog and having the ability to use a system with higher voltage would be best in the initial conversion.

OR, swap in an MTH DC Pittman motor assembly an than you can go with the Cruise CDR.   Lots of folks back fit the MTH motors into there Pre War, especially the ones they like to Run.    Here is a picture for an electric.  G

To JGL: Thanks and answers:

1.  Is the high voltage reading is simply a 'no-load' condition? Placing big resistors to draw almost 2 amps before train is started does reduce the voltage drop; not quite done testing that, but you have nailed part of the problem with stable voltage.

2.  Yes I have ERR electronics inside the pre war locomotive to run it with TMCC.  There are 3 of the motors involved, 2 on one TMCC board and one on a second board to run as a lashup.

3. Do you only have the single pre war locomotive that is a problem, or are you running several that would need a 'fix'? Yes in the sense there are 3 of the same motor involved in this one train.  Only ONE train, though. Pride Lines COD A and 2 B's.

4.  Are you running the pre war locomotive on the same track/power block with any other engines with modern electronics in them at any time?  For any mainline operation, the answer is never; this is a long distance passenger train that will run from Chicago terminal to staging yard, and no need to run any other loco of any sort in the same block for center rail power.  So whatever higher voltage I use with the COD train is only needed when it's on mainline.  It can leave common areas where 18v is used, like terminal and staging, on 18v because speed is slow in all those places.  For one grade situation, if needed we'll decide that a helper is needed and add a concealable coupler to obs car.  A LionChief loco would serve nicely for that duty  :-)  Whole layout has common outside rails as a single entity; no breaks on common rail side of circuit anywhere on layout.  Each block has paired + and common with circuit protection, but all common returns are bussed together at the phased-together power sources and united by the common rail as well.

5. Re: Arduino control of relays and diode drops: still very interesting, and I'm sure the learning curve would not be too steep.  Just a matter of devoting the time to learn the tool, and then apply it to this case. Let me finish with the testing of the system with the 1.5-2 amp baseline load before I ask you to work out any Arduino details for my situation.

To GGG: Thanks---the MTH power idea is interesting.  Would be a higher cost option since I would have to change out 3 motor units.  Something to not forget if the overall solution remains elusive.  If done, would be a fine permanent solution!

Don

Ok, Don.

With only the one train to worry about, and only the one track, I think GGG's suggestion of modifying the engine is probably the way to go for simplicity.  This would let you drive the motors at 22-24 VAC to get optimal performance out of them.  I think George here may have a better understanding of which boards need what voltages, but I'll have a look if you can provide any documentation on exactly what is in these engines.  

For GGG, is it only the R2lc that has problems with high voltage, or are there other components as well?  I'm not familiar with the various products offered by ERR.  

___

As to the Arduino route, we can still play with it if you like, seems like fun.  

JGL

I wouldn't recommend higher voltages on any of the TMCC/Legacy electronics.  The only people that can really answer the absolute limits are Lionel, and they've made their statement, 19 VAC maximum.

The DCDR motor driver is a simple device with Triacs, Caps, a few resistors and MOC.  No IC chips other than the MOC.  The trigger circuit is driven by 5VDC, so as long as the CAP and Triac can handle the voltage all is good.  They can, CAPs are 35 or 50 V and the Triacs are 200V if I remember correctly.

The R2LC and the RS can't exceed the 19VAC.  G

I think working around the voltage limitations of the electronics is what we're trying to do at this point.  I'll take GGG's word for it on the limits of each of the boards in this case, as it seems reasonable to me.  The R2LC and Railsounds boards cant go over 19VAC, but the motor driver DCDR looks good up to at least 35V.  I have to assume this is an on board 35VDC, so the max AC would be what... 35 / 1.4141 or 24.7 volts AC on the input.  That will be fine with the 22VAC output of a Z4K.  

So, what we need to do is cut the voltage to the R2LC and Railsounds boards so that it cant exceed 19VAC even if track voltage is up to 22VAC.  Unfortunately there is no such thing as a 'simple' AC voltage regulator, so I think the simplest solution here is to use pairs of diodes in series to lower the voltage.  if we assume the input will never exceed 22VAC, the limit of the Z4000 transformer, then we need at least 3 volts of drop on through the diodes.  That's 5 diodes at the typical .6volts each.  I'm going to add an extra one in for a little wiggle room.  So what we need is 12 diodes wired up in reversed pairs to allow AC to flow through them for a 3.6 volt drop.  This string of diodes then has to be placed in line with the power to the R2LC board, which is pin #1 on the board.  

I'm not sure what the best way is to do this.  I've thought of unsoldering, or cutting off the pin then soldering wires to each side of the socket on pin1 and inserting the diodes there.  Another option would be to build a short jumper cable between the boards with the pin1 wire extended to place the diodes in series.  The jumper is probably a better option if you want to preserve the boards, but cutting off the pin is quick and easy.  The same process will have to be repeated on the sound board if you have one, and I'll have to look up what pin the AC Hot connection is, or exactly how the railsounds board is hooked up.  As said in an earlier post, I'm not super familiar with ERR electronics.  

Two small issues I see with this method are that the "turn on" voltage to the engine will be higher, going from maybe 8 volts up to almost 12, but I doubt this will cause any problems for you.  The second issue will be making sure any lightbulbs in the engine or cars are rated for the higher voltage.  

A third issue worth mentioning is that you'll have to make sure you don't forget to turn the voltage down if you run another engine on that track.  

JGL

Thanks, both of you. To confirm some facts queried above: the motor driver board is ERR Co's as I understand it.  The R2LC radio board is a Lionel board, and the RailSounds board I don't know for sure; may be Lionel's.  I will email Ken at ERR Co with cc to both of you, and ask if the motor driver board can handle 22v, as is true for Lionel.  The general directions for the AC Commander board set is not to exceed 20v, so it's not clear whether the driver board would be an exception.

If this is feasible, I like the "jumper" idea for sure: the less snipping or soldering at or near boards, the happier I'll be.

Don

An ERR Driver board is different than a Lionel Modular A or  DCDR.  I am talking about the ACDR/DCDR.  G

So, it seems the easiest solution would be to acquire a Lionel ACDR?  For the more difficult solution we could scratch build a motor driver board

JGL