model railroad power supply

[KeepItSimpleStupid]

That IC is about $900 and that's just the IC.

One real issue to ask is how current limit is handled in the RR controller. If it's fixed, that's one thing.If they vary it all over the place, then that's something else.

When you say 0-24V for your outdoor railroad, how is reverse handled? Mechanically? Electrically? Both?
Are you expecting to automate reverse on your outdoor train?

 

[danjones@nethere.com]

But there are other simnilar cheaper ones available, like the ones used on combat robots, for example.

 

[4pyros]

Danjones; The Smart Cab Control Module you are looking at does it all. But it is only rated at 50 watts or about 2 amps at 24 volts. It needs a fixed voltage input so you can use your big power supply just fine. Speed and direction is controlled by your computer. For more power you use more than one Smart Cab Control Module and devide your layout into blocks. One Smart Cab Control Module per block all powered off your big power supply. YOU NEED NOTHING ELSE!! For manual control you would have to switch between your small controller and the new setup at the track feeds. Andy

 

[danjones@nethere.com]

4pyros,

Sorry, you're missing the point. Each of my trains pull 15-18a starting up and 10-12a running. Voltage (0-24) doesn't seem to change current draw all that much. Permanent magnet motors(8 of them) and incandescent lighted passenger cars. Total train length approaches 50ft. My layout is 350' end to end, 7 blocks. I like the SmartCab modules, they just won't supply sufficient power to operate my trains without over currenting (overheating).

Thanks, Dan

 

[KeepItSimpleStupid]

Do you have an answer to these questions yet?

When you say 0-24V for your outdoor railroad, how is reverse handled? Mechanically? Electrically? Both?
Are you expecting to automate reverse on your outdoor train?

 

[KeepItSimpleStupid]

Here is something else worth checking out: https://www.electro-tech-online.com/custompdfs/2010/11/PAD127_Rev_D.pdf

This looks even cooler: **broken link removed**

 

[danjones@nethere.com]

Currently I am using a Bridgewerks controller that has 25a 0-24vdc with a built-in reverse. Of course, I can only run 1 train at a time this way, but I do have sufficient current at all voltages! I was just looking for a way to automatically run several trains at the same time. My layout consists of a loop to loop with 2 passing sidings in the midddle. I put together a totally mechanical system that will allow me to run 1 train continuously without my having to do anything but watch. That does get boring after awhile....

 

[4pyros]

Danjones; Ok then I am asuming the small power supply you said you have in your first post is realy the Smart Cab Control Module and what you want to do is incress the power handling of the controler and run it off your big power supply. Is this right? The first thing I would do is cheak with CTI and see if thay have a higher power unit. Trying to make a current bosster for this controller is tricky, the output is PWMed some of the time, it gose reverse polarity to change direction and it shorts out the output to stop. Redesigning the unit with out knowing the computer interface will be almost imposible. You will still need one for every block. Andy

 

[danjones@nethere.com]

4pyros,

I started with CTI, offering to pay them to make me one, and they haven't even replied to either my phone or e-mail messages.

Yes, you need to connect a suply to each block that you intend to use, serially. This is done through a program.

My intention was to boost the power in some manner and handle the direction thru use of a relay.

Thanks

 

[4pyros]

danjones; It seems the CTI system is not well suted for grardan rail roads, but I asume someone has tryed this before. Have you tryed the CTI forms or some of the modle RR forms for info? I am shore it can do done but how easy it is, is the questoin. Having the direction control after the booster mite make it simpler, but you still need braking. You could do that with a relay and a power resister. Can you disable these functions in the Smart Cab Control Module and use it for speed only? Andy

 

[danjones@nethere.com]

no, I don't need braking. the braking function just lets you stop the train and restart it without having to do an actual throttle down and up process. nice, but not necessary. the reality is that the braking function adjusts the voltage, so if I have a voltage mimicing circuit, it will "brake" for me.

 

[KeepItSimpleStupid]

Tonight, I read the CTI manual and the FAQ and the manual differ in how SmartCAB works. It suggests that the system uses 1 power supply and a 100 step linear switching regulator. All good. The FAQ says it pulses the voltage at low speed, the manual does not say this. Pulseing is a form of PWM. The max voltage is set with a pot on the SmartCAB module. That's a good thing.

There is two heat sinked parts on the module and what looks like a relay. Questions. What are they and what do they do? Relay could change direction. One transistor could be the output device for the variable power supply and the other could be the switch which can pulse the motor. This seems very plausable.

There is an idle voltage adjust. It artificially makes a zero at some other voltage. I BELIEVE it has two functions:

1) Train detection (current in a given section of track indicates train is there) and 2) Could possibly run interior lights or LEDs in the train.

There is a residual 1.2V on the rails at "Zero" throttle. Bridge recifiers located in the train for lighting would null out this 1.2V and with a given zero throttle and a possible switching regulator in the train lights could be illuminated. This is a guess.

It appears it needs current sensing although based on the instructions, you can't set any limits. Short circuit protection is part of a safety aspect. Reversing direction HAS to know when the train has stopped. Thermal protection exists.

The controller has to be aware of current.

I'd love to know what parts are on the board.

This is my SWAG as to what's happening in the controller.

Thus I see two approaches:

#1) Find the power control signal and assume it's a voltage and condition it to run a commercial switchmode PS that is current limited. That's like OUCH! Been there, done that. Make a higher current pulse FET. Find the direction signal and use it to operate a similar higher current relay. Remove the effect of the existing relay. Remove the effect of the existing FET. Feedback the scaled current (significant glue required and isolation probably required [been there]). Advantages: less heat. Disadvantages: relay, some amount of reverse engineering, lots of mods to existing circuit.

#2) High power OP amp with unregulated bipolar rails. Probably custom xformer (been there, done that). Scale SmartCAB output voltage as input to power OP amp. Feedback the scaled current. (significant glue required). Current limit the power OP amp. Advantages: No changes to existing circuit. Disdvantages: OP amp dissapates heat, probably requires fan cooling, noise

In both cases, you could demonstrate proof of concept with 3 Amp rated parts. Some aspects would be the same for #1 and #2.

Questions remain:
1. Which is easier?
2. Which is cheaper?

 

[KeepItSimpleStupid]

When I sleep on things, I sometimes get other ideas. One that came to mind is in relation to current feedback.

In any event, two conditions MUST be sensed. No ifs,ands or buts.
1. Overcurrent/Overload/short circuit or whatever you might want to call it.
2. Direction changes CANNOT be allowed to happen until the motor has stopped.

These above safeties MUST exist. 20 Amps * 23 V is is 460 W which is significant. All sorts of bad things will happen if the motor is not stopped when the direction is changed. It;s bad for the motor and the electronics.

If the controller really doesn't care about actual current and doesn't need that info, then the following might be possible.

1. Define overload as the current exceeds X Amps for X time and after such time the system shuts down and requires a manual reset. The aux system can crowbar the controlling system so it knows. Current limit woud be built into the design anyway.

2. Within the direction control system only allow switching when the sensed current is zero.

 

[4pyros]

Instead of designing something new maybe we can modify the existing controller to handle more power. If not we can do over-current, direction, and brakeing and fail-safes externaly and use the controler for PWM and voltage control only, but this may mess up the software control.

danjones; do you have the new controller now? Andy

 

[KeepItSimpleStupid]

Instead of designing something new maybe we can modify the existing controller to handle more power.

That's the idea, but a 20 A, 0-24 V PS is not an easy animal to design.

If not we can do over-current, direction, and braking and fail-safes externaly and use the controler for PWM and voltage control only,

Over-current is usually part of a commercial supply, so with an analog out for current (unfortunately it's referenced to power supply positive unless isolated), you can force a crowbar across the original supply. e.g. Make it draw 3.5 A

Braking Reducing the voltage is not as effective as a 4-quadrant power supply,but probably OK. Has been for years for model trains.

Direction control Yep, force the controller to always be forward and reverse the direction with the proper protections externally.

software control mess up

- It's probably not that smart .

danjones; do you have the new controller now? Andy

Good point

 

[4pyros]

KeepItSimpleStupid;

Over-current is usually part of a commercial supply, so with an analog out for current (unfortunately it's referenced to power supply positive unless isolated), you can force a crowbar across the original supply. e.g. Make it draw 3.5 A

Here we can use light bulbs in series with the track power that will not light under normal load. Put a LDR on it to see the light from a track short and have the computer disconnet the power to the track.

Braking Reducing the voltage is not as effective as a 4-quadrant power supply,but probably OK. Has been for years for model trains.

Not reducing the power but putting a braking resistor across the track power to stop the motor.

danjones; How many differant trains would you like to run at the same time? With the CIT controller anything more than two becomes quit complex. Have you consitered DCC?
Andy

 

[KeepItSimpleStupid]

My train layout some 50 years ago had one of those clicking breakers. Never used the one on the transformer.

LDR's at 15 A, I don't think will work. I have an army tube tester that used a light bulb as a fuse.

I automated a gantry crane for a model railroad some 20+ years ago and I had to do that. OFF/Brake shorted the motor and stopped it instantly. 3 of the motors were about 1 Amp @ 6V or about 6 Watts. I did not use PWM at the time. Overcurrent like what would happen if the program failed did stop the motor when it ran into an immoveable object. Power had to be recycled to reset.

I can't see any inexpensive way out of this. High power is just that, expensive. Some amount of reverse engineering and measurements are necessary.

It's tough to use PWM at some fixed condition for ANY motor. I tried it once for a car heater motor. It sung to me. Pulseing a motor is about the only way of making a motor move really slow.

I think your in agreement with the requirement of short circuit protection and the need to have the motors stopped when reversing direction and hopefully that it must be built-in and not up to the operator.

 

[4pyros]

KeepItSimpleStupid;

I think your in agreement with the requirement of short circuit protection and the need to have the motors stopped when reversing direction and hopefully that it must be built-in and not up to the operator.

Yes we are in agreement.

Look here for the lamp over current control. It looks like he used one controller with one big transistor on it, High Power? **broken link removed** Andy

 

[KeepItSimpleStupid]

Note to Dan:

In electronics, the power supply and the case make up a large percentage of the cost of a device, this is lots of power supplies and lots of cases which means lots of $.

Is DCC not an option where the rail carries fixed power and communication and the loco's just have to switch (PWM) the power to the motor?

Thinking out loud: The number of trains on a given track also increases power requirements.

So, I guess wer're back to distribuing power is the best option.

 

[KeepItSimpleStupid]

Devil's Advocate & Murphy:

Suppose at low throttle CTI's system reverts to pulsed DC. Would the boosted system be able to overcome inertia? PWM is more suited for high current motors. Would there be a problem?