Arduino Applications and Programs

[PatB]

1 hour ago, melmerby said:

Seems to very similar to the Velleman KA03 but that  uses an L298D.

 

Couple of topics on the Arduino Forum (there are others):

https://forum.arduino.cc/index.php?topic=277150.0

https://forum.arduino.cc/index.php?topic=332397.0

 

Thanks for that. I'd actually found the first but the second was new to me. After a bit of fiddling about I've got the LEDs that indicate output to the motor terminals blinking and fading nicely in response to various sketches. Current issue is, once one motor terminal has faded up and down, to then swap the up/down output to the other terminal to reverse the direction of the loco motor. I can't see it being too hard, but, working in the evenings, I'm a bit tired and slow so it might take a few goes. 

[melmerby]

1 hour ago, PatB said:

Thanks for that. I'd actually found the first but the second was new to me. After a bit of fiddling about I've got the LEDs that indicate output to the motor terminals blinking and fading nicely in response to various sketches. Current issue is, once one motor terminal has faded up and down, to then swap the up/down output to the other terminal to reverse the direction of the loco motor. I can't see it being too hard, but, working in the evenings, I'm a bit tired and slow so it might take a few goes. 

I'm like that all the time!

[Simond]

Updated info can be found at

 

https://www.westernthunder.co.uk/index.php?threads/arduinos-applications-stuff.7856/

 

I’ll update here too in due course.

 

atb

simon

[PatB]

As promised a few posts ago, Here's the sketch for a basic, timed shuttle. You can muck about with the parameters to adjust speed and acceleration (and thus time and distance from end to end), and the waiting time at either end. Sorry about the "goto", which, I gather, is rather disapproved of, but I spent ages shuffling brackets to try and get it to loop indefinitely, to no avail. If anyone can get it to work without (probably very simple, but I appear to be simpler), please feel free to share.

 

Here's a short video of a Hornby J13 shuttling happily back and forth on a temporary benchtop oval.

 

 

[Simond]

Pat

 

I was puzzled by your sketch, and your “goto”.

 

I suggest you remove the “for i” and “for k” loops, as they don’t seem necessary.  If you split your loop into two sections, going up & going down, and write your outputs for direction in each section, and then use the accel / decel loops you have, you don’t need the goto.

 

i have a very tatty version of your sketch (I just used the serial monitor to simulate the outputs) which I could upload if you really want it!

 

atb

Simon

[PatB]

2 hours ago, Simond said:

Pat

 

I was puzzled by your sketch, and your “goto”.

 

I suggest you remove the “for i” and “for k” loops, as they don’t seem necessary.  If you split your loop into two sections, going up & going down, and write your outputs for direction in each section, and then use the accel / decel loops you have, you don’t need the goto.

 

i have a very tatty version of your sketch (I just used the serial monitor to simulate the outputs) which I could upload if you really want it!

 

atb

Simon

Thanks for the tips. Much appreciated. Although I now have a sketch that does what I need, it's been bugging me that it is less tidy than it could be. I'll have a go at what you suggest next time I have a few minutes to sit down and play with the code. Any working example, tatty or not, would be great. Please do upload it. 

 

Once again, thanks for taking an interest. I'm very new to the Arduino, although, many years ago, I did some moderately sophisticated programming in C+ (as it then was) as part of my engineering degree. It's been so long, though, I've no idea if it'll come back. 

[Simond]

Pat,

 

it's like riding a bike...  ('cept it hurts less when it falls over!)

 

I'm at work so just looked at it quickly.  try copying this into your IDE and uploading it, I think it will work  as I've commented it.   I have not used the board you are using to drive the motor, but guess they are all similar.  Presume you write one rail low, and PWM the other high, and then reverse them to go the other way.

 

I would change your program so that the "delay(30)" becomes "delay(delaytime)" and set the value of delaytime in your setup.

 

Then, assuming it works, I'd have a Boolean variable "direction", and call a function "RunTrain(direction)" from the loop.  The function would have the lines that set the pin low, and the loop that PWMs the other pin.  You would need an "if / else " using "direction" to set the correct pin low and PWM the other pin.    You can then use external triggers (buttons, etc, connected to other pins)  to update variables within your loop to set the train off, rather than it always doing exactly the same thing.

 

hth

Simon

 

 

 

 

void setup() {
  pinMode(10, OUTPUT);  //sets PWM pins to output mode
  pinMode(11, OUTPUT);  //pins 10 & 11 correspond to Motor B on shield
}

void loop() {
  start:
  int x = 1;  //Set increments for loop counters
  int y = 1;  //sets increment for forward acceleration step
  int z = 1;  //sets increment for reverse acceleration step
 
//   for (int i = 0; i <= 1; i+x) {                 // DONT NEED THIS

// GOING ONE WAY
    digitalWrite(10, LOW);                      //sets negative rail polarity
    for (int j = 0; j > -1; j = j + y) {
      analogWrite(11, j);                         //sets positive rail PWM value 
      if (j == 255) {
        y = -1;                                            // switch direction at peak  - SURELY SWITCH FROM ACCELERATION TO DECELERATION??
      }
      delay(30);                                         //sets rate of acceleration in forward direction
    }
  delay(5000);                                        //sets wait time at end A of shuttle track  

 

 

//  for (int k = 0; k <= 1; k+x) {              // DONT NEED THIS EITHER

// GOING THE OTHER WAY
    digitalWrite(11, LOW);                     //sets negative rail polarity (opposite rail for reverse)
    for (int m = 0; m > -1; m = m + z) {
      analogWrite(10, m);                       //sets positive rail PWM value (opposite rail for reverse)
      if (m == 255) {
        z = -1;                                                 // switch direction at peak
      }
      delay(30);                        //sets rate of acceleration in reverse direction
    }
  delay(5000);                          //sets wait time at end B of shuttle track
 //    goto start;                           //returns sketch to start as I couldn't get it to loop by other means  // DONT NEED THIS
  }                                               // DONT NEED THIS
}                                                 // DONT NEED THIS
}                                                //  YOU DEFINITELY DO NEED THIS!!! - this is the end of the loop function

Edited June 26, 2019 by Simond

[PatB]

On 26/06/2019 at 15:17, Simond said:

Pat,

 

it's like riding a bike...  ('cept it hurts less when it falls over!)

 

I'm at work so just looked at it quickly.  try copying this into your IDE and uploading it, I think it will work  as I've commented it.   I have not used the board you are using to drive the motor, but guess they are all similar.  Presume you write one rail low, and PWM the other high, and then reverse them to go the other way.

 

I would change your program so that the "delay(30)" becomes "delay(delaytime)" and set the value of delaytime in your setup.

 

Then, assuming it works, I'd have a Boolean variable "direction", and call a function "RunTrain(direction)" from the loop.  The function would have the lines that set the pin low, and the loop that PWMs the other pin.  You would need an "if / else " using "direction" to set the correct pin low and PWM the other pin.    You can then use external triggers (buttons, etc, connected to other pins)  to update variables within your loop to set the train off, rather than it always doing exactly the same thing.

 

hth

Simon

 

 

 

Well whaddya know? Worked like a charm. Thanks so much. I haven't split it into separate functions yet, as I don't yet need to, but it's all looking much tidier. I found, though, that declaring "delaytime" (or, rather, "AccelDelay" to distinguish it from any other delays I might need to use) in setup didn't seem to work. I got "AccelDelay not declared in this scope" from the compiler. Moving it to the top of the loop, along with the acceleration step increments seemed to work though.

 

Currently adding a few lines to use some pins to drive a simple LED 2-aspect signal at each end of the shuttle track, to add a little more visual interest. Everybody likes lights that do stuff ;-).

[Robin2]

3 hours ago, PatB said:

 

 I found, though, that declaring "delaytime" (or, rather, "AccelDelay" to distinguish it from any other delays I might need to use) in setup didn't seem to work.

If you declare a variable within a function it is a local variable which only exists within the function. The simplest solution is to make it a global variable by declaring it before setup(). You can then access it anywhere in the program.

 

...R

Edited June 28, 2019 by Robin2

[Simond]

Thanks Robin, my mistake.

 

Atb

Simon

[Paul4256]

Hi everyone,  I am playing around with a few Arduinos on my DC analogue OO gauge layout, mainly to operate points & decouplers (cheap servos) and home made signals so that the signals can react correctly to point settings and loco presence detection (using TCRT5000 sensors).  I decided to see if I could control the trains from the Arduino code so that the software can know where trains are, their direction & speed to make other stuff on the layout happen, mainly sound files played via cheap MP3 players.  All good fun. When first attempted I got a horrid whine from the loco motors, regardless of speed which stumped me.. until I found a super library & code by Ken Shores called "Trainmotor" on the web that totally eliminates the issue.  Fantastic, and all worked perfectly on a test track.  Transferring to my layout, however, causes all kinds of power fluctuation issues whenever a loco traverses some of my live frog points or an isolation section join. I guess it's mini 'shorts', but I can't think how to fix it. Has anyone else done this and found a solution (apart from tear up & re-lay all the track!!) ??

 

[Robin2]

Please post a link to where you got the TrainMotor code.

 

What Arduino board are you using?

 

Did you try the code prior to TrainMotor on your layout and if so did it exhibit the same problem as the TrainMotor version?

 

What sort of motor driver h-bridge are you using? - post a link to their datasheet. You can get motor driver h-bridges that can withstand short circuits.

 

What power supply (volts and amps) are you using for the track power? Is it separate from the Arduino power?

 

...R

Edited July 14, 2019 by Robin2

[Paul4256]

Arduino is a standard UNO R3, The motor driver is a Deek-Robot L298p off Ebay, can't see a datasheet.... https://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItemVersion&item=291942380591&view=all&tid=1696068034019

 

Code with previous library had same issues but also loud noise from the motor.   Power is from old Hornby controller 12v/1Amp uncontrolled output, UNO powered from 9v battery.  Snipped through the power feed from the L298 to Uno to ensure independance.

 

Problem much less with a newer class 08 Hornby shunter than with the older S15 class so maybe I have issues with this loco?

 

A capacitor across the track feed outputs has helped quite a lot.

 

Library if from Github https://github.com/KensCode/TrainMotor

 

 

 

 

 

[Robin2]

There is an awful lot of code in that library - sorry, but I am too lazy to study it all. I don't think there is as much code in the program I wrote for wireless control of trains on a club layout

 

I don't think there is anything technically unusual in the program. However it writes directly to the Arduino registers rather than using analogWrite() for motor control and it would take a long time to understand the register code well enough to know if there is anything unusual.

 

The L298 is an old-technology and inefficient motor driver chip. It has a large ON resistance that wastes a lot of energy - maybe as much as 2 volts get lost in the motor driver. Also I don't think they take kindly to short circuits. On two separate occasions I brought L298s to club meetings and wet home with one side of the L298 broken  .   I have been using Infineon TLE 5206 h-bridge chips - they are good for up to 5 amps and have short circuit protection.

 

...R

Edited July 17, 2019 by Robin2

[PatB]

I'd suspect that the whine is the audible result of feeding the motor PWM rather than a steady DC voltage. The loco in my project does it too, but not loudly enough to be a problem. There are circuit diagrams on the Web for "converters" to turn PWM into smooth(ish) DC, which might help. My guess is that this is what your output capacitor is doing to some extent. 

[Paul4256]

Thanks for your thoughts.. the Infineon TLE 5206 sounds interesting and worth further investigation.  Robin, any chance of a copy of some code to work with it?  Thanks.

[Robin2]

On 17/07/2019 at 17:20, Paul4256 said:

Thanks for your thoughts.. the Infineon TLE 5206 sounds interesting and worth further investigation.  Robin, any chance of a copy of some code to work with it?  Thanks.

It's a very simple thing to use. There are two inputs one for FWD and one for REV. To go FWD (for example) set the REV pin LOW and analogWrite() to the other pin. Vice versa to go in reverse. This does mean using two PWM pins on the Arduino. And, of course, which is FWD and which is REV depends on how your motor is wired. Be sure to study the datasheet for the part.

 

The code I use is like this

if (direction == 'F') {
    analogWrite(revPin, 0);
    analogWrite(fwdPin, pwmVal);
}
else {
    analogWrite(fwdPin, 0);
    analogWrite(revPin, pwmVal);
}

 

...R

EDIT ... 19 July to add code snippet which I thought I had added earlier

Edited July 19, 2019 by Robin2

[fcwilt]

The Infineon site has this part marked as "not for new design".

Am I looking at the correct part?

https://www.infineon.com/cms/en/product/power/motor-control-ics/intelligent-motor-control-ics/integrated-full-bridge-driver/tle5206-2g/

 

Thanks.

[Robin2]

This is the RS Components link for the part I have used

https://uk.rs-online.com/web/cp/1658151,9062978,9062978P/?pst=TLE52062SAKSA1&sra=p

 

As far as I can see they are still readily available and I don't think the newer Infineon products are available in the same convenient package. Unfortunately the trend is to design everything for surface mounting.

 

...R

PS ... I thought I had included a code snippet in my earlier Reply - i must have omitted to press "save". I have added it now.

Edited July 19, 2019 by Robin2

[Simond]

I wouldn’t worry about the “not for new design” thing.  If you were committing lots of money in product development, you’d want to know if a key component would be discontinued within the product lifecycle, but I’d guess there will be parts around for a few years yet, certainly in model-making quantities.

 

atb

simon

[fcwilt]

16 hours ago, Robin2 said:

As far as I can see they are still readily available and I don't think the newer Infineon products are available in the same convenient package. Unfortunately the trend is to design everything for surface mounting.

 

...R

 

 

Thanks.

 

Just out of curiosity do you know (or know how to find out) what the suggested new part is?

 

Frederick

Edited July 20, 2019 by fcwilt

[Robin2]

4 hours ago, fcwilt said:

Just out of curiosity do you know (or know how to find out) what the suggested new part ?

If I recall correctly it took me a long time to find the TLE 5206 by searching through RS Components' online catalogue. One of the problems is that things are "filed" according the the words used by the manufacturer to describe the part which means that a search term may not find all the items that would actually be suitable. And often finds dozens of things that are unsuitable.

 

All I can suggest is that you begin searching. I did not search the other big suppliers such as Farnell, Digikey or Mouser simply because I can collect stuff from RS.

 

...R

[Paul4256]

Still playing with the L298 motor shield... came across this funny.  If I test my potentiometer directly on the Uno just using +5v, GND & A0 for the reading, I get a lovely smooth progression from 0 to 1023 with 512 ish at the centre of rotation. If I now drop the motor shield on top, connect through to the same pins from  the shield  and rotate the knob it reads from 0 to around 260 at about 95% turn, then jumps to 700/800 then to 1023 at full turn. Most strange and I can't figure out what's going on. Makes no difference if the shield is powered or not.  Code on the Uno is next to nothing....

 

int potPin = 0;    // select the input pin for the potentiometer
int val = 0;       // variable to store the value coming from the sensor
void setup() {
Serial.begin(9600);
}

void loop() {
  val = analogRead(potPin);    // read the value from the sensor
  Serial.println(val);
}

 

 

[Robin2]

10 hours ago, Paul4256 said:

 it reads from 0 to around 260 at about 95% turn, then jumps to 700/800 then to 1023 at full turn.

My suspicion is a bad connection somewhere.

 

Or maybe the motor shield uses A0 for something. Have you tried connecting your pot to a different analogue pin?

 

...R

Edited July 28, 2019 by Robin2

[RobjUK]

10 hours ago, Paul4256 said:

If I test my potentiometer directly on the Uno just using +5v, GND & A0 for the reading, I get a lovely smooth progression from 0 to 1023 with 512 ish at the centre of rotation. If I now drop the motor shield on top, connect through to the same pins from  the shield  and rotate the knob it reads from 0 to around 260 at about 95% turn, then jumps to 700/800 then to 1023 at full turn.

 

That sounds like the shield board has a resistor between the wiper and ground connections you are using; from the figures probably around a third the resistance of your pot.

 

Try disconnecting the pot ground and see if you still get a varying input? If it does, though not very linear, if there is a resistor. If not it should just read full scale all the time.

 

A fixed resistor on the input may be a stray pulldown or it could be intended to allow a two wire variable resistance input, or an active input from an opamp.

 

Edited July 28, 2019 by RobjUK
clarification