The Unitrack crossover changes four turnouts at a time, but it has only two positions -- straight or crossover.
Thank you for your explanation Mark. I'm using 1000uf caps on my turnouts and they work fine. So then, with my Unitrack crossover application, you think that my 4700uf is okay to use? (It's 0.175 times more than 4000uf.) I've tested it and the crossover works well. I'd otherwise go out and simply buy something closer to 4000uf if required, but I'd pay ten times their cost in shipping. Thanks again!
4700 uF is fine for the double crossover. It would even work fine for a single switch, as long as you were not quickly toggling it back and forth, repeatedly for an extended period. In a bridge rectifier circuit, the low pass filter shaves off the tops of the peaks and fills in the valleys of the rectified AC waveform. Kinda like railway/highway construction in hilly terrain: you use the energy/dirt shaved from the peaks to fill in the valleys.
Ah, I'm relieved. Good to know I'm not going to burn it up. I really appreciate you and @Mark Ricci sharing your electrical engineering knowledge with us. This isn't the sort of application where answers to are easy to find!
True that it has 4 turnouts...but...it only has to change 2 at the same time...for the crossover maneuver to work. Those 2 turnouts are one function at the same time. 2 opposing deverging routes connecting to each other. Unlike setting 1 turnout to deverging to go into a siding. You dont have to set the turnout at the far end of the siding to deverge back to the main from the siding to get into the siding at this end.
Just to clarify a little more... In a RC circuit, the time it takes a capacitor to charge to 63% of applied voltage is known as the RC time constant.. eg; for a 10ufd capacitor and a 1 M Ohm resistor in a circuit with an applied voltage of 20V, the time constant is 10s.... Assuming the cap is fully discharged, and power is then applied, the capacitor will charge to .63 x 20V =12.6V, after 2 time constants, Vc = 17.2 V, after 3TimeConstants VC = .95 x Va, 4TC .98 x Va 5 tc =.99 x Va. AFter the 5th tc, the capacitor is in its steady state and Vc = Va. Vr = 0 For discharging, it essentially reverses. after the 1st tc Vc remaining = .37 , 2tc = .14 and after the 5th tc, the cap is essentially fully discharged. In power supplies... For a full wave rectifier connected to 60hz AC in, means a 8.3 mSec period from 1 peak to the next peak at the rectifiers output so ideally, the value of capacitance is selected to provide voltage via discharge in smoothing out the peaks. When a higher instantaneous load ( a significant decrease in R), or demand for power occurs, the time constant decreases and the capacitor discharges faster. The function of the cap is to keep the voltage as constant as possible and then discharging effectively, providing additional power to throw the power hungry 4 solenoid beast. I love the double crossover (thinking my next layout will have 4), but power requirements significantly reduce options for control under DCC.
Maybe its been so long since I used the double crossover with electrical switches that I dont remember if all 4 turnouts move at the same time or not. I do all my turnouts manually. Doing them manually means only 2 turnouts need to be 'flipped' at a time for a 'crossover' to work. Edit: I found my answer... The four turnout machines are wired to a single cable coming out of the turnout, so all four turnouts will switch at the same time. I will crawl back under my rock now... .
Technically true, but (there's always a butt!) the other two turnouts are useless if not also thrown, since their through route would run against the thrown switch at the other end of the same track. This is the same reason you would decide to throw the two opposite corner switches from a single control: you don't have to, but it's easier than throwing each one, and there are no useful cases where you would only throw one. Therefore, there really is no useful reason not to throw all four turnouts at the same time. A larger capacitor, but fewer (electrical) switches, indicator controls, less wiring, etc. are required to throw all four switches together. Butt-in-chief,
I see what you saying and I missed that. Kato thought of simple minded folk like me and provided only one pair of wires emerging from the crossover. Update: I see your follow up post now.
BUT...I manually throw only 2 turnouts on the double crossover to get from A main to B main. The other 2 do not have to be thrown to do one crossover. Hmmmmmmmmm I am running DCC...so add that into the equation...
Very well stated. I would only note that, after 5 TC's have elapsed, the capacitor is effectively fully (dis)charged. In real life, parasitic (internal) resistance affects the charge/discharge rate. Likewise, parasitic resistance also slowly discharges the capacitor over time, even if left open-circuited. Very high quality (and expensive) capacitors have very little parasitic resistance, but they all have some. As the capacitor approaches Va, the current through the capacitor approaches 0. This is called asymptotically approaching a limit: you keep getting closer and closer to the limit, but never reach the limit (whether Va or 0VDC). But for most practical applications, 5 TC is the time required to effectively completely (dis)charge the capacitor.
In boolean logic, we would describe the other two switch positions as "don't care", which provides options of whether to throw them or not, and the control logic can be minimized accordingly (which Kato did, by providing only one pair of control leads, wired to all four solenoid coils.) There are a few DCC decoders capable of throwing all four switches at once in the Kato DXO. One that I am aware of is the Digitrax DS74 (and the discontinued DS64). These actually can individually control up to four DXOs independently, and can be programmed with routes (sets of positions on multiple switches, controlled with a single command), etc. These can be powered and controlled by DCC, but should not use the main track DCC bus, because inadvertent track shorts can not only render the DS74 temporarily incapable of throwing a switch (that might in turn clear the track short), but can also cause it (or at least its DS64 predecessor) to lose its address programming. I suggest separate booster or CB for track and stationary DCC decoders. You can use the command station to directly power/control both the stationary decoders and a booster which then powers/controls the track and locomotives on it. Note, when using DCC control for the DS74s, any DCC system that can control stationary decoders should work. There is room underneath the DXO to embed separate stationary decoders for each switch, powered from the track. While this greatly simplifies the layout wiring, it also presents the potential for the problems noted above when (not if) the track is shorted.
Thought I added that in the 2nd paragraph. Oh yes, we have the theoretical and the practical as no electrical or electronic component is perfect. The electrolytic capacitor while having the plus of larger values of capacitance in similar size packaging, tend to have higher leakage currents. The exponential curve in which the RC constant follows best describes the characteristics and never truly, as you mentioned, achieves Va but close enough for govt work. Oh well, the imperfections we have to deal with.. Yes after the 5th tc, capacitor works much like an open switch for basic analysis. Vc = Va Vr=0 It=0 as compared to the inductor which inversely works like a closed switch after the 5th tc. the same %s apply and the time constant = L/R
Wanting to revise the track layout at my yard throat to expand my engine terminal, I was faced with how to lift my Unitrack that had been lightly fixed in place with a spot or two of glue every foot or so. In a rare creative moment, I recalled that I had an old computer case card slot blanking plate and it's the perfect ersatz tool -- thin, strong and bent at 90 Deg. Just insert at an unglued location, work slightly slide to side and pry up. By the way, I'm now a huge fan of Aleene's Tacky Glue. It's strong, colorless, remains flexible after drying and does not permanently adhere to Unitrack.
Great idea, fortunately I saved a few to use but just threw out dozens. Shucks... They could have been repackaged and "sold" as "Attached Track Remover Tool" @ $4.95 + shipping. LOL
Off topic, I made a 40 Mile round trip across town yesterday in evening rush hour traffic to my tiny local hobby shop to buy some Kato Unitrack. He has a pretty good selection on hand and he told me that his business in Unitrack has been growing strongly. Nice that he's recognized a market and is serving it. I suppose that's the nature of Unitrack, that when you need a special length or pieces, you want them then. With higher shipping costs and local taxes now added on to Internet orders, his pricing has become more competitive and his delivery is instant.
neat use of that computer thingyboberdo . i too use tacky glue it works great track or buildings . i have lots of computers in boxes . been saving for a rainy day projects.
