SCRs & Latching relays

Thread Starter

maassmi

Joined Oct 30, 2011
81
Ok, I think I am getting a better picture in my mind.

I just don't know which wire goes where. I drew a picture with two of the styles of relays you sent along with my wiring needs. Can you match a "numbered" pin to a "lettered" wire for me? This is where my understanding fails. Thanks.

Oh, the resisters are soldered onto D and F, hense their longer length.

Thanks again.
 

Attachments

djsfantasi

Joined Apr 11, 2010
8,651
I have quickly sketched out a simple circuit to drive the latching relay and your switch machine from your existing Atlas Switch Control Boxes. It does depend on the Atlas Switch machine receiving enough power when a diode is put in series with a coil, but since diode matrix ladder control is done all the time, I believe we are in good shape here. Using diodes in this way, we are powering the latching relay in what SgtWookie meant by "If 16VAC is half-wave rectified, the output will average ~11.3v". One diode will conduct "half" of the AC "wave".

Note the two horizontal diodes - they are used to switch polarity between pins 1 & 10 depending upon which way the turnout is switched. The vertical diodes are to prevent current following a sneaky path and trying to energize both switch machine coils at the same time. Since the switch machines draw a lot of current, I would use 1A diodes (have in the past), but you should measure the current draw of the machine and size accordingly.



As far as the power supply that SgtWookie is referring to, once you try to light too many LEDs or incandescent bulbs, you will see them start to get dim, or stop working. That is because you are drawing too much current from your "old standard trasnformer". Searching on this forum, you can learn how to "recycle" an old computer power supply, that will give you much more power for your lights. It is not difficult to use multiple supplies and in your case, they likely can all be kept separate. But I see that as a separate issue from getting this latching circuit working.

Happy Modelling!
 

Attachments

Last edited:

SgtWookie

Joined Jul 17, 2007
22,227
The 2N5062 SCR (Thyristor) 2N5062GOS-ND will stay on when powering a 20mA, 2.1VDC LED and a 470 Ohm resistor. http://search.datasheetcatalog.net/key/2N5062
I've already mentioned a big problem; the SCR's "going stupid" when the power is removed. In order to "fix" all of the lights after power is cycled, each switch would have to be energized for at least a moment.

Another problem (pending) is that our OP made a statement that their LEDs are already too bright.

One way to reduce the intensity is via PWM, which would render an SCR useless as a latch. Another way is to limit current in a linear fashion, which at some point will cause the SCR to fail to latch. The only other method I can think of offhand would be to shunt current around each LED individually, but that would require a ridiculous amount of wiring.
 

Thread Starter

maassmi

Joined Oct 30, 2011
81
Hi djsfantsi, I am unable to retrieve your sketch. On my screen your attachment appears as an X marking a photo or bmp that would not download. Can you save it in PNG format and try again please?
 

SgtWookie

Joined Jul 17, 2007
22,227
I sent djsfantasi a quick PM about their image not showing up.

Meanwhile, I re-formatted the datasheet image of the relay so that it would be a bit easier to understand what the internal connections are:



The contacts are shown in the RESET position.
 

Attachments

Thread Starter

maassmi

Joined Oct 30, 2011
81
This appears to be diagram of the 2-coil relay you sent me earlier. My problem is I don't know which of my wires connect with which of the pins or leads on the relay. Can you help me with this?
 

SgtWookie

Joined Jul 17, 2007
22,227
I came up with a version for the two-coil relay; I tried to keep it pictorial rather than schematic-like.

The resistor value will largely depend on your DC supply voltage, and somewhat depend on your LED's Vf (forward voltage) and the amount of current you wish to flow through them.

The diodes can be any common diodes; 1N914, 1N4148, 1N4001 thru 1N4007, etc.



As you can see, the dual coil relay requires a little bit less wiring; only two diodes per relay is needed.

Another handy thing about the dual coil relay is that you can install them in a socket, and it doesn't matter which way you plug them in; they will work just the same.
 

Attachments

Last edited:

Thread Starter

maassmi

Joined Oct 30, 2011
81
djsfantsi: I think I understand which wire goes to which pin, but I think there may be a "short" in this system. I attaching a sketch of what I think you are trying to show or say to me. I am drawing a "blank" on this one little thing of where do the wires from the switch connect to the relay.

I think I am almost there.
 

Attachments

Thread Starter

maassmi

Joined Oct 30, 2011
81
Wow, SgtWookie,

This picture I fully understand!:D Placing the resister where you have it will cut my need for resisters in half! This is cool. Now, for the diodes. I am unfamiliar with them. From what I have read, they will cut the AC power flow in half or something? I also noticed that there is a line or band across one end of the diode. Does this band indicate which way the electricity must flow through it in order to function correctly? I am assuming it does. Are they durable enough for me to solder them directly to the wires? What can I expect to happen if I don't use a diode? This set up is nearly identical to the Atlas #200 snap relay. This makes it very easy for me to understand. Thank you so much.:)
 

djsfantasi

Joined Apr 11, 2010
8,651
I went with the one coil version because it was most like the Snap Relay's the OP is trying to replace; the relay controls a DPDT set of contacts. Then he can use one set to control the frog of the switch machine and the other set for indicator lights or relay logic.

Sgt., I think there is a sneak path for current through the solenoid coils.

IMHO, if one were to put resistors in both legs of the LEDs in your circuit, instead of the power source, it would be better when putting indicator lights in parallel. Often times, one set of indicators are used on the panel, and there may be two others on the model layout - one in each direction of traffic of the track.
 

Thread Starter

maassmi

Joined Oct 30, 2011
81
Ok, Djsfantsi has a point. I will use this to power two sets of lights at the same time so it does make sense to put the resister on the power feed end of the LEDs instead of from the power source prior to the relay.

But, Djsfantsi, I still have a question about the AC power feeds in your diagram with the single coil version. Based on what I see on the 2-coil version, would I connect the wires the same way on the 1-coil version?
 

Thread Starter

maassmi

Joined Oct 30, 2011
81
djsfantsi, after looking at your drawing a little longer and after seeing SgtWookies drawing, I am assuming that the N wire in your diagram can be connected to either the 1 or 10 pin on the relay and the R wire in your diagram is connected to the reamaining of these two pins, and that the C wire in your diagram connects to both the 5 and 6 pins. Is this correct?
 

SgtWookie

Joined Jul 17, 2007
22,227
Wow, SgtWookie,

This picture I fully understand!:D Placing the resistor where you have it will cut my need for resistors in half!
Since the red and green LEDs cannot be on at the same time, you can use just one resistor for that pair. You should not attempt to "share" a resistor between more than one LED connected in parallel; as their forward voltages will be somewhat different, and the one with the lower Vf will get more of the current.

Now, for the diodes. I am unfamiliar with them. From what I have read, they will cut the AC power flow in half or something?
They only allow current flow in one direction.

I also noticed that there is a line or band across one end of the diode. Does this band indicate which way the electricity must flow through it in order to function correctly?
The end with the band is called the "cathode", the end without the band is called the "anode". Current flows from the anode (the end without the band) to the cathode (the end with the band).

This is conventional current flow; electron flow is just the reverse of conventional current flow - but don't let that confuse you.

If the voltage on the anode is more positive than the voltage on the cathode, then current will flow through the diode.

Are they durable enough for me to solder them directly to the wires?
Well yes, but you should use pliers or hemostats to wick the heat away from the body of the diode. You should not have to apply heat for more than 3-4 seconds to get a good solder joint.

What can I expect to happen if I don't use a diode?
If you do not use diodes as shown, the relay will be left latched in an unpredictable state.

This set up is nearly identical to the Atlas #200 snap relay. This makes it very easy for me to understand. Thank you so much.:)
That's why I drew it that way. :)
 

djsfantasi

Joined Apr 11, 2010
8,651
OK, I drew my sketch out in more detail, showing the indicator lights like SgtWookie did.

By the way, N stands for the Normal switch position, R stands for the Reverse switch position. C is the common power. C also is wired to pin 10 of the latching relay. S/R is the set/reset signal. Because of the diodes connected to the Switch Control Box, when one is pressed, a + voltage is applied between pins 1 & 10, causing the relay to set; when the other button is pressed, a - voltage is applied causing the relay to reset.

The other two diodes (shown vertically) prevent current from sneaking through pin 1 back to the other switch machine coil.

The indicator lights are then switched by the latching relay contacts, as can other circuits such as the switch machine frog.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,227
I went with the one coil version because it was most like the Snap Relay's the OP is trying to replace; the relay controls a DPDT set of contacts. Then he can use one set to control the frog of the switch machine and the other set for indicator lights or relay logic.
Ahh, OK! I forgot about the frog polarity.

I used to have a Lionel "027" gauge train; started off with it in 1957. I collected more stuff over the years, but then I left it at my folks' place when I went in the service. It basically sat there for over 30 years. I sold it a few years ago. Opening those old boxes up was like jumping in a time machine. But I digress - the "0" and "027" gauge trains run on AC and are 3-rail, so the frog polarity is not a problem.

Sgt., I think there is a sneak path for current through the solenoid coils.
I don't see how?
The switch control box represents an Atlas #56; the light violet wire is the AC return (common). The blue wires are N and R; I can't remember which is which, as I haven't seen one for a number of years. However, they can't both be powered at the same time.

IMHO, if one were to put resistors in both legs of the LEDs in your circuit, instead of the power source, it would be better when putting indicator lights in parallel. Often times, one set of indicators are used on the panel, and there may be two others on the model layout - one in each direction of traffic of the track.
If indicator LEDs are going to be paralleled, then they will need separate resistors. I was trying to avoid that, as having the resistor in the +V line near the supply would reduce the possible amount of damage were a short to occur somewhere in the system.
 

Thread Starter

maassmi

Joined Oct 30, 2011
81
I see that by having two diodes in the line with the connection inbetween the diodes, any "short" is eliminated because the electricity is allowed to flow in only one direction. Got it! Thanks! :D
 
Top