Product for use with a diode matrix?

[jbonkowski]

   

    I'm trying to update some of the electronics on my club layout. We have several large yards that use diode matrices along with Tractronics Switchlock boards to control the Tortoise machines. Unfortunately, Tractronics is no longer in business.

    Is there another product out there that performs the same function? There seem to be a few products out there for controlling Tortoise machines, but they seem designed around computer/DCC control. I need something that will throw the Tortoise based on two momentary contact button inputs, when either input is grounded, the Tortoise will go in one direction. I think Circuitron makes something for that, but it's ~$25 per Tortoise! Is there anything cheaper?

    Does anybody even do a diode matrix anymore for yard ladders?

    Jim
     

[Rasputen]

You could use a latching relay - they are in the $5.00 to $6.00 range.  I would add a snubber to make the relay contacts last longer.  The tortoise is a pretty big inductive load.

[mmagliaro]

I think you could do this with latching relays as Rasputen suggests.  I would use the DIP style that have higher-ohm coils (like 400 ohm or more) so that even if 10 or 20 of them fire at once when you push a pushbutton to select a route through your diode matrix, you don't have to worry about the relay coil current overloading your pushbuttons.   There are DPDT latching relays like this for sale all the time on surplus electronics sites, eBay, etc, and many of them have the exact same pin-out, so if one ever fails and you mount them all in IC sockets, you can just unplug it and replace it.  I have used latching relays like this for years.

If you want to keep using the non-mechanical solution (the Tractonics board or other)...

Any chance you can post close-up photos of the old Tractronics board so I can see what the circuit looks like?
If I understand this correctly, you push a pushbutton that is connected to your diode matrix, the various outputs of the diode matrix
power a bunch of those Tractronics boards, which "latch" even after you let go of the button, so they power a bunch of Tortoises to select the route through your yard ladder.   Sounds like that Tractronics board is something like a simple flip-flop that could be reverse engineered easily enough if we could see the board close-up.
The Circuitron board you mentioned (I am guessing  their "TC-3") probably does the same thing, but I notice in their instructions that they say that on power-up, it will randomly select one direction or the other, which would be bad.  You want it to select NOTHING on power-up, so the switch machines all stay where they are.
--
If somebody were to devine a new board that is functionally equivalent to the Tractonics or Circuitron board, but you had to buy the components yourself and solder them on, would that appeal to you?   Once the circuit is devised, I can see from the Circuitron board that nothing on that board is very expensive.  I'm sure all the parts could be had for 2 or 3 bucks at most - cheaper if you buy all the parts in lots of 20-30 if you are making a bunch of these.  I'd be intrigued to see if I can duplicate the functionality of your board, and then you could solder them up yourself.

[John]

This is an example I think

http://www.circuitous.ca/xSingleCoil4TrackLadder.html


http://www.circuitous.ca/StallMatrix.html

[jdcolombo]

Another link to the Canadian site with the actual controller boards.  I use about 30 of these boards to control some 80 Tortoises on my layout, and they have worked perfectly for 20 years.

http://www.circuitous.ca/556Stall08.html

John C.

[John]

Another link to the Canadian site with the actual controller boards.  I use about 30 of these boards to control some 80 Tortoises on my layout, and they have worked perfectly for 20 years.

http://www.circuitous.ca/556Stall08.html

John C.

The site that keeps on giving .. Not sure if he is still producing boards - but they seem simple enough to order some from a jobber

I think this might just be the ticket for the OPs question

[Rasputen]

I wouldn't put the LED's in series with the tortoise as the one ciruitous.ca example (and the Circuitron instructions) shows.  When the LED fails, you will lose function of your switch machine until you replace the LED.
When the voltage is reversed on a stalled motor, there is quite a large spike that will appear across the motor leads.  I measured over 300 volts when reversing a 10 volt supply.  There is an A/C component to this spike, which I think is caused by the piano wire throw rod that is unwinding itself as soon as the voltage changes.  Thus you get about 5-6 fluctuations every time this event takes place.

[haasmarc]

I acquired Tractronic's inventory years ago and still have some parts.  I know we have diode matrix boards and I think we still have SwitchLocks if you want to keep with the same electronics.

[mmagliaro]

The thing I don't like about all of these circuits is that they all rely on a flip-flop that "assumes" one output state on power-up.  So all your turnout ladders are going to move to that state when these things power on.  One of Paisley's circuits on that page has a feedback from the spare contacts of the Tortoise to prevent this (the one that has "memory" to keep the last position).  But I'm not really keen on that one either, because it only works for remembering the position of a sinlge turnout, not the selection of an entire ladder route through a diode matrix.   The circuit really needs to be a tri-state output: Hi 1, Hi 2, or "neither", instead of just "hi" or "lo".
--
Considering all this, the latching relay idea of Rasputen's is really starting to look like a winner here.  Those things will say where you put them and the power-up won't move them.
--
Rasputen: 300v kickback from the tortoise?  Wow.  Well, however the OP does this, you definitely want a simple snubber on all those Tortoises to keep them from destroying electronics or relay contacts.  I'm surprised more of those Circuitron and other boards don't get damaged by this.  But then, maybe they do and users don't know why.

[Rasputen]

[ Guests cannot view attachments ]

Here is the ugliness you get back from the tortoise when you have it stalled in one direction and then reverse its polarity.  I didn't bother with the really high frequency stuff as it is beyond what my portable scope could measure.  The bottom of the screen is zero volts - if you look closely at the very lower right hand corner you can just see the trace is starting to stabilize at about +10 volts. If there were a simple inductor there, you would expect to see just one spike.  Since the piano wire is mounted on the tortoise lever, it will start to wind the motor and gear train back towards the middle as soon as the voltage changes.  I think this is what is causing all of the AC.  I didn't take the time to clamp the tortoise arm in one position to prove this.

[peteski]

Rasputen:  we use all Tortises on my friend's layout. Totaling about couple of dozen. They are powered by a couple of 12V DC power supplies (large wall-warts), and we have LED indicators (2 reverse-connected LEDs in parallel) connected in series with the motor.    If I recall the operating current in that circuit is about 5mA and stalled probably around 12mA.  I don't remember exactly because we installed them about 20 years ago.  Few are powered bu stationary decoders, while most are controlled by DPDT toggle switches.  The layout has been operated regularly (usually one 2 hours operating session once par week, plus othr special events) and we have never had an LED get damaged.

Even if there is a high voltage inductive spike, I suspect that the reason the LEDs survive is because each LED protects its peer by clamping the voltage to around 2V (Vf for either LED while it is conducting).  Wiring indicator LEDs into the motor circuit is actually widely recommended for Tortioses.   

LEDs in general are damaged by either forward overcurrrent, or reverse voltage (excessive reverse voltage creates a destructive reverse current of the PN junction).  I suspect that the current caused by the motor's high BEMF voltage is very low, and the forward-conducting LED can easily absorb it, clamping the voltage to very save 2V (thus protecting its reverse-wired peer LED).

[peteski]

This is an example I think
http://www.circuitous.ca/xSingleCoil4TrackLadder.html

http://www.circuitous.ca/StallMatrix.html

John,
that schematic implies that the switch machines will only remain powered up for the duration of the throw (momentary push buttons).  Stall switch machines should remain powered up after the throw is complete (that's why they are called "stall switch machines).  Almost seems that this circuit is meant for the solenoid-type switch machines which throw one way or another depending on the polarity of the power fed into them. Like Kato Unitrak machines.

[Rasputen]

Peteski, I have about 24 tortoises on my layout, which rarely gets run.  I have had to change about 8 of the LEDs out, and it always seems to be the red one that fails.  I made the mistake of using LEDs with integral leads and a plastic bezel, so they are very difficult to change.  A local gentlemen is building a large S scale layout in his basement, and he is suffering from the same phenomenon, even though his layout is only 70% complete.  My point is that when one of the LEDs fails, the tortoise will not function in both directions until you replace the LED, when they are wired in series.

[John]

Peteski, I have about 24 tortoises on my layout, which rarely gets run.  I have had to change about 8 of the LEDs out, and it always seems to be the red one that fails.  I made the mistake of using LEDs with integral leads and a plastic bezel, so they are very difficult to change.  A local gentlemen is building a large S scale layout in his basement, and he is suffering from the same phenomenon, even though his layout is only 70% complete.  My point is that when one of the LEDs fails, the tortoise will not function in both directions until you replace the LED, when they are wired in series.

I've used some of the christmas light LEDs in the past .. left them in their holder -- one goes bad - pull it out

[peteski]

Peteski, I have about 24 tortoises on my layout, which rarely gets run.  I have had to change about 8 of the LEDs out, and it always seems to be the red one that fails.  I made the mistake of using LEDs with integral leads and a plastic bezel, so they are very difficult to change.  A local gentlemen is building a large S scale layout in his basement, and he is suffering from the same phenomenon, even though his layout is only 70% complete.  My point is that when one of the LEDs fails, the tortoise will not function in both directions until you replace the LED, when they are wired in series.

Interesting.  This is the first item I've heard of this problem. On my friend's layout we use standard green and yellow LEDs.  Their chemistry is similar to standard red LEDs.    By "standard" I mean the Gallium Arsenide material with Vf of 1.7 - 2.0V.  Same as has been used since LEDs were first invented.   If anything I would have expected the newer blue, true-green, or white (blue LED with yellow phosphor) LED getting damaged. Those are much more sensitive to voltage.   

Are your LEDs wired the same way as my friend's?  Are you using a 12VDC power?




Also, if the LED failure model is an open circuit then the Tortoise would be able to still throw in one direction (not that this  would be very useful).  On the other hand, if the defective LED was shorted internally, the switch machine would still function normally (but the other working LED would not light up).

Like I mentioned,  I'm not dismissing your claims of the problem, I'm just really surprised to hear about this problem, since if you do a Web search for Tortoise wiring diagrams, a majority show using LED indicators wired in series with the motor, and in the past I have never heard of the LEDs getting damaged (likely due to to the reasons I mentioned).  If this is as serious of a problem as you say, I suppose adding a 0.1uF (500V ?!) in parallel with the motor terminals might flatten out those voltage spikes.  Those caps are available dime a dozen, although 500V rated might be a bit harder to find. Maybe a 200V rated cap would suffice?

EDIT: This whole 300V thing just doesn't seem to add up to me.  Was a x1 or x10 scope probe was used?