Author Topic: Train destination display - up and running  (Read 9131 times)

0 Members and 1 Guest are viewing this topic.

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Train destination display - up and running
« on: December 10, 2015, 09:57:40 pm »
Hello everybody,

I'm new to this forum and I'm sorry to be looking for help here already with my first post. I'm not sure this is the right place, but I really don't think there's many other sites out there that gather as many electronics enthusiasts and professionals as the EEVBlog does :)

I'm not really a newbie when it comes to electronics, soldering and stuff... I've been handling soldering equipment regularly for the past twenty years or so. But still, my experience with circuitry is pretty much limited to digital stuff, where there's either voltage or no voltage... I don't have much practical experience with analog circuits, which I believe is what I'm lacking in the given case.

What I have sitting on top of my bench right now is the drive motor assembly of a somewhat older train destination display that used to be installed on pretty much every station on the local subway / tube / underground (I'm not sure what term to use... we call it "U-Bahn" here) network here in Munich. I remember them myself until a few years ago, when they were successively replace with newer, modern LCD displays. The design of this unit might date back into the 1980s or even 70s (1971 was the year when the subway started operation in Munich; however I'm not sure if these displays were already in place back then).

The challenge, of course, is to get it working again in some kind of way... and that's where I'm lost. I can't really get the circuit at hand to work in any useful way, and my hope is that somebody here will recognize some pattern, or derive in any other way how to drive the inputs in order to make it work the way it's supposed to.

I'll try to provide as much detail as I can about the circuit, but first: here's a short video of somebody who claims to have repaired one, just so you get an idea of what I'm talking about:



In the comments there's a reference to a website (that is gone by now) which was supposed to contain more information about how he did it - unfortunately there was not much information there, and we never managed to get in touch with the uploader.

Speaking of "we": I got this drive unit from a friend who is an active member of a local society called "Münchner U-Bahnfreunde" (German website: http://www.muenchnerubahnfreunde.de/). They are striving to collect historical data, photos and items, and they also feature displays at local fairs, where they show off their collection. They bought a few of these old train destination units, and they are already putting them on display regularly, even though non-functional. They asked me to try and "resurrect" the old circuitry inside. The basic idea here is to combine the old electronics with something like an Arduino and / or Raspberry Pi (or whatever it takes) and let the fair attendants control the display in some way.

------------


Alright... now for the stuff I know so far. I'll attach some pictures to clarify things, due to the file limitations I have placed the complete imageset here: http://www.justjakob.de/zza/. You can also find the Eagle schematic there. If there's any questions go ahead and ask, I will of course try to answer them :)

The unit consists of a rather chunky 3-phase motor, a set of gears with some magnets on them, and a PCB that has a 10pin connector for an input. Whatever once was connected to that input has been lost, but from what I'm told I suppose it wouldn't make things much simpler.

I have tried to reverse-engineer the PCB and look up the components on it, I have attached a screenshot of what I worked out. Also check out the image link above for the Eagle file.

Here's the parts list and what I know about each of the components:

MOTOR: "Berger Lahr" RSM63/12 - That's the primary drive motor, a 3-phase synchronous motor. the closest kind of documentation I could come up with is pages 101 and 102 of this document: http://www.schneider-electric.com/download/ww/en/details/2081705-Synchronous-motors-RSM/?showAsIframe=false&reference=1148964249
C2 & C3: appear to be motor caps that allow the 3-phase motor to run on single-phase AC. I don't have their exact capacities, but I remember that they somewhat matched up the values in the motor datasheet.
C1: I'm not sure if that's a capacitor at all. It's marking reads "104 M3B" and I couldn't come up with any information about it. Might as well be some kind of fuse / MOV.
T1: marked "L201E3", appears to be a TRIAC to drive the motor AC. http://www.alldatasheet.com/datasheet-pdf/pdf/114977/TECCOR/L201E3.html
T2: marked "MPSA92", seemingly a PNP transistor. http://www.onsemi.com/pub_link/Collateral/MPSA92-D.PDF
T3: marked "MPSA42", seemingly a NPN transistor. http://www.farnell.com/datasheets/1747986.pdf
IC1 & IC2: marked "03F024", can't find any information on them, but I guess they are some sort of hall effect sensors. They are mounted on the copper side of the PCB, right underneath a big wheel that sits right on the motor hub, that has permanent magnets sitting on it.

There is also another component that I forgot on the schematic. It's marked "S14K 60" and sits parallel to C2 & C3. It seems to be pretty clearly a MOV that - if I understand correctly - is for protection and probably does not influence the regular function... correct me if I'm wrong here.

My guesses about the magnets and hall effect sensors is that they provide positioning feedback to the controlling circuit. The math adds up: There's four magnets on the outer diameter (the one that IC1 picks up) on the motor wheel, the gear train reduces the motor movement by 1:10, and there's a total of 40 display "sheets" on the display hub, so each magnet passing would mean that the display has flicked by one sheet.

There's one more magnet on the inner diameter of the motor wheel that mechanically interacts with the 1:10 reduced wheel on the gear train: It is pushed down (near the hall effect sensor IC2) only on full turns of the reduced wheel, which would correspond to a full run-through of the display sheets. That seems to be some kind of calibration input for when the unit is powered up since the output of IC2 is routed directly to the 10pin connector and has no other connection to the rest of the circuit.

I tried to measure the output of IC2 with a multimeter (I don't own a scope, unfortunately) and it seems like it is actually pulled to ground when the magnet comes close to it.

From what is written in the copper, it seems the circuit used to run on two different voltages: 42VAC for the motor and 15VDC for the rest. I have tried to get some life out of it and could get the motor to actually spin when applying the voltages directly to the TRIAC, so I believe there should be some way to get it to work.

The 42VAC most certainly go into pins 2 and 4 of the input, at least that's what makes most sense to me. That way T1 can be triggered to close the actual motor circuit. The big question is: How do I trigger T1?

I don't know exactly where the DC is to be fed into, for me the most obvious combination would be pin 6 to be positive and pin 1 to be ground (derived from the guessed power supply lines of the hall effect sensors). Might of course be wrong here.

I have no real idea how to drive the remaining pins, unfortunately.

I can see that there's some interaction between the hall effect sensor IC1 and the transistors, and I'm really not sure how it's supposed to work. My guess would be that you can drive this circuit in a way that stops automatically at the next pass of the magnet (which would mean at the next sheet of the display) but I can't really "see" that behaviour in the circuit. Also, my guess could as well completely wrong :(

I'm also guessing that there is some kind of way to read the magnets passing by IC1 externally, since it would be kind of hard to stop the motor at the right time otherwise. With all the passives attached to IC1, I'm not sure how that works either.


------------

I guess that's all the detail I have at this point. Sorry for the long post, I just wanted to make sure that there's nothing missing in the puzzle. If I have overlooked anything, please let me know, I'll try to fill it in.

One more thing: Yes, it would be feasible to design my own PCB and attach the motor onto that, and I have spent some time thinking about that, too. I consider this as a "plan B", given the historical background of this project I'd rather keep as much of the original components, of course.

I really appreciate any help I can get with this project, thank you all for reading up to this point, and thanks in advance for any input :)
« Last Edit: December 17, 2015, 01:53:05 am by jketterl »
 

Offline station240

  • Supporter
  • ****
  • Posts: 859
  • Country: au
Re: Train destination display - looking for help
« Reply #1 on: December 11, 2015, 12:48:02 pm »
What an odd circuit.

C2 is a run capacitor, C3 is likely just noise suppression.
T1 is indeed a Triac, T2 & T3 are drivers for it.

IC1 and IC2 are hall effect sensors, which output a digital pulse when a magnet passes over.
Seems to be setup that it rotates the motor once, triggers IC2. An external circuit then takes the signal from pin 8 of SV1, it then sends a pulse to pin 10 to make the motor to rotate enough to pass over IC1. the motor then continues to rotate a full turn.

* note full turn means far enough to advance the destination, likely this is 90 degrees.

try shorting pins 8 and 10 together (with a push button switch), it should rotate the destinations continuously.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #2 on: December 13, 2015, 02:15:16 pm »
Thank you very much for your reply :)

unfortunately... I still can't get any movement out of it. I connected the circuit up with the AC on pins 2 and 4, 5VDC across pins 1 and 6, with AC pin 4 connected to DC GND (from what I know about triacs, this is necessary, but might as well be connected to VCC... don't know what's more likely) and tried connecting pins 10 and 8 to each other. Since I don't know the exact behaviour of the hall effect sensors (they might require a pull-up), I also tried a straight connection from VCC to pin 10 - that didn't work either :(

Still looking for help here...
 

Offline alho

  • Contributor
  • Posts: 37
Re: Train destination display - looking for help
« Reply #3 on: December 13, 2015, 09:21:16 pm »
 I don't remember how triacs work  :-[ so don't take my word for this.

Pin 1 is DC ground and Pin 6 is 5VDC

Pins 2 and 4 are 42V AC

Pin 8 is output from IC2

Pins 3 and 9 need to be connected to something or T2 and T3 will do nothing. To What?

Which leaves pins 5,  7 and 10.

Pin 7 is either input to simulate pulse coming from IC1 to start the circuit or output to get trigger from IC1 when magnet passes over.
Unfortunately Pin 10 could be for same purposes.

Pin 5. For setting  DC voltage on T3 base? Some kind speed control (hysteresis on RC circuit)? And again pin 10 could do same.   


On the video posted motor keeps running after the destination sings stop. So more info on the mechanics might be needed.
Could the motor change direction after it flips trough all destinations or does it loop for ever?

And files in  http://www.justjakob.de/zza/ gives 403 Forbidden to me.

 
« Last Edit: December 13, 2015, 09:26:32 pm by alho »
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #4 on: December 13, 2015, 10:20:08 pm »
Well, as far as i know... you put a voltage across the gate and the A1 terminal (that's the one on pin 4) to trigger; the triac will then conduct as long as there's a current flow across A1 and A2 (here: until the end of the AC phase... so basically at 50Hz the motor will stop within 1/100th of a second after the gate voltage is gone). The problem is, this should work with both negative and positive voltages... so do I connect pin 4 to GND and apply a positive voltage to the gate, or should i connect pin 4 to VCC and try to make the gate negative? :-//

Speed control? I don't really think so, since the motor speed is pretty much given by the AC frequency. I don't think there's some re-modulation going on.

As for the video: I believe the sound you hear after the display stops is the "rear side" of the display working. You can't see it in the video, but those units were two-sided, so that they could be viewed from either end of the platform. I'm not sure if they were individually controllable (it would not really make sense to display two sets of information on one platform) but I'm guessing that both sides had their individual set of motors and drive assemblies. It seems that whatever modification was done to that display in the video allows the two sides to be controlled individually.

Also, the top and the bottom row of the display would usually not stop spinning at the exact same time. One more clue that makes me believe that the display in the video is not fully controllable (i.e. you can tell it the exact slide you want it to show) but just allows a basic movement of the motor, leading to a random outcome...

The motor direction cannot be reversed for two reasons: a) you would need to put the motor cap(s) and power supply on different terminals of the motor, and b) there's a mechanical block on the main motor wheel that engages with the gear teeth and prevents it from spinning backwards.

The pictures should be fixed now, sorry for that. I failed to set the file permissions properly |O
 

Offline SL4P

  • Super Contributor
  • ***
  • Posts: 2074
  • Country: au
  • There's more value if you figure it out yourself!
Re: Train destination display - looking for help
« Reply #5 on: December 13, 2015, 10:36:29 pm »
I took a look at your schematic, and it could help you (and others) figure it out - if you untangle the diagram, now that you've chased the PCB and wiring...
e.g the hall-effect sensors are 'inputs', and the motor is an 'output'...
so it can help to visualise/understand the flow if redrawn from left to right, and ideally with +voltages at the top, and ground/-voltages below...
Don't ask a question if you aren't willing to listen to the answer.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9590
  • Country: us
Re: Train destination display - looking for help
« Reply #6 on: December 13, 2015, 11:52:07 pm »
I cannot provide detailed help without spending some time looking through all of the pictures and details, but in English these displays are called split flap displays, or just flap displays. According to Wikipedia two common manufacturers were Solari (Italy) or Progotron (Czech Republic). This may help with your searches. Split flap display turns up many hits on YouTube with people who have experimented with them or made them.

I believe the basic principle of operation is that the display will move through one display message for each pulse received. So to bring up a particular message or destination it is necessary to send an appropriate number of pulses. I think there is most likely a "home" setting to return the sign to the beginning of the sequence. To make the sign work you will need to provide it with a source of power of the kind that it needs, and then work out how the control signals work in order to make it switch.

I'm afraid that without technical documentation from the manufacturer it will take some time to reverse engineer the operating principles. I think it would be worth trying to find the name of the manufacturer written on it somewhere and then contact them for information about how to make it work.
I'm not an EE--what am I doing here?
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #7 on: December 14, 2015, 12:45:40 am »
I tried my best to clean up the circuit... I replaced the supply lines with symbols and used a bus for the 42VAC. I'm still sticking with VCC since I'm still unsure whether 5VDC will do the job. I tried to get rid of as many crossings as I could, didn't work out that well. I still hope it's an improvement.

Thanks for the information about the other manufacturers. The PCB I have at hand was manufactured by a company called "Krone" - the exact type is "6280 3 160-00 K1/9". I forgot that in the original post, but I had no luck on google with that anyway. I'll still try the other manufacturers, maybe that will turn up something similiar.

 

Offline alho

  • Contributor
  • Posts: 37
Re: Train destination display - looking for help
« Reply #8 on: December 14, 2015, 01:17:31 am »
Speed control? I don't really think so, since the motor speed is pretty much given by the AC frequency. I don't think there's some re-modulation going on.

I was thinking that adjusting DC voltage at base of T3 would delay it turning on and cut part of half wave from going to motor reducing power.

Yeah, AC motors and controlling them is not my thing.

Heres my attempt to untangle schematic.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9590
  • Country: us
Re: Train destination display - looking for help
« Reply #9 on: December 14, 2015, 02:16:38 am »
Thanks for the information about the other manufacturers. The PCB I have at hand was manufactured by a company called "Krone" - the exact type is "6280 3 160-00 K1/9".

Note that the PCB is just a component of the complete unit that might have been contracted out. Krone didn't necessarily make or supply the finished display. You would be looking for a name plate or a rating plate on the outside of the casing somewhere.
I'm not an EE--what am I doing here?
 

Offline SL4P

  • Super Contributor
  • ***
  • Posts: 2074
  • Country: au
  • There's more value if you figure it out yourself!
Re: Train destination display - looking for help
« Reply #10 on: December 14, 2015, 07:43:17 am »
cheers, the schematic looks a lot easier to make sense of!
I'm curious if R6 really does indeed go to GND... small style tweaks can be improved, but heading the right way to support analysis and documentation!!

I you've reached your limits, it may help to look for a local Hackerspace, and a semi-professional tinkerer with experience developing on microcontrollers - 'even' Arduino will handle this type of application easily.
I'd love to look at it, but I'm not exactly local to work with you!

Id guess with a keen hobbyist, you'll have the flaps turning within a day or two, and the sign 'sort of' working in a few more days (assuming the motor and flap solenoids are functional!)
Don't ask a question if you aren't willing to listen to the answer.
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1831
  • Country: us
Re: Train destination display - looking for help
« Reply #11 on: December 14, 2015, 08:27:05 am »
Pin 10 going low, dry contact or open collector, will start the motor. Pulse low should be about 95% of the time it takes to get to the next message.  The capacitor connected to IC1 keeps it low as it slowly charges up.  When IC1 is in position the output goes high pulling pin 1o high stopping the motor.   All this would free up the controller and allow speed variations due to wear ot temperature  not to effect signage.  A little UNO like controller would be just the thing to make a neat display.  Use random number generator to vary display times.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #12 on: December 14, 2015, 03:28:08 pm »
Wow, there's been a lot of input in those last few posts :) Cheers everybody :)

Let me give a quick reply to a few, I will work some more on it later.

I was thinking that adjusting DC voltage at base of T3 would delay it turning on and cut part of half wave from going to motor reducing power.

Ok I understand now where you were going, just like the "simple" dimmer circuits for good old lightbulbs. I still believe it won't work with a motor, especially not with a 3-phase one. It would be a nice feature to have it spin up to speed and brake it accordingly, yes, but it doesn't seem like the guys who designed this back in the days were going for something that sophisticated ;D

Heres my attempt to untangle schematic.

I see what you did there. I really like it :) even I start getting an idea of what it might do now, at least the transistors :)

It seems like the circuit was designed to sink current from the triac's gate via T2, which in turn is activated by T3. That makes me conclude that the A1 terminal of the triac needs to be connected to VCC instead of GND in order to run. Please correct me here if necessary.

It would also make sense to just connect (3) and (9) to ground. At least I cannot make any much more out of those at this point...

Note that the PCB is just a component of the complete unit that might have been contracted out. Krone didn't necessarily make or supply the finished display. You would be looking for a name plate or a rating plate on the outside of the casing somewhere.

I don't have the complete unit to check that, but I'll see if I can get somebody to find that out.

I have been googling extensively last night, and I did manage to get at least some information. It seems like there once was something called the Krone System 6000 (which this might belong to). In particular, I found this post: http://unknowndomain.co.uk/2013/07/11/a-visit-to-brighton/ that shows one very old installation in the UK which hase some very similiar drive assemblies. Also, in the pictures you can see some kind of ICs labeled "6412 ..." with matching 10pin-connectors on their PCB. That might be some of the original controlling circuitry. Unfortunately, not very much technical detail, and nothing that would catch with google.

I also found a reference to an old book from 1980 that is of course out of stock, and I don't have much hope of getting my hand on that.

Most of the technical documentation I found (for all the companies I tried) usually talks of some kind of bus control (mostly serial, in some cases different... but always digital) to address these displays. While that makes of course sense, I'm afraid that the digital part of the system at my hands is gone and needs some kind of replacement, so that kind of documentation is unfortunately useless as well.

I'm curious if R6 really does indeed go to GND... small style tweaks can be improved, but heading the right way to support analysis and documentation!!

I will check that. I guess the best way to make sure that the circuit is correct would be to re-route the board in eagle and see if the layout matches. I might eventually do that, but will have to create the specific components first.

I you've reached your limits, it may help to look for a local Hackerspace, and a semi-professional tinkerer with experience developing on microcontrollers - 'even' Arduino will handle this type of application easily.
I'd love to look at it, but I'm not exactly local to work with you!

Yes, we have thought about contacting local hackerspaces, but they are kind of hard to find. Not sure if it's just because you have to make "first contact" first or if the local people here really are not that much into tinkering and hacking. Thanks for the offer though :)
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #13 on: December 14, 2015, 04:31:58 pm »
Pin 10 going low, dry contact or open collector, will start the motor. Pulse low should be about 95% of the time it takes to get to the next message.  The capacitor connected to IC1 keeps it low as it slowly charges up.  When IC1 is in position the output goes high pulling pin 1o high stopping the motor.   All this would free up the controller and allow speed variations due to wear ot temperature  not to effect signage.  A little UNO like controller would be just the thing to make a neat display.  Use random number generator to vary display times.

Would you please elaborate a little on this? It's just that bringing pin 10 to ground did nothing for me yet, the reason being most likely that the remaining pins were not properly setup. I believe pins 3 and 9 are especially important, as well as the bias of the triac.

Also, I'm not sure about this, but from my measurements both IC1 and IC2 will be low when the magnet passes over them. Can't be sure about that without a datasheet, though.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #14 on: December 14, 2015, 04:34:27 pm »
I had another go with the eagle schematic (mostly for learning purposes). Here's the result.

I'm curious if R6 really does indeed go to GND...

I just checked... Yes, one end is indeed connected to GND - at least as long as the assumption that pin 1 is GND holds.
 

Offline SL4P

  • Super Contributor
  • ***
  • Posts: 2074
  • Country: au
  • There's more value if you figure it out yourself!
Re: Train destination display - looking for help
« Reply #15 on: December 14, 2015, 09:05:53 pm »
ok, a big jump here...
the details we have been discussing so far are for the motor only...
the trial turns the motor on or if. no speed control.
the sensors tell the controller where the drive shaft is in its rotation.
This all seems fairly complete.

EDIT: From here in I'm assuming there is only one motor assembly as shown for the whole sign panel...?  If there are multiple motors, then it gets easier, as the controllers are only handling individual flap drums.

What appears to be missing are (solenoids?) that act as (a brake?) on each flap wheel...
My guess being that the motor starts under no load, with a simple slip clutch rubbing against each flap drum.
As soon as the drive shaft has completed one or more revolutions past the index position sensor (i.e. up to speed), the controller releases the brake on those flap drums it needs to change. (one tile one each flap set has a index flag so the controller knows when it is 'home').

The rest is simply counting pulses, and setting or releasing he flap brakes.
Probably not the most elegant method, but low parts count, and easy for non electronics techs to grasp and repair.

My guess is based on watching these things work in the past, and the typical reset/refresh cycle they would do when changing a message.  To reset the board to a known state (home)...
Start the motor, release 'all' the flap brakes... then reapply each flap brake as the home flap index tab comes around.
it seems to me there should be more bits & pieces to control the individual flap wheels...?  Or at least a hint to how they were individually started/stopped.  This display is still a sort of electromechanical XY display from a drive perspective.
« Last Edit: December 14, 2015, 09:43:46 pm by SL4P »
Don't ask a question if you aren't willing to listen to the answer.
 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9590
  • Country: us
Re: Train destination display - looking for help
« Reply #16 on: December 14, 2015, 09:28:39 pm »
Did you discover this web page before? If not, there is someone there you might reach out to for information:

http://www.jave.de/blog2/?p=111

Unfortunately it seems like the display he had was lacking sufficient parts and information to operate it in the original way. Therefore he created a new circuit to drive the display.

Also someone here reverse engineered a Solari one:

http://rodyne.com/wp-content/uploads/2014/04/solari.pdf
« Last Edit: December 14, 2015, 09:34:40 pm by IanB »
I'm not an EE--what am I doing here?
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #17 on: December 15, 2015, 01:57:15 pm »
Alright, first of all, let me bring the good news: I felt courageous last night and tried to wire the board up, and for the first time it gave me some kind of reproducible behaviour.

I took the 42VAC to (2) and (4), and 5VDC to (1) and (6), as usual. In addition, I hooked +5V to (4) to give the triac a positive bias, and (3) and (9) to GND since it made sense to me. I started to pulse (5) by hand and the motor started to twitch when I applied positive voltage. The motor movement however was pretty unpredictable, so I switched over to an Arduino.

I found that I could make the motor spin in it's quarter-turns kind of reliably, when applying a HIGH level for a certain amount of time (the motor does not spin during that period yet). If i then switch the output off (as in: put it into high-impedance INPUT mode) the motor starts turning and will stop once the next magnet passes over IC1.

Here's a quick video, showing the maximum speed I could achieve with that method:



So something indeed does charge while applying voltage on (5) (I'm guessing it's C5) and then activates the rest of the circuit including the motor, while it discharges.

There's a few points that I have to work on now...

1) Reliability: sometimes, the motor seems to be kind of sluggish to start, and then does not complete the quarter turn as expected before whatever charge is remaining is depleted. I also noticed that when I removed the wheel with the magnets on it, the motor would still continue to make kind of predictable quarter turns. So I'm guessing the capacity of whatever gets charged is matched pretty tightly against the motor speed. It might also be that the 5V I'm feeding in is indeed too low and that I'd have to go up to 9V or even 12V to get more charge into the circuit, and such more potential turn time for the motor.

I also noted that the motor started to behave better over time, I have no idea how old this unit is, and if the motor on it would be fit to be installed under the original conditions. Maybe some WD-40 will work, too.

2) Speed: as I said, what you see in the video is displaying the max speed after tuning delays to a minimum, without losing the reliability. It seems much slower than in the video I posted before (which matches my own memory of how fast these displays were). I might achieve some more by using a transistor to charge (5) since arduino outputs are very limited when it comes to delivering current (I am actually still scared of frying mine). I might as well need to switch to (10) for charge, but (10) seemed more unreliable while fiddling around.

3) Sweet spot: There seems to be a sweet spot when applying voltage to (5) or (10) that makes the motor turn continuously. I found that by accident when I was trying to put an LED on to see when the arduino output was active. I somehow ended up with both LED and its resistor in series to (10) or something like that. I'm guessing the voltage drop of the two diodes combined triggered something. I will have to investigate this, it might help speeding the display up.

From here in I'm assuming there is only one motor assembly as shown for the whole sign panel...?  If there are multiple motors, then it gets easier, as the controllers are only handling individual flap drums.

What appears to be missing are (solenoids?) that act as (a brake?) on each flap wheel...

No, it's not like that. There is one of these assemblies on each flap drum. In the video I posted originally I'm assuming a total of four motors (one for each of the visible flap displays, plus two more for the same thing on the rears side), at least two (can't really be sure about the rear side).

There is no solenoids and / or clutches. As soon as the motor turns, the flaps will start coming down. And from my mechanical deductions, that's one flap for each pass on IC1.

The homing is done via IC2, since it doesn't get engaged on every turn, but only when the drum reaches its home position (i.e. every ten turns). That part should be easy, since there's no additional circuitry on that. I'm guessing a simple pull-up on pin 8 routed to an arduino input should do the job.

Did you discover this web page before? If not, there is someone there you might reach out to for information:

http://www.jave.de/blog2/?p=111

Unfortunately it seems like the display he had was lacking sufficient parts and information to operate it in the original way. Therefore he created a new circuit to drive the display.

Also someone here reverse engineered a Solari one:

http://rodyne.com/wp-content/uploads/2014/04/solari.pdf


Yes, I've seen them both (the PDF on the latter is nice though. I only came up with a dead link and was lucky to be able to ressurect that information through archive.org).

The first guy pretty much bypassed the electronics, which is a shame, since his units look very similiar to the one I have here, mine does not have a processor though. Rewiring the board is still an option for me, even though I don't want to go down that route yet. I'd only need to access the triac gate and the two hall effect sensors directly, wire it up to an arduino and I guess it would work.

The latter is unfortunately different from the electronics perspective, his circuit looks very much different from mine. What I share with him though is the wish to retrofit a new external controller circuit to make it work again :)

----------

Right now, I wish I had a scope to probe around. That Siglent one looked very interesting. Santa? ;D

I found this website: http://www.falstad.com/circuit/ that lets you design circuits and gives some nice visual representation of voltage and current. I will try to get the schematic into there, maybe that will help.

As always, thank you all for your input, I'm getting a lot of fun out of this :) Cheers :)
« Last Edit: December 15, 2015, 01:59:18 pm by jketterl »
 

Offline SL4P

  • Super Contributor
  • ***
  • Posts: 2074
  • Country: au
  • There's more value if you figure it out yourself!
Re: Train destination display - looking for help
« Reply #18 on: December 15, 2015, 02:10:32 pm »
aaah ok,  one motor per flap drum makes the control logic simpler.
you are certainly well on the way to getting this up!
My thoughts were based on those huge flight info boards, where motor per character seems like a difficult way to do it!

Now, it's mostly a hunt for optimum voltage and current for fast start and speed / reliability.

Suggestion... WD40 is a water displacement fluid - not technically a lubricant.  Use a drop light (sewing machine) oil instead!
Don't ask a question if you aren't willing to listen to the answer.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #19 on: December 15, 2015, 02:25:48 pm »
I was actually joking on the WD40 part ^^ I'm not even sure if the motor is serviceable. As far as the whole project goes... It seems that I'm actually woking on a spare, and the guys have some more parts on their hands, so I'd probably go for a replacement of sorts. I'll have to see about that.

A little update on my sweet spot theory: I just hooked up a potentiometer and measured that when applying somewhere between 1.0 and 4.7 V to (10) the motor will spin continuously. The same goes for (5), the voltage range is slightly different, about 0.9 to 4.4V. Should be easy to achieve with a set of resistors.

That will come in handy for homing the display. If i find a suitable signal to get IC2 pulses, I might even be able to jump to the right slide at full speed.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #20 on: December 16, 2015, 12:06:36 am »
OK folks I'm really happy right now because I think we've done it  ^-^

After discovering the sweet spot I really only needed to find out how to measure the right amount of steps the motor has passed... and so, in lack of a DSO I tried out something I had read about some time ago... measuring analog voltages - like a scope - with an Arduino. It's not exactly state of the art, but it couldn't hurt... and I guess it was the best option at hand :) (Yes, I did start an inquiry among friends if somebody has a scope ;D but that takes time)

Poking around with that I found... well, nothing on the rest of the pins. I could measure a slight voltage drop on (5) when driving the circuit through (10) and vice versa, but not really sufficient to be picked up by an Arduino GPIO. So I figured and tried an in-line resistor. I don't know how big the range would be, and how lucky i was... I picked up the first resistor on my bench, which was a 10k one, and put it inbetween the Arduino drive pin and (10). This resistor brought the drive voltage down into the sweet spot window... and allowed me to measure the voltage drop of IC1 very clearly :)

So i hooked the other end of the resistor to a digital input and started to write some code, assuming that the arduino would - when reading continuously - pick up a LOW value. And it did :)

So all was left at this point was to hook up the homing sensor and write some more code.

I think the video shows pretty good how well it works now :)



Once again: THANK YOU ALL SO MUCH :)

Next, we well try to bring this together with a real display drum, and see how it works. I'll keep this updated when it happens :)

8)
 

Offline SL4P

  • Super Contributor
  • ***
  • Posts: 2074
  • Country: au
  • There's more value if you figure it out yourself!
Re: Train destination display - looking for help
« Reply #21 on: December 16, 2015, 01:51:33 am »
Fantastic!
Sorry I distracted you with the single motor path!?!!
Don't ask a question if you aren't willing to listen to the answer.
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - looking for help
« Reply #22 on: December 16, 2015, 01:17:47 pm »
Fantastic!
Sorry I distracted you with the single motor path!?!!

I'm not Australian, but I guess this is the right time to use the term: no worries ;)
 

Offline jketterl

  • Contributor
  • Posts: 14
  • Country: de
Re: Train destination display - up and running
« Reply #23 on: December 17, 2015, 02:09:00 am »
Just a quick update here, for those who are following this.

One of the guys came by my place today and dropped off two assembled displays, a full-length and a half-length one (basically the two units that are installed on one side of the complete box).

We mounted the motor unit to the smaller one and did fiddle around with it, and it did work, even though it had some reliability / repeatability issues. After some tweaking of the code and a realignment of the driving gears, it seems to work pretty well :)

Right now the control is very rudimentary, it's just the Arduino serial console where you can type in the number of the slide you want to display, but this might very well be the base for some kind of user interface already. We are currently debating / evaluating the best solution, it's pretty much determined that we want to go wireless with some kind of website, so basically we could add a Raspberry Pi, an Arduino Yùn or a Particle Photon. But that shouldn't be a big issue any more now, I've already built similiar things before :)

I've also changed the thread title since this has pretty much evolved into a successful repair now :)

As usual, a quick video for everybody's enjoyment :)



Next steps involve ordering some materials to get the circuit off the breadboard and onto - most likely - an Arduino shield, so the next update will probably take a few days.
« Last Edit: December 17, 2015, 02:12:49 am by jketterl »
 

Offline Macbeth

  • Super Contributor
  • ***
  • Posts: 2550
  • Country: gb
Re: Train destination display - up and running
« Reply #24 on: December 17, 2015, 02:54:25 am »
I really like the flap display. I remember them in clocks from the '70s.

Also, nobody does transport as cool as the Germans.  :-+ (Oh, the Japs do... wrong... they don't have no limit autobahns and only sell super power cars for export!)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf