Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 25, 2017 12:40:27 GMT -5
SOS ... I tried Märklin flex track, and bought 1 part to test on a piece of plywood. First it is so "strong" that it's impossible to fix it correctly in curves with nails (0,5mm), even R 400mm, and the elasticity tends to bend it back ; I cut the plastic part on the outside of the curve, every 2cm, but it solves nothing. And then I want to use 3mm plywood to hold the track in "mountain" parts of the layout. I think more is too hard to bend, to get transitions between different slopes. Problem : the nails are not fixed in a very strong way ... So I want to know some things about Rokuhan ; yes, I know, the cost for flex track is about 3,5 times that for Märklin ! But the turnouts are less expensive ... 1) About flex track, the ballast does not stay "closed" on flex track in curves. I think with a radius of 400mm this is not too visible between the rails ; re-ballasting again OUTSIDE of the rails is less difficult. 2) I read that Rokuhan flex rail does NOT tend to bend back. Is that true ? So glueing would be sufficient. And what glue would be the best on plywood ? Rokuhan recommends nails ... My Arnold layout used 3mm plywood too and there was no problem to hold the flex track in place ! 3) The turnouts seem to have a very simple way to control them : no current, straight ; current, deviation. I saw the power control module ; so there would not be any problems with monovibrators, capacitive discharge, relays etc, like Märklin ; a simple photocoupler controlled by one output of the Atmel processor, followed by an appropriate transistor, would fit ; and the turnouts would stay "synchronized" with the outputs all the way. Thanks for any answers ...
|
|
|
Post by markm on Feb 25, 2017 13:13:03 GMT -5
Alberich, If you don't like the Märklin flex track you might want to look into the peco flex track: www.peco-uk.com I've used the Rokuhan flex in a tunnel and didn't need to worry about the outside gaps in the roadbed but IMO on curves greater than about 220mm the gaps on the outside can be easily covered with ballast. The gaps between the ties and inside curve aren't noticeable. I measured the overall height of the track at .21" (5.25mm?) The Rokuhan turnouts are bipolar (two wire): + - straight, - + divergent. They use a capacitive discharge in their control module. The important thing is that the current should be momentary. Hope this helps, Mark
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 25, 2017 13:42:13 GMT -5
Thanks for Your answer.
With the bipolar system, control is even more difficult than with simple discharge. The best solution would be a bridge like L293D with two monovibrators ; or 2 capacitors and a symmetrical power supply (I want to switch with only one output bit). But I think I will look at Peco Flex rail ... Unfortunately the problem with the nails seems to stay ! Or can I use nails "thicker" (0,8mm) than with Märklin (0,5mm) ? But, very interesting : the cost seems to be less than Märklin (I think I will need more than 22 meters) ! So I would use Peco track and Märklin turnouts. But, problem : I read that Märklin uses code 55 rails and Peco is 60. Doesn't this bring problems with compatibilty ? On the other side I saw (in this forum ! ) that Atlas manufactured a Z flex track, code 55 ; and if the effort to "come back" to the straight shape is less strong, perhaps nails will be sufficient ... especially if I can use thicker ones ? And, if I come perhaps back to Rokuhan : what is the current needed by the Rokuhan turnouts (Märklin turnouts work with about 450-500 mAmps) ?
Many questions, thanks for every answer !
|
|
|
Post by markm on Feb 25, 2017 20:04:20 GMT -5
Alberich, One thing to understand about the track is that each has different tie size and spacing: Something to consider when you mix and match track.The Peco, Märklin and Rokuhan all use some variation of the European/Asian dimensions. Atlas and MTL do the U.S. standard. The height differences are not generally a problem. Rail height varies between code 55 and code 63, with the American track close to 55mils and the non-American about 63 mils. I haven't measured the current to throw a Rokuhan turnout, but they use a 10ufd capacitor through a 100 ohm resistor so it can be calculated. Mark
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 25, 2017 23:45:17 GMT -5
The next test I will do is to buy a section of Rokuhan flex track and a turnout for testing purposes, their track system seems to be much more reliable in terms of solidity (zscalehobo has published a Youtube sequence where they show how to fix the track with SCREWS, like the old M-track system by Märklin, which was for me the best track they ever made (I used it in the 60's) ; I can hear some people shouting about the realism, but in terms of reliability it was unbeatable ! And I think for myself that the reliabilty is the first thing to consider, the realism comes in second place ; I want to build a working layout, not a static diorama. I made this error once with the Arnold tracks (realistic, but rusting, not to speak about the problems with the turnouts ! ), I won't make it for the second time ... The Märklin turnout I bought for testing will go to the "Bay" and the piece of flex track to the bin (some parts were already destroyed )... To solve the electronic problem, I have a complete measurement equipment (even a new digital scope), and I am from that generation when personal IT was equivalent with "HardSoft" and who held the soldering iron in hands, long time before they could use a mouse ! As a first "Personal Computer", I began with the "KIM-1" by controlling a test railroad layout. For control I would prefer the system with the L293D (one IC could control 2 turnouts ... if the power spikes are not out of the range of the IC ! ) and the monostable pulse generators (each output connected to the inhibit of the other generator to avoid transients on the inputs, both inputs connected together so one generator is activated by the negative edge and the other by the positive edge of the common control line), with a good decoupling of the power supply for the IC's. A 10nF capacitor from the common input to Gnd and a 10K resistor to Vcc will be connected too, and the common input of the monovibrators will be controlled locally by a photocoupler (to Gnd), the link to the central processor would be the line controlling the LED, and so not noise-sensitive (like the MIDI lines for instruments). Because the turnouts switch the power to the tracks any other relay for that purpose seems to be useless ... if I take care so that groups of "stop" sections can't be "bridged" by a conductive wheel on another track, the turnouts switch only one rail off (too long to explain in detail ... ). What do You think of this solution ?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 26, 2017 13:13:38 GMT -5
1) I will post the circuit when it will be working ; first I must buy a Rokuhan turnout for testing ... 2) Another advantage of optoisolation is that the diffferent power circuits can be completely separated, without any common Gnd or else. 3) On Youtube Rokuhan shows how to "bridge" the power switch in the turnout, by unmounting the switch (and cancelling the guarantee ... ). They said they would deliver a version of PCB with that modification already made ... but I did'nt see it till today !
|
|
|
Post by markm on Feb 26, 2017 14:51:48 GMT -5
Regarding #3, I think you are looking at an old video. The original Rokuhan turnouts, R006 & R007, were always power routing, much to the surprise of Z scale users. There were videos produced for how to make those turnouts non-power routing. They have since been obsoleted and been replaced by R022 & R023 (wooden ties) and R055 & R056 (concrete ties). The later models use two small screws underneath to allow the user to select the mode of operation.
Mark
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 26, 2017 15:50:53 GMT -5
Thank You Mark ! That means that the staff at Rokuhan is listening to the users ... And what do You think about the reliability of that PCB concerning short circuits (I think this is the cause for the trouble) ? Should I better use power routing and so need no additional relay, or make the turnouts operate like Märklin's, add a relay, but "save" the contacts inside of the turnout ? Or is the whole contact system available as a spare part and easily replaceable (after unmounting the turnout) ? But for myself I think the best method is a "quick" fuse in the line to the tracks, so the PCB does not act itself as a fuse ! Thanks for any answer ...
|
|
|
Post by markm on Feb 26, 2017 19:58:43 GMT -5
I guess really the question of short circuits depends on the throttle. I've found the Rokuhan throttle cuts out very fast. Since the AZL locomotives' power consumption is on the order of 10s of mA you really don't need too much power.
It's probably just a matter of common sense. I've been using the power routing for passing sidings and light industrial switching without problems. It don't think I'd recommend using the power routing to direct power for half a layout.
I haven't heard of Rokuhan offering a spare part. But if it does fail, isn't it a matter of adding a relay?
Mark
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 27, 2017 2:38:23 GMT -5
I wanna answer especially to Greg. Because I don't want to make many modifications to my layout once finished, I prefer give up the "power switching" from the beginning, and use a relay instead. Because my system needs to switch 2 lines for every track, I think I could use a relay with 4 switches, the remaining 2 for the turnout (or group of 2 turnouts). So I attach this schematics, drawn by hand and scanned ; it's only a project, not tested, using a symmetrical power supply. I hope it's correct, especially for the freewheel diodes, and it can be understood ... The operation is to charge one capacitor while the other is discharged into the solenoid, and the length of the pulse is determined by the capacitor. I think the resistor Mark was speaking about is only to limit the charge current for the capacitor, and not in series with the coil. Because the picture is not stored on the net, I have some problems to make it appear in real size, so You must open the link in a new tab (with Firefox) for viewing. Sorry ...
|
|
|
Post by markm on Feb 27, 2017 5:36:33 GMT -5
If it helps, the Rokuhan functionally looks like this: There is additional circuitry to route the power for capacitor recharge. Disadvantage is that it takes a few seconds to recharge the capacitors. Advantage is only a single power supply. Mark
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 27, 2017 7:31:44 GMT -5
I think that, unless there is a very sophisticated regulating circuit to stop the charging (and the costs would be much more for the system ! ), some current is always flowing through the coil, and this is the reason for the time the capacitors need to re-charge : this current must be low enough, so the coil can't become too warm. The other solution would be to use a momentary switch, but this doesn't work with a relay. A complete simplified version (without freewheel diodes) could then look like this : The value of R2 should be high enough to limit the current through the coil, but low enough to allow a quick charge ! R1 is the 10 Ohm resistor used in the original version. The main advantage of the dual-supply system would be that there is no current flowing through the coil during recharge (but i should add the 10 Ohm resistor). On another side I think that a few seconds of waiting before using the turnout again is not a major problem ! PS for Mark : I went today to my local dealer, he sells Peco SL200 flex track. The problem is simple : this track is absolutely NOT compatible with Märklin, we didn't succeed in joining it to a piece of Märklin, it does even not fit into the rail joiners ! I wonder why Peco took code 60 rails, because the only manufacturer for tracks in Europe is Märklin with code 55, and Peco themselves don't have any other items like turnouts etc in Z ; the only thing one can build with Peco is an ... oval, without a station ! Unfortunately the Atlas Flex track (code 55) is very difficult to buy in Europe for one or two units, I want to test first and not make the same negative experience as with Märklin ... with the whole stock for the layout !
|
|
|
Post by markm on Feb 27, 2017 21:43:07 GMT -5
Alberich, I didn't say too much about the Atlas track because it's new even here...only about 3 months. The manufacturer bundles the track in units of 5 and many distributors only want to sell it in multiples of 5. Thanks for the information, but I'm a bit surprised with your problems mating Märklin and Peco track. The key measurement is the rail base which is 0.047" and 0.049" respectively. I was able to mate the two tracks: Both track were purchased pre-2010, so not current production. The Peco fits well with the Rokuhan, a bit loose in the MTL joiners and very tight with the Atlas. I'm not sure I'd call the Märklin rails code 55. I measure them at 0.60" to the Peco 0.063" and as you can see in the image (Märklin on the right) the rails are similar. Regarding Rokuhan turnout control, after I posted the previous schematic, I kept feeling something was wrong, so I opened the Rokuhan unit again. The internals are completely different, and I have no idea what I was thinking (English term: Brain Freeze). Anyway they do it with 2 electrolytic capacitors, two diodes a 1kohm resistor and a SPDT switch. I'm still working on cleaning up the real schematic. Mark
|
|
|
Post by smr on Feb 28, 2017 2:03:42 GMT -5
Unfortunately the Atlas Flex track (code 55) is very difficult to buy in Europe for one or two units, I want to test first and not make the same negative experience as with Märklin ... with the whole stock for the layout ! Try Joerg Erkels shop in Germany. He has Atlas track on stock and is an AZL dealer as well. He speaks English. Spur Z Shop: www.1zu220-shop.deBest, Sven
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 28, 2017 3:20:34 GMT -5
Thanks for Your replies ... To Mark : I think the system should work almost like my last schematics, perhaps the 10 Ohm resistor is not needed ; the best way is "trial and error" (I hope errors won't be destructive ! ) ; I think by using 9v DC and increasing the value of the capacitors from 10uF to 1000uF until it works there should not be a too big risk. I will post the answer when tests are finished. To Sven : I went to the indicated address, typed "atlas" into the search engine and got only construction machinery (made by Atlas in 1/1 scale ! ) in HO by Kibri. So I think some problems with Atlas are still remaining. And meanwhile I did read about some serious contact problems when passing Märklin turnouts at low speed, for example in ... stations ! It's too bad, because the capacitor (with 1000 to 2200uF) discharge works well with only one polarity. And, sorry, I think Rokuhan Flex track is ultimately too expensive (about 57€/m ! ) My project in Z was an adaption of my former N layout, but : 1) One third would have been tunnels, so I fear some problems with cleaning the track. Somebody said to me (yesterday) that contact problems were a major problem with Z and that he gave up Z for that reason. And my tunnels would have been only accessible from the edge of the layout (even near the middle) ... No hatch possible, because I think one sees the "cuttings" in the landscape. 2) One "round" would have been 20m (= 4.5km ), so a freight train running at the scale speed of about 30miles/h would have needed about 6 min for one "turn" ; and this means 2mn of invisibility in tunnels, with only the running noise to indicate that there was no problem ! 3) The cost with Rokuhan flex track would have been about 1600 to 1700€ ; and, sorry, I am still thinking that Rokuhan flex track is not very beautiful ... or really ugly ! I find that somewhere it looks like a caterpillar (the animal, not the bulldozers). So I made a new design, according to the "KISS" principle (Keep It Simple and Stupid), the same size, 2.9m to 1.45m, a simple oval with only one station (my former design had 2 of them) with 6 turnouts, and one tunnel near the 1.45m edge, less than 1.2 m in length. The whole track would be Rokuhan (so no ballasting problems), standard tracks with only one section of flex in the tunnel (so it will be invisible) to "close" the oval, and the visible curves made with the R490 track sections ; so I can admire my trains running most of the time, and there are no major cleaning problems. One "round" will be about 6 to 7m, I think it's enough ! And the costs of about 600€ I calculated for the track, too (add about 1400€ for 4 trains, and then come the buildings, not to speak about the other deco elements). I saw one guy saying on german TV his Märklin layout in HO would cost about 60000€ ; no comment ! Such people destroy the miniature train hobby, more than the DCC problems . But this brings another problem : with flex track by Märklin, I could consider soldering a pair of feeders every 66cm (to the joiners). But now ? I can't sold feeders to all joiners ! What do You think about it ? Somebody wrote that no feeders are necessary in Z. What can be said about the "quality" of the contact at the joiners by Rokuhan ? Or are they absolutely unreliable ? If two joiners make a "bad contact" in the section between 2 feeders, there would be no solution ! And Murphy says ... Rokuhan says it would be better to have several feeders on the layout ... but not at each track section !
|
|