Maintenance

Post Reply
Penri
Posts: 1284
Joined: 18/05/2017, 21:28

Maintenance

Post by Penri »

The railway team report that the underground system is not working, although I do not have the details of what the problem is.

They have carried out an investigation and suspect that the issue rests with the Pi.

As and when I have time I’ll see if I can spot some obvious issue.

If anyone reading this has any more information please post it here.

Terry, I’ve posted this to alert you plus if you do have any pointers on what I should look for please let me know.

Penri
TerryJC
Posts: 2616
Joined: 16/05/2017, 17:17

Re: Maintenance

Post by TerryJC »

Penri,

When you look at it, see what the LED on the Pi is doing. It should be on with the occasional flicker. Then remove the power and wait for about 30 s before re-applying the power. The LED should flicker for one or two minutes before steadying with the occasional flicker. If that is Ok, then the the Pi itself is probably OK and the fault is most likely in the motor drive board.

Or the power supply could have failed.
Terry
Geoff
Posts: 17
Joined: 19/02/2019, 14:59

Re: Maintenance

Post by Geoff »

I have checked the power supply and there is a good 12 v supply.

I cannot be certain about the LEDs on the Pi but they seem normal! I can check next Monday.

One peculiarity is when the Pi Initialises the traffic lights work as expected as it runs the warmup sequence. But there is no power to the track. But after this the traffic lights do not respond from the front switch! The train not moving.

The first problem suggests that it could be the motor drive board. I believe that there is a spare one in the store which I should be able to swap and test on Monday. But to me the fact that the traffic lights are not working from the front switch suggests it could be something else?

Geoff
Geoff
Posts: 17
Joined: 19/02/2019, 14:59

Re: Maintenance

Post by Geoff »

This morning I checked to Pi on the underground railway and on initialisation the led was flickering after it stayed on steadily, the same as the chocolate factory Pi.
On removing the Motor Drive board I Found that the orange output wire had broken at the crimp, on fixing this it made no difference. I replaced the Motor drive board with a spare and powered up after initialisation which seemed normal I realised the Motor drive board was excessively HOT. no output power. I turned off immediately.
I disconnected the track output from the main board and checked the underground with a normal controller and it ran normally as expected, no short in the track.
I reconnected the output as normal. I have put Tape round the power supply plug to prevent it being plugged in!
Any suggestions?

Geoff
Penri
Posts: 1284
Joined: 18/05/2017, 21:28

Re: Maintenance

Post by Penri »

Hello

As I understand it the Motor Drive board is a H bridge circuit controlled directly from Pi I/O pins. If the Motor Drive board is getting very hot I would guess that either: the O/Ps are shorted to each other or to something (power rail or ground) or the; or one or other of the I/P control lines is stuck in an On state, which could be either a H/W or S/W issue.

Terry, is the drive to the H bridge PWM?

Penri
TerryJC
Posts: 2616
Joined: 16/05/2017, 17:17

Re: Maintenance

Post by TerryJC »

Right. I discovered this morning that I wasn't subscribed to this particular Topic so didn't respond to it. Sorry about that.

After a phone conversation with Penri, we have agreed that he will remove the Interface Board from the layout and I will test it at home. Once I have tested it and repaired it as necessary, I will return it to the WMT.

Geoff,

The broken wire would probably be the most likely cause of the problem, although I'm not sure why the traffic lights aren't working. I'll keep you posted.

FYI I have had numerous problems with brittle wires breaking in the crimps, so I will probably replace all the wiring with softer equipment wire if that seems to be the issue.
Terry
TerryJC
Posts: 2616
Joined: 16/05/2017, 17:17

Re: Maintenance

Post by TerryJC »

Penri dropped the Interlace Module off a short time ago and I can confirm that:
  1. There is nothing wrong with the Pi; it is running the code correctly.
  2. The DC/DC Converter is producing 5 V as designed.
  3. The L298N Driver chip is getting extremely hot.
I've only had time to do a quick check and will need to remove a few things to establish if there is a problem in the Interface Board itself. I may be wrong, but if the Motor Drive Board is run with its motor terminals open circuit, then no current should flow in the driver devices, so the chip shouldn't get hot. That implies one of two things:
  1. There is a short somewhere in the interfacing circuit (ie on either or both of the motor drive terminals (Pins 5 and 6 of TB1). I've done a quick check between each terminal to 0 V and 15 V and found only resistances in the MegOhms. I will buzz everything out IAW the circuit to establish if there is anything else wrong, but there really isn't much there to short to.
  2. Something on the track is shorting the motor drive terminals and damaging the chip. That is possible if the rest of the interfacing circuit is OK, but I'll just have to check it all very carefully before I replace the Motor Drive Board a second time.
I won't get through this today, but should be able to spend more time on it tomorrow.
Terry
TerryJC
Posts: 2616
Joined: 16/05/2017, 17:17

Re: Maintenance

Post by TerryJC »

I'm fairy certain that the Interfacing Board is operating as it should (apart from the short circuited Motor Drive Board). If I remove the input wiring from GPIO Pins 17, 27 and 18 (Enable Out, Enable Back and PWM), but leave the power going to the Motor Drive Board, the heat sink becomes so hot that it is untouchable. If I remove the power, then it doesn't as you would expect.

I have put a scope on the three GPIO Pins and the lines are moving as they should:
  • Enable Out goes high, the PWM ramps up.
  • PWM ramps down for the first stop.
  • PWM ramps up.
  • PWM ramps down for the stop at the Main Door End.
  • Enable Back goes high, the PWM ramps up.
  • PWM ramps down for the first stop.
  • PWM ramps up.
  • PWM ramps down for the stop at the Store End.
There really isn't anything else connected to the Motor Drive Board so I will replace it in due course and test it against the test circuit in the Documentation. I have torn that down now so will have to spend a bit of time finding all the parts and assembling it again. This probably won't happen for a day or two because it's my son's birthday today and I have appointments tomorrow.

If anyone can see any flaws in my logic, pleas shout before I blow up yet another Motor Drive Board. :cry:

In the meantime if someone could check for shorts on the track (track to track, track to 0 V and track to 15 V) it would eliminate another problem.
Terry
Penri
Posts: 1284
Joined: 18/05/2017, 21:28

Re: Maintenance

Post by Penri »

Terry

Analysis seems sound to me.

The LM298N is quite a simple and robust devices with not a lot to go wrong within it without some external assistance. If there's a chance of a repeat of the problem it may be worth considering some current limiting.

Penri
TerryJC
Posts: 2616
Joined: 16/05/2017, 17:17

Re: Maintenance

Post by TerryJC »

There is always a chance of a short circuit in this kind of application. It occurred to me that the train motor might be a good suspect.

Passive limiting won't work because to be effective the volt-drop on a series limiter would drop too many volts and the train wouldn't run.

That means some kind of active limiting circuit; either a simple transistor device or a dedicated current limiting IC. The problem is going to be where to put it because there is limited space on the Veroboard, so it might be that a piece of Veroboard outside of the Interface Board may be easier. This could go between the PSU and the power input or between the motor drive terminals and the track.

What would you do?
Terry
Post Reply