I ran into an interesting problem as I was trying to enhance my layout by adding some dwarf signals to indicate the position of select turn outs; red when the turn out is diverging, green for mainline. I am running TMCC, with the switches controlled by an SC-2. The SC-2 is configured to run 4 switches with the remaining two terminal blocks set up for accessories. The SC-2 is using the first switch block for power, i.e., I am not using the wall wart nor external power connections. Each switch's common terminal is tied to the C terminal of a 1033 (16V) which is thereby powering the SC-2 via the switch tied to block one. The switches are Atlas Super O.
Here is a diagram of the configuration with everything working:
I am only showing one switch, however, all four of center / left terminals are hooked to other switches identically to the one shown.
I decided to use Atlas Snap Relays to control the aspect of the dwarf signal. They are similar to the switch machines in that the are a solenoid with a common tap to the center of the coil and two connections to each end of the coil. Completing the circuit on either end move the core toward that and sliding some internal contacts. I wired them per the Atlas directions. The result looked like:
I now basically have two coils in parallel, one through the switch and one through the snap relay. I threw the switch via the SC-2 and immediately blew the 1.5 amp fuse I had in line to the C terminal on the 1033. I tried another fuse with the same result and proceed cautiously with a DVM. I deemed it safe to temporarily bypass the fuse with a jumper. I was able to get the snap relay to work, however, the switch machine would barely move. It tried, but it acted as it it was not getting enough current. I also noticed that the snap relay was getting very warm. I also determined that even with the power removed to the snap relay common connection (connection H in the diagram) the switch machine would not work – it tried but just did not have enough power. When I completely removed the snap relay the switch worked fine. When it was reconnected, the same result – the snap relay worked rather crisply and the switch machine would barely move and would not definitively move the points to either side.
With the snap relay getting hot, I decided to measure the resistance across the switch and relay when they were disconnected from the circuit. I figured I had created a basic current divider with the two coils (or more appropriately inductances) in parallel. The resistance with a DVM measured 27.5 ohms between D and E (the switch machine) and 4 ohms between G and H (the snap relay). So I assumed that the snap relay was using 7 / 8 of the current and the switch machine was only getting 1/8 of the current. Also, with a combined resistance of about 3.5 ohms (I realize that the coils do not act as a true resistor when AC is passing through but rather an inductor) I estimate the current draw to be somewhere around 5 amps. Hence the untimely demise of my 1.5 A fuse!
So, armed with this information, I tried to put in first 20 then 50 ohms of resistance on the common side of the snap relay to raise its effective resistance to be close to that of the switch machine, hoping to balance the current between the switch machine and the snap relay. This then looked like:
Still, the switch machine is lethargic, as is the snap relay. So I have no solution to this dilemma.
Any thoughts of what is occurring and any ideas on getting the SC-2 to drive both the snap relay and the switch machine? Any thoughts would be appreciated.
I had searched through the forum and found some postings from Roger Bielen regarding the use of the DZ switch machines which have an add on to enable the driving of signaling. That is an option but I already have a decent investment in the Atlas products. If I can not get the current configuration to work, I may have to switch, as I did check and the DZ's claim to be Atlas compatible.
Another option I have considered is building a capacitive discharge power supply http://www.awrr.com/cdsupply.html. These do have the “problem” of a 2 – 3 second recharge which will cause trouble for routes with the SC-2. I believe if I drop the optional ready circuit, I could easily build a supply for each switch (8 total) driven by the SC-2 to get around the route problem with the recharge time. My only question with this option is that the switch would now be DC powered. Will I still be able to use the SC-2 in the same configuration as above. Since the #1 switch block would no longer be powering the SC-2, would use external power or the wall wart? However, Lionel states in the SC-2 manual to not use any type of external power if the SC-2 is driving switches. Any thoughts?
Sorry for the long post folks – I appreciate any assistance!
- Luther
You seem very knowledgeable have an interesting electrical problem. I am not qualified to solve the electrical isse but would offer the following.
dz stuff does offer Atlas compatible switch machies. changing to them could become a sizeable investment [$s and time] even for only four switches. basically, as you will see from their website, a switch machine and a relay are needed to manage the application you want -- drive the switch and activate the dwarf signal. It will work fine. i use it with ross/gargraves.
an all in solution --dz2500 (+/- $25), a dz 1008 relay (+/- $9) per switch . you would also need a dz2001 databus driver (+/- $25) and the wire for the bus....and the time to do it. they work well with SCs and ASCs.
They have a couple of other products as well that you may want to investigate---like a dz1011 that look and work like dwarf signals but are optical sensors --like lionel's 153ir -- with a much smaller footprint.
I would consider using the atlas, the way it was, without the "snap" relay and pick up the dz 1008 relays and operate the signals you already have with isolated rails.
ralph
Luther,
I am using a setup almost exactly like yours, but I use the snap relays to control 10 amp power relays to power the track as I change routes with the switches. I use an 18 volt tranformer to power the switches and snap relays which puts out 4 amps. It is protected by a 5 amp CB.
One thing that I read on a few forum posts is to never use the #1 terminal on an SC2 as it creates problems due to some internal wiring. I use the wall wart to power the SC2 and have never had a problem with this setup.
One additional note, you can change the #5 & #6 position on the SC2 to momentary contacts and connect switches these positions so that you can control 5 switches with each one without using term #1
If you want more info on this do a search on SC2 or SC-2 on this forum or go to the other forum and search the TMCC section. There has been a lot of discussion on this subject.
Your wiring diagram looks OK to me. The difference in resistance between the snap relay and the switchmachine, I think is due to the fact that the snap relays are actually modified HO switch machines and the coils are probably wound with smaller gauge wire. This would account for the difference in measured DC resistance. As you mentioned this is not really relevant to AC operation.
Good luck on solving your problem, hope I have been of some help.
After thinking about your problem a little, I may have overlooked the simplest explaination. Do you have other loads connected to your 1033? The problem may be as simple as not having enough power available to meet the inrush current requirements of two solenoids operating at the same time. If your fuse is a fast blow variety, I don't think it is large enough.
John,
Thank you for the follow up.
I do have the variable outputs on the 1033 in fact tied to a 10V bus on the layout (the 16V for the relays and switch machines are pulled from a fixed post). I have two SC-2s and a Gantry Crane tied to the 16V bus. On the 10V bus I have a blinking water tower, fueling station and ice house. During my testing, nothing was running on the 10V bus – I was only testing the switches. But I can check the current draw on the 10V bus to make sure of this, or even completely disconnect the 10V side. I seem to be pulling a steady ~170mA on the 16V bus with nothing operating – I suspect this is power for the 2 SC-2s.
Also, I thought it may have been a voltage issue as the switch machines and the relays are listed for 16 – 18 V. I tried a Lionel accessory transformer set to 17V (I believe it can provide 30W). That would only be about 2A and I believe that you said your set up was pulling 4A for both. I suspect that it could be too much for the accessory transformer thereby negating my test, but I suspect that a lone 1033 should be OK. However, the 1033 is a new addition for me as of last Monday, but I would think if it is supplying 16V everything is well with it and it could handle the amperage – but I am not that familiair with the 1033.
I will definitely try to disconnect the 1st terminal set on the SC-2 and use the wall wart (on both SC-2s). That should be simple enough to do. I had seen some of the posts and warnings you referenced about using the 1st terminal. However, I tried it and everything worked with just the switch machines. I figured I would give it a try as I hate to waste 1/6 the capacity of the SC-2!
I will do some more testing this evening – as your feedback is telling me I am at least on the right path and this should be doable. If I still have issues, I can try a CW80 which I have (with its own set of issues) and also try to remove the SC-2 completely from the equation by using just the Atlas control switches provided with the switch machine.
Thanks again for the help.
I was able to do some further testing this evening. I removed the power / switch connections from the #1 terminal of the SC-2 and used the wall wart. Still no luck, the switch machine was very lethargic. I then tried a CW-80 and both the switch and relay moved as they should. The switch was not moving to the full extreme but it was locking the points in place. The Atlas switch machines are funny in that they do not need full throw to lock the points based on the internal slide mechanis.m
I then decided to do some more testing with the 1033. I had it plugged into an APC surge protector / power strip combination. I tried various plugs on the strip and also directly to the wall. When plugged into the wall the switch seemed a little peppier but it was still not moving fully. I then attempted to disconnect the 10 volt bus from the A – U terminals on the 1033 and the switch seemed to move even more, but still not to the full extent and definitely not locking the points. I reconnected the A – U terminals and made sure there was no current being drawn. Showing 0 A on the 10V bus, I still got lethargic performance and only partial travel on the switch.
So it looks like the 1033 was the problem in this case. It just did not have enough current push. It was worse when there were connections to the A – U variable posts even with no draw. So I guess I will be going back to the CW-80, which the 1033 was bought to replace!
Our community is FREE to join. To participate you must either login or register for an account.
Get the Classic Toy Trains newsletter delivered to your inbox twice a month