We’ve Been Working On The Railroad-V: Turntable Indexer
Back in issue #51 (still available for sale) I wrote an article on a variety of uses for the Atlas O Universal Switch Controller (USC). One of this uses was a turntable indexer and I did a proof of concept using a paper clip as a target for the IR sensor. Well, I decide it was time to put the indexer to use as I am now working on the White Top module of the layout.
First a refresher. For those of you who may not have seen the article, or who, like me, are too lazy to go dig it out, here is the circuit diagram published in #51.
The Sensa-Trak II is an IR transmitter/receiver pair from TCH Technologies. It’s output is normally “high”, i.e. a positive voltage and when triggered the output goes “low”, i.e. to ground (0 volts). Since the USC does not like to have a ground on the input constantly, I used a small trigger circuit that will go low when the sensor is triggered but then charges quickly back up to the the source voltage. This way, the turntable can be restarted. Without that trigger, the table would be stuck in place.
Here are the various circuit boards secured under the layout next to the Millhouse Studios turntable.
I use a Radio Shack Wall Wart that puts out 18 VAC. That is controlled by my remote power strip. That 18 VAC feeds a 12 VDC power supply which provides the power for both the USC and the Sensa-Trak sensor. I feed that 12 VDC over to a 5 VDC power supply to run the turntable. My previous testing showed that 5 volts was more than enough to run the table at a nice slow steady speed.
What I needed to do to make this all feasible was make an adjustable target so that I could align the table to incoming and outgoing tracks. The target had to be easy to make with no special machine tools and it also needed to be fairly cheap. I thought about this for a while and even searched online for something I could buy off the shelf. What I came up with is shown below.
That is a strip of 0.032″ thick brass stock (K&S from the LHS), two 6-32 hex nuts, a 6-32 square nut, a 6-32 acorn nut, a 6-32 brass screw 1-1/2″ long and a piece of .040″ brass wire. The hex nuts are soldered to the brass strip while threaded onto the screw. Then I drill out the nuts so the screw just passes through the nuts (#29 bit, 0.136″). I put the screw back through the hex nuts with the square nut threaded on the screw and put the acorn nut on the end with some Walthers Goo on it. Now when you turn the screw head the square nut slides back and forth along the brass strip.
The target is held onto the bottom of the turntable drive wheel with Scotch double-sided foam tape. I lined up the table and incoming track by hand. Marked the location of the sensor head on the drive wheel and then attached the target so the square nut was in the middle and aligned with the mark on the wheel.
To align the table I manually backed the bridge away from the incoming track and then triggered the table. Once the table stops because the sensor is active I check to see where the table stopped. From that I know to adjust the screw so the table advances a bit more or stops a bit earlier. I go back under the table, make the target adjustment and manually move the bridge back away from the sensor, then restart the table and check the alignment again. It did not take long to get the table lined up dead-center. It took maybe 4 trys at 10 minutes, tops.
I assembled two micro-pushbuttons with about a foot of wire attached on some perf board and then built a small cabinet out of styrene siding scraps to house the perf board. The buttons get connected to the start and emergency stop inputs on the USC. The housing looks like a utility cabinet next to the turntable. Flip open the top and the button on the right starts the table. The button on the left is an emergency stop button.
Finally, since the USC has a built-in output for an Atlas O dwarf signal head, I placed a dwarf next to the incoming lead to the table. It’s lit green when the table is stopped (and aligned). It’s red while the table is running. Cool!
I realize now that I did not need such a long screw. I just was not sure how sensitive the target/sensor pair was going to be so I wanted a lot of adjustment. Now that I’ve done it four times, I could likely have used a screw half as long which would allow more targets to be installed for more tracks.
Category: !Joe's Projects
Saving...
Good bit of inventiveness Joe. You also demonstrate a comfort with things electronic…a handy capability in model railroading. I’m woefully challenged with that.
I try to take something away from most articles involving ingenuity…that use of brass strip, nuts, machine screws, and solder to create that “widgit” will stick with me going forward.
Bob
Glad you like it. I really wanted to make it simple enough for anyone to duplicate.
I’ve always had an interest in electronics. My Dad owned a Television sales & repair shop so I grew up around electricity and electronics. In 1972 I scratchbuilt a digital clock. My first journal publication had to do with digital electronics and I had an article published in Jan. 81 MR on how to built an IC throttle in a Bud box. It’s in my blood.
Will this circuit align the turntable with equal accuracy rotating either clockwise or counter clockwise?
Charlie
No, it will not. I discussed this in the original article. Because neither the target nor the sensor beam are infinitely thin, there is hysteresis in the sensor response. I’ve considered ways to reduce the hysteresis, making a slit for the sensor, reducing the target diameter, and more, but in the end I decided to accept one direction rotation as a limitation. The whole set-up is less than $50. A bi-directional encoded indexer would cost several hundreds of dollars. For me, it’s an acceptable trade-off.
I suppose one could measure the difference between CW and CCW rotation and then attempt to attach two target pins that distance apart, but I think that way lies madness!
My solution for avoidance of madness for my two turntables is to align by eye. However, one technique a friend has for this problem is, when necessary to rotate the other way, to go past a short distance and then come back using the alignment feature.
Charlie
That would work but would have entailed a reversing switch run up to the little shack and back to the motor, or adding another Atlas USC and a third push button. It already looks like spaghetti under there so I opted not to do that.