I’m very rusty with my circuit design skills which I haven’t used in decades, looking for some guidance. In short, I’m looking to limit a constant voltage supply into a load with a max of 4 Amps at 14 V without significantly reducing the supply voltage in the process. Fortunately, I think I can tolerate a few volts of supply voltage drop and still operate ok. The application: I’m trying to help out a friend who wants to convert a G-scale model train to use a Blunami DCC system controller operating on a constant 14V bus. The controller is pretty sophisticated but has a strict 4 amp total load limit. He has at least two locomotive units which will be used on the same track at the same time. Each locomotive has a stall current (max power to the DC motors) at or above 6 Amps. This current draw can be sustained almost indefinitely by the locomotives if something is stuck, so I can’t use a solution that only protects for start-up or transient power conditions. Im not certain about all my possible options here but the simplest would be if I can individually limit each engine to pull no more than 4 amps while still providing 12 v or more to the DCC controller being added to each train which pulls power from the 14v source through the tracks. I can’t limit at the power supply since each train can pull power individually, hence the supply must support 8 amps or more in total but each load must limit the individual current to avoid damage to the DCC controllers.
I suppose I may be able to limit the power to the DC motors (2 per train) to 2 amps each so that if they both stall at the same time, the load on the DCC controller is limited to 4 amps. Though I worry that trying to balance the power this way won’t work well with the train as I have no idea what problems will occur if the two wheel sets aren’t sharing the load equally. So I really prefer a source limit into the DCC controller over a load limit to the dc motors for the wheels.

The other issue is space constraints, since this circuit and any heat sinks need to be mounted inside the locomotive housings. So I need a minimum part count solution that doesn’t require a circuit board or a lot of volume.
I looked at threads on this forum for ideas but I didn’t see any solutions that seemed to fit the bill. I’m trying to avoid any solutions including non-resettable fuses or anything which would require opening the locomotive housings for repairs as they are a real bitch to get into. I'm hoping for something better than a self resetting fuse as each time the fuse opens , he would lose connections with the locomotives until the fuse resets and the software may just get stuck in a loop.
Thanks in advance for any suggestions on this.

 

Welcome to AAC.

I think a fast acting resettable fuse might be your only practical option. A device capable of sinking 28W for a sustained period is not going to fit your space constraints.

Maybe the fuse could be placed exclusively inline with the motor's supply so anything else in the locomotive stays online. Or perhaps it could disconnect only one of the two motors to allow for a kind of "limp mode" operation.

I'm not familiar with the controller, but a more sophisticated approach might be to add an additional device based on an MCU that talks to current sensors on-board the locomotives and throttles back the misbehaving train to maintain maximum load.

 

I've always been interested in model railroading, but I can't understand why DCC is still prevalent.
If the controller can't handle overcurrent, it's not very sophisticated. Limiting the controllers current without limiting the load will result in voltage collapse. With a 6 amp stall current, a 2 amp limitation will almost certainly promote a stall condition. Sounds to me like the load needs to request capacity from the system manager, who, at the time, may or may not grant, depending on system priorities.

 

I’m very rusty with my circuit design skills which I haven’t used in decades, looking for some guidance. In short, I’m looking to limit a constant voltage supply into a load with a max of 4 Amps at 14 V without significantly reducing the supply voltage in the process. Fortunately, I think I can tolerate a few volts of supply voltage drop and still operate ok. The application: I’m trying to help out a friend who wants to convert a G-scale model train to use a Blunami DCC system controller operating on a constant 14V bus. The controller is pretty sophisticated but has a strict 4 amp total load limit. He has at least two locomotive units which will be used on the same track at the same time. Each locomotive has a stall current (max power to the DC motors) at or above 6 Amps. This current draw can be sustained almost indefinitely by the locomotives if something is stuck, so I can’t use a solution that only protects for start-up or transient power conditions. Im not certain about all my possible options here but the simplest would be if I can individually limit each engine to pull no more than 4 amps while still providing 12 v or more to the DCC controller being added to each train which pulls power from the 14v source through the tracks. I can’t limit at the power supply since each train can pull power individually, hence the supply must support 8 amps or more in total but each load must limit the individual current to avoid damage to the DCC controllers.
I suppose I may be able to limit the power to the DC motors (2 per train) to 2 amps each so that if they both stall at the same time, the load on the DCC controller is limited to 4 amps. Though I worry that trying to balance the power this way won’t work well with the train as I have no idea what problems will occur if the two wheel sets aren’t sharing the load equally. So I really prefer a source limit into the DCC controller over a load limit to the dc motors for the wheels.

The other issue is space constraints, since this circuit and any heat sinks need to be mounted inside the locomotive housings. So I need a minimum part count solution that doesn’t require a circuit board or a lot of volume.
I looked at threads on this forum for ideas but I didn’t see any solutions that seemed to fit the bill. I’m trying to avoid any solutions including non-resettable fuses or anything which would require opening the locomotive housings for repairs as they are a real bitch to get into. I'm hoping for something better than a self resetting fuse as each time the fuse opens , he would lose connections with the locomotives until the fuse resets and the software may just get stuck in a loop.
Thanks in advance for any suggestions on this.

Are you trying to protect the engines or the controller ?
I'd be very surprised if the controller can't handle over current , what happens if the train de rails and shorts the track. Even my set in the 1960s had a thermal limit to protect the controller in this condition.
If you trying to protect the trains , why ?
If the trains are designed to take 6 amps , you either need to get a controller that can supply the current you need , or use the trains in a way they don't take that amount of current.
Any current limit at the motor , will drop the voltage the motor sees .

 

Are you trying to protect the engines or the controller ?
I'd be very surprised if the controller can't handle over current , what happens if the train de rails and shorts the track. Even my set in the 1960s had a thermal limit to protect the controller in this condition.
If you trying to protect the trains , why ?
If the trains are designed to take 6 amps , you either need to get a controller that can supply the current you need , or use the trains in a way they don't take that amount of current.
Any current limit at the motor , will drop the voltage the motor sees .

This current limiter would be intended to protect the DCC controller from the higher loads pulled by the motors if the train stalls. If one of the cars being pulled derails , that is most likely when something like this would happen in a way that doesn’t short the track supply to one or both engines. I’m not a model rail roader myself, but I doubt you can prevent derails outright. So trying to design for the occasional mishap such that I don’t burn out the controller if/when that happens.
Thanks for the interest.

 

This current limiter would be intended to protect the DCC controller from the higher loads pulled by the motors if the train stalls. If one of the cars being pulled derails , that is most likely when something like this would happen in a way that doesn’t short the track supply to one or both engines. I’m not a model rail roader myself, but I doubt you can prevent derails outright. So trying to design for the occasional mishap such that I don’t burn out the controller if/when that happens.
Thanks for the interest.

Are you certain that this "expencive" controller does not have the current limiter already ?
I've just checked my grandsons system, don't tell him, and his cheap ddc controller can handle over current no problem.

 

I've always been interested in model railroading, but I can't understand why DCC is still prevalent.
If the controller can't handle overcurrent, it's not very sophisticated. Limiting the controllers current without limiting the load will result in voltage collapse. With a 6 amp stall current, a 2 amp limitation will almost certainly promote a stall condition. Sounds to me like the load needs to request capacity from the system manager, who, at the time, may or may not grant, depending on system priorities.

Thanks for the reply, I’m not sure why he didn’t select a DCC controller that has built in protection for overcurrent. He mentioned he looked, but I agree with you a more sophisticated controller would include its own protection. I’ll take a look myself and see if I can find/suggest an alternative for him to consider. I’m not sure what he paid already and perhaps he just doesn’t want to eat the cost, but I imagine a workaround may cost him more. He wants the DCC controller as it provides many features for controlling sounds, lights, smoke and speed / direction control over a two train consist. The way this works is through a smart phone App that controls the train speed through a BLE connection. I can see the attraction to combine the other features but the unit he is hooking this up to wasn’t designed with DCC compatibility up front. Other units exist which do. He already has another DCC compatible engine already that came DCC ready. In short though , I think you and another reply both suggest a load current limit on the motors to force them to shut off while leaving the input alone. That approach makes sense to me. Since each engine has two motors (independent wheel sets each with its own motor), I was suggesting a 2 amp limit to each one separately- 4 amps in total per engine. They both are run from the same output of the DCC controller. So I’m either sharing one 4 amp limiter or balancing the load. Another idea I had last night would be to disconnect one motor set from each engine, assuming each motor is limited to 3 A stall current. I can test this to see. I’m not sure how effective running each train would be with half the working power but at least that might be compatible with the controller he selected. Though it might be more prone to burning out those motors which are working twice as hard. Maybe if he really doesn’t need all that power though, it might be the easy solution.

 

Welcome to AAC.

I think a fast acting resettable fuse might be your only practical option. A device capable of sinking 28W for a sustained period is not going to fit your space constraints.

Maybe the fuse could be placed exclusively inline with the motor's supply so anything else in the locomotive stays online. Or perhaps it could disconnect only one of the two motors to allow for a kind of "limp mode" operation.

I'm not familiar with the controller, but a more sophisticated approach might be to add an additional device based on an MCU that talks to current sensors on-board the locomotives and throttles back the misbehaving train to maintain maximum load.

I like your suggestion of limiting the loads individually. If one wheel set is shorted out due to a derailment; it’s possible the other wheel set wouldn’t be and could still be pulling the unit along which could lead to other issues. However, I see your point and agree limiting load to the motors only makes more sense than limiting input power to the DCC controller. The resettable fuse was my first thought on the subject but I’ve gotten other suggestions to look into more suitable DCC controllers. So I’ll do some more research into that first to see if I can suggest any better options. Thank you!

 

If an actual fast acting circuit breaker is not fast enough, then an electronic one.

 

You might want to look at the DCC-ex approach. For low dollars, an Arduino and motor shield, and well supported open source software, your buddy could be the envy of the town.

 

Below is the LTspice sim of a simple electronic fuse that may work for you:
The shown value of R1 causes it to trip and shut off the current at about 4A.
Removing the power resets the fuse.

The value of C1 can be increased to reduce the effect of transient currents spikes, that may otherwise trip the fuse.

The P-MOSFET can be just about any that has a low enough on-resistance to not significantly reduce the voltage for normal load currents.
If it has a sufficiently low on-resistance, it will require no heat-sink, since it's either fully ON or fully OFF.

The voltage drop during normal operation is small, and equals the load current times the value of R1 plus the P-MOSFET on-resistance.

 

Below is the LTspice sim of a simple electronic fuse that may work for you:
The shown value of R1 causes it to trip and shut off the current at about 4A.
Removing the power resets the fuse.

The voltage drop during normal operation equals the load current times the value of R1 plus the P-MOSFET on-resistance.

View attachment 339771

LIke I suggested in post #9 BUT I did not provide any details because I was in a hurry.

 

LIke I suggested in post #9 BUT I did not provide any details because I was in a hurry.

That's why I posted the circuit, as I'm not.

 

Thanks all for the various input. I Have enough for now. Got some support from the application engineers at Soundtraxx and discussed with my friend to make a plan. I will still have to make a number of changes to lighting and add relays for power management that should all be simple enough. lots of small tweaks still needed. The application engineer suggested a different approach to dealing with high stall currents. What has been used successfully with other customers that used the same brand of G-scale USA locomotives is to simply rewire the parallel motor configuration and convert the two wheel sets into a serial drive configuration. The average short track user typically doesn’t need more than half the full throttle speed which results from serializing the drive system. running the motors in series works fine with the controller and this cuts the stall current in half while keeping the torque needed to drive the load. The track my friend has built has no long straight sections and his 9 car setup will be constantly going in and out of turns , limiting the useful speed range. If he takes the trains to other tracks he’s also fine with the lower top speed. So it is unlikely any additional protection will be required since the DCC controller unit does have sufficient built in protection to prevent damage during the more typical derailment scenarios. It would need protection like that which has been discussed in this thread if keeping the speed range intact was needed using the parallel configuration. There are other controllers out there which can handle these trains but none with the desired BLE and APP based control. This is latest and greatest but perhaps they will offer a higher current capable version down the road. It would be easy to switch out in the future if they do once the rest of the unit conversion is done. Greatly appreciate all the suggestions and especially the spice sim work to validate these fast acting current limiters. I will have to keep this handy just in case plan B becomes necessary. Happy new year all!

 

Below is the LTspice sim of a simple electronic fuse that may work for you:
The shown value of R1 causes it to trip and shut off the current at about 4A.
Removing the power resets the fuse.

The value of C1 can be increased to reduce the effect of transient currents spikes, that may otherwise trip the fuse.

The P-MOSFET can be just about any that has a low enough on-resistance to not significantly reduce the voltage for normal load currents.
If it has a sufficiently low on-resistance, it will require no heat-sink, since it's either fully ON or fully OFF.

The voltage drop during normal operation is small, and equals the load current times the value of R1 plus the P-MOSFET on-resistance.

View attachment 339771

Thank you! This is more than I expected from the community, much appreciated!

 

If the controller did have a higher current output, how long would a stalled train motors' last, before burning out?
What would happen in a 'running ' system if they did?