02-03-2019, 10:50 AM
Mike, it might be a good idea to fit a higher rated fuse to match your increased overload current limit!
|
Model train controller
|
|
Hi Terry
Yes, thanks, I have not got any fuse in the circuit at the moment, it is still very much a lash together affair across the bench job at the moment. My plans are to try and incorporate the speed adjust/control, together with the 555 PWM from the first circuit with the Mosfet circuit, from the second schematic I posted, to control the output . I will be using a 317T regulator to set the max supply voltage to about 14v, that stabilised supply will supply the variable control circuit and Mosfet output control. I will also wire in the transformers second 12v winding to help improve the transformer regulation. The transformer is an old RS 0 - 12, 0 - 12, claimed to be 2A per winding, but it does get warm after being left on occidentally with no load for a few hours. The rectifier is a 6A bridge and smoothing is 4700uF. I will use a 7805 to set the supply for the 555, if that is to low to drive the Mosfet then I will use another 317T to adjust the voltage for the 555 to the required level. I prefer the 555 circuit in the first diagram, as it describes how to set the mid frequency of the pulses. I have read that Hornby train motors don't like the PWM mid frequency to be too high. thanks Mike
I constructed the circuit in post #14 and can confirm it does work.
However the slow speed control and torque of Hornby type trains seems to be only very slightly better with a PWM controller than when powered by a LM317T type of regulator which can supply a variable voltage from a low impedance source. The drawback with the PWM controller is the mechanical noise generated through the motor due the the pulsing current. It initially was like a high frequency hum, (about 260Hz) and is there all the time even at stationary. It improved slightly and was less noticeable when I changed C5 from 100nF to 47nF. But adding what is termed "inertia Control" (automatic slow start, coasting, and slow stopping) has proved difficult to do using the PWM circuit and is beyond my knowledge and understanding of the workings of the 555 timer chip.. Anyway I went back to experimenting with a linear regulator, LM317T chip, used in a "Slow turn on" circuit as below. In theory this circuit should work with a larger capacitor to give a 30 second slow rise in the output volts, and together with suitable switching and selected resistors to shunt the capacitor the circuit can give the effect of breaking and finally stopping. I found that a 100uF cap in place of C1 gave a slow rise in voltage of about 40 seconds to full output. (full output is set by a 4.7k variable resistor in place of R2) Various resistors were experimented with to give the breaking effect. During testing the setup worked very well with a 300mA bulb, 40 seconds to full voltage, 30 seconds to about half voltage to simulate coasting, and 5 seconds to the final stop at the station. However in practice things were totally different. When powering a train on the track the voltage was all over the place, rising then falling randomly, and even coming to almost a stop. The odd thing though was, if the engine was held still allowing the wheels to spin on the track the controller worked perfectly, the voltage rose steadily to full volts. Release the train and it momentarily sped off but the voltage was soon all over the place again. My conclusion, the slow start circuit in the Texas application notes DOES NOT work with a model train. I dont know why. Possibly because of back emf and connection to the track issues being amplified by the transistor. Role back a day. I had made this circuit on a little piece of Veroboard, and I had thought with out any wiring faults, it worked first time (allowing for the experimentation with capacitors and resistors) We were happily playing with it for most of the day. It was not until later in the day that I discovered that the collector of the PNP transistor was not connected to ground, I had used the wrong copper strip for ground. Oops, why was it working at all, oh well I will rectify that and try again. The capacitor C1 value had to be drastically reduced from 4700uF to 100uF, the breaking resistors all had to be changed to cater for the gain of the transistor. Final circuit. I reverted back to the circuit with the collector disconnected. The engines need about 3 to 4.5 volts to start to move from stationary pulling a load with 4 carriages. The final value for C1 was 4700uF. Shorting C1 with a 2.7k resistor simulates slowing down for points or approaching a station, dropping the voltage to about 6 to 7 volts. A further resistor 270 ohms switched in discharges the capacitor to 3v and simulates final breaking. There is a short video here https://www.youtube.com/watch?v=Y1Ci2T-RVkc What fun Mike
11-03-2019, 06:18 AM
Hi Gents, a good design in ETI many years ago. I used a damped HV generator across the tracks. It seemed to keep the tracks/ wheels low resistance.
Ed
11-03-2019, 11:16 AM
I can remember reading the how it works for that circuit in the day.
It used a high voltage current limited so that it was low enough not to set the traction motor running with a tough capacitor across the tracks to blow off any finger marks and so on. The traction supply was simply diode gated in. There is a web site in the US that holds an archive of most of ETI. There is a link to the archive web site on Paul Stennings site.
11-03-2019, 01:47 PM
Why not include the link here? https://www.americanradiohistory.com/ETI_Magazine.htm
From memory, the way it works is that the near short across the track by the motor damps the oscillator but the instant that the short is removed, the oscillator kicks in with a sufficiently high voltage to create an arc that blasts away the contamination so that normal traction can resume. I've no idea how high the voltage is but I imagine that you have to be careful of putting your hands near the track when it is on!
11-03-2019, 02:25 PM
I would have to find the article and re-read it in order to remember how it worked.
This was the very simple design I used, it gave almost as good slow speed results as the PWM controller but without the annoying noises, and worked better than the original circuit shown in the Texas Instruments Application notes.
Would anyone like to comment on how it works. Mike
13-03-2019, 11:10 AM
I'm not sure why you want a switch between the 1N4148 and the IC, unless it is for shunting movements where precise stopping is essential so I'm assuming that it is permanently closed and that R2 is set to zero at the start of the sequence.
Turn up R2 to the value required for full speed running. The 1N4148 conducts, connecting C1 across R2. C1 commences to charge, dragging down the voltage across R2 because of the initial high charging current. As C2 continues to charge, the voltage across it rises, allowing the train to start and accelerate until it reaches its maximum, 0.7V below that across R2, and no longer has any effect, so that train now runs at full speed. When R4 is brought into play, it starts to discharge C2, which drags down the adjust voltage via the diode, causing the train to slow down to the point at which R4 in series with the diode are in parallel with R2 and the speed stabilises until either R4 is disconnected, allowing C1 to start recharging again and increasing the speed or R5 is brought into circuit. This has the same effect as R4 but, due to its much lower value, has a much faster and more severe effect on speed. I assume the value of R5 has been chosen so that the train doesn't quite stop, allowing R2 to be used to stop it at exactly the right point.
Hi Terry
Thanks, that more or less confirms my understanding. The switch is there just to change from normal variable speed control to auto speed control. Can you confirm if R3 plays any part at all in the operation of the circuit. There will be a 3rd button so there will be a final stop control. Thanks |
| Users browsing this thread: |
| 1 Guest(s) |