Fix the loco wheels, they should meet the standards gauge, and then you shouldn't have any problem.
People love to complain about Atlas turnouts not meeting NMRA standards, but EVERY single time I have had a loco or car roll roughly through an Atlas turnout, it was proven that the wheels were out of gauge. Fixing the wheels fixed the problem, EVERY time.
And my standards on trackwork are pretty high - my most difficult locomotive has to work forwards and backwards along all paths at warp speed far in excess of normal operating speeds, both along and when pulling AND pushing various cars. If at any point there is a derailment - the trackwork is not good enough and I fix it. This is predicated on the test loco and cars all having properly gauged wheels per the NMRA gauge.
Now, a Snap-Track turnout is an 18" radius, which is the absolute minimum, and really pushing it for 6 axle diesels and the steam locos MTH makes in HO, even with their special adaptations for sharp curves. So there certainly is an argument for replacing them - but you will find similar bits of the Custom Line turnouts not matching the NMRA gauge as well.
And in another bit of harkening back to the past - this sounds somewhat similar to an old problem Cuda Ken had when trying to use a 5 amp MRC or Bachmann booster on his layout - the short circuit protection was less than adequate and allowed things to fry. MTH might be explained by the fact that their stuff was originally designed for 3 rail O, which draws a lot more current than HO locos.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
There really isn't anything more to be said. You summarized it pretty well.
It is hard to say which, or both, are responsible, the snap track turnouts or the MTH locos.
I agree with you that (1) don't run MTH locos on those unmodified turnouts or (2) replace the Atlas snap track turnouts with Atlas Custom-Line turnouts.
Rich
Alton Junction
Folks, a quick update -- while I was waiting for the two MTH HO locos to be fixed, I eliminated some of the gaps on the figure 8--because they seemed to be implicated in the shorts/failures in the 'death zone' of my layout--and migrated the reversing section to the outer loop.
I also added 2-Amp quick-blow inline fuses to each of the track circuits to protect the engines.
I got the engines back two weeks ago and, running with the quick-blow fuses revealed that the locos were experiencing arcing/shorting on just about every Atlas 540 and 541 turnout on my layout.
This was probably happening all along but it was not enough to fry the PS3 boards; but with the fuses in place it became clear that, if the locos were moving slowly, the arcing & shorts were causing current spikes high enough to blow the fuses I'd installed.
After a few blown fuses, close inspection revealed scorching & pitting on the 540/541 closure rails where they converge next to the isolated frogs. When traversing a turnout, the back of the wheel flange was close enough to the inside of the closure rail to allow arcing; even the back of the wheel flanges show pitting.
It's possible that all along the frequent 'undetected' shorts were causing cumulative damage to the boards, and that the arcing that occured on the insulated gaps in the 'death zone' was enough to trigger the outright failures.
The Walther/Shinohara #4 curved turnouts and Atlas 562 #4 turnouts show no evidence of the same problem--but they have measurably wider gaps between the converging closure rails.
Aside from the too-narrow gap on the 540 and 541 turnouts, I measured the MTH HO wheel sets on my GP-35 and F3 locos, and am sorry to report that according to my Vernier calipers they do not appear to meet NMRA standards for RP-25 wheels; they are too close together for either Fine:HO or minimal Proto standards. This would explain why they have a problem with the narrow gaps on the 540/541 turnouts' closure rails.
So, in my opinion:
Anyone running MTH HO locos on a layout with power supplied by a DCS Commander should not use unmodified Atlas 540/541 turnouts on their layout. Nor vice-versa.
In my opinion the products are incompatible out of the box for the reasons outlined above. I'm not running my MTH DCS HO locos on my layout again until I've dealt with the 540/541 turnouts. I suppose my options are:
1-File away a portion of the inside of the curved closure rail to widen the gap;
2-Coat the inside of the curved closure rail with enamel to prevent arcing/shorts;
-or-
3-Change power routing to the curved closure rail to eliminate the shorts (difficult, as I'm not currently using switch motors capable of power control...).
Other suggestions welcome. Thanks again for reading.
At least you (and we) now know that the layout works as expected with that DPDT manually controlling the reversing section. That pretty much also eliminates the 3-way turnout and that turnout below it as culprits.
richhotrain OK, that would seem to make the Bachmann AR unit suspect, but then, as I recall, you were frying decoders before you installed the AR unit. Without being glib, my advice is to dump the MTH Z-1000 and the DCS Commander on eBay and use the proceeds to buy a starter DCC system, Digitrax or NCE. Keep the Bachmann AR unit for the reversing section. Those two DCS locos will run on DCC. Add a decoder to the DC loco that you bought for testing purposes. That's what I would do. Rich
OK, that would seem to make the Bachmann AR unit suspect, but then, as I recall, you were frying decoders before you installed the AR unit.
Without being glib, my advice is to dump the MTH Z-1000 and the DCS Commander on eBay and use the proceeds to buy a starter DCC system, Digitrax or NCE. Keep the Bachmann AR unit for the reversing section. Those two DCS locos will run on DCC. Add a decoder to the DC loco that you bought for testing purposes. That's what I would do.
Rich --
Yes, the first two of the three failures occured before the AR unit was installed, so I know it's not the AR.
I know the exact position of the engines for the first and third of the three failures--with a wheelset from the rear truck stopped exactly on the gap--and the other one may have been on that very same spot. So yes, something about the combination of that insulating gap, the Z-1000, the DCS Commander, and the PS3 boards is deadly.
I appreciate the guidance re: DCC instead of DCS. I will give it serious consideration.
Regards,
-Rick
I tried it first hard-wired (swapping alternately to match the polarity with the inner and outer loops) and then with a DPDT switch to control polarity. No issues.
Did you remove the Bachmann AR and replace it with a DPDT?
richhotrain OK, we gave you four weeks to conduct the test. What happened? Rich
Rich,
In four or five hours of operation I experienced zero problems with the DC engine running in either direction in forward or reverse over/through the 'death zone', stopping on the gap, etc.; no shorts, no fault found.
The DC engine also crosses the NE DXO in any direction without stopping.
So I'm still stumpted about the cause of the problem.
A little over a week ago I spoke on the phone with one of MTH's techs and sent him the drawings linked in my previous posts, but have not heard back from him regarding what aspect of the layout might be blowing the DCS PS3 boards.
I will get the HO DCS engines repaired, add the 2A circuit protection to the track power, stagger the 'death zone' gaps, sacrifice a goat and then maybe--just maybe--try again.
zstripe Rich, What test? I didn't take no test. Respectfully! The Shadow
What test? I didn't take no test.
Respectfully!
The Shadow
We don't need no stinkin' tests.
rallison Bought an inexpensive DC loco and will proceed to test after swapping out other bits. Will post an update after testing with the DC loco.
Bought an inexpensive DC loco and will proceed to test after swapping out other bits.
Will post an update after testing with the DC loco.
Also: I put the GP-35 and one boxcar on the unpowered track, positioning them as they were when the third engine failure happened (after a derailment of the following boxcar) in order to determine exactly where the loco's wheels were when it died. It turns out that, where it was halted by the derailment, the lead wheelset of the rear truck of the loco was smack on top of that insulated gap in the 'death zone'.
So, for at least two of the three failures, a wheelset from the rear truck was stopped bang on the gap when the loco died. Hmmmmm...
Rick, I am going to take one last shot at this issue before we close it off.
- Your two locos negotiate both outer loops without any problems, as would be expected since there are no apparent trouble points.
- Your two locos negotiate the figure 8 without any problems, which is somewhat surprising since the "death zone" is within the figure 8.
- The NW and NE DXOs seem to work fine except for the loss of power when exiting the NE DXO.
So, if this were my layout, here is what I would do.
- Remove the 3-way turnout and the Atlas snap switch below the 3-way turnout and replace those turnouts with flex track.
- Disconnect the Bachmann AR unit and replace it with a DPDT toggle switch to manually control the reversing section.
- Purchase a cheap DC locomotive and run it into and out of the figure 8, using the DPDT to control the polarities in the reversing section. You can find Atlas HO scale DC diesels at MB Klein for under $40.
In this way, a thorough test of the entire layout can be conducted without putting those repaired MTH locos at any further risk.
If all goes well, add the 3-way turnout back into place and do the testing all over again.
If all goes well, add the Atlas snap switch back into place and do the testing all over again.
If all goes well, report back to us before attempting to run the repaired MTH locos on the layout
That's what I would do at this point.
Or at least beating a model of a late-era iron horse. But yes, Rich, agreed that this thread is played out and next steps are as discussed above.
No shorts occured when test-running the complete figure 8 while it was still powered by jumpers. But that's in line with the previous and subesquent paragraphs in that post, as the locos ran over the 'death zone' section without incident many times before/between each event, so I don't think there's any learning to be had from that.
Agreed, NE DXO issue is probably a lack of power brought about by the interaction between the distance between insulated frogs/crossings and the distance separating power take-up wheels on the loco trucks. I can probably confirm that just by putting the dead loco on unpowered track at that point on the problem headings, and looking at where the wheels are.
Thanks again to all responders. No need to reply further unless new possibilities/thoughts occur; I will provide an update if and when the trouble is well and truly shot.
rallison And FWIW both engines ran through the 'death zone' many times (when the soon-to-be-reversing section was still hard-wired with the jumpers) before the each of the first two engine failures described in problem #1. And that is very perplexing. After completing the figure 8 but before installing the AR unit I ran an engine around the 8 numerous times, and across the NW/upper-left DXO--all without issue. But did any shorts occur? And again after installing the AR, the repaired GP-35 ran successfully on all three main sections, operating without issue for a number of hours over several days before the car derailment that led to the third engine failure. And that is very perplexing. (The only limitation was that it could not traverse the NE DXO at upper right from figure 8 to outer oval without stalling, as described in problem #2.) If, indeed, it was a stall, not a short, that clearly suggests a lack of power.
And FWIW both engines ran through the 'death zone' many times (when the soon-to-be-reversing section was still hard-wired with the jumpers) before the each of the first two engine failures described in problem #1. And that is very perplexing.
richhotrain [...] Were you ever able to run engines around the outer oval without any problems? Were you ever able to run engines around the inner oval without any problems? Besides removing the 3-way and that one Atlas snap switch, another test would be to remove the AR unit as well and see if you can run an engine around the figure 8. If an engine could successfully run those 3 routes (outer oval, inner oval, and figure 8) a number of times, that would at least allow you to hone in on the problem. [...]
[...]
Were you ever able to run engines around the outer oval without any problems?
Were you ever able to run engines around the inner oval without any problems?
Besides removing the 3-way and that one Atlas snap switch, another test would be to remove the AR unit as well and see if you can run an engine around the figure 8.
If an engine could successfully run those 3 routes (outer oval, inner oval, and figure 8) a number of times, that would at least allow you to hone in on the problem.
Of the three scenarios, #2 seems the least plausible. I would rule that one out.
#3 is the one that I suggested earlier in this thread when I posted that photo of the 3-way turnout. If a metal wheel bridges those two rails and just sits there, you would expect a circuit breaker in the booster or controller or power pack to shut down the layout. If the circuit breaker doesn't cut power, I could see it frying the PS3 board.
#1 is certainly a possibility, but I am skeptical of intermittent shorts on turnouts. Either they work or they don't.
I would normally suspect the AR unit, especially one like the Bachmann which is slow to react to a short, or so they say. But, then, you were experiencing shorts without the AR unit wired in place.
I don't really see where you have much choice but to remove the 3-way and that lower Atlas snap switch and test the figure 8 that way.
But, as I see it, you are caught between a rock and a hard place, having already fried a couple of PS3 boards. So, you are going to need nerves of steel to run those repaired engines in DCS mode.
Here is another thought. If you want to test that Bachmann AR unit to be sure that it is operating correctly, you could run the lighted caboose back and forth through the reversing section. A simple 12 volt incandescent lamp could be used to verify normal polarity versus reverse polarity, instead of relying on your ear to hear the clicking.
But, this thread has already been running for a week, and you are no closer to resolution of either electrical issue. What you ought to do is buy a cheap DC loco, disable the AR unit, and run the layout in DC mode.
You need to start eliminating some of the potential issues.
1. Intermittent short in the Walthers/Shinohara 3-Way Turnout OR the Atlas #541 snap-track turnout at the bottom of the left lobe of the figure 8. Of the two, I actually think the Atlas is the more likely source, because of something recently remembered: After the completion of the auto-reversing section and the addition of the Bachmann AR unit, I could hear the AR unit ‘trip’ when the loco entered the reversing trackfrom either end and from the NE DXO. It recently occurred to me that sometimes I heard it trip even when the entire train was already on the reversing section. It would happen when the train had entered the East/right lobe of the figure 8 coming from the outer loop, and was descending the curve of that lobe clockwise. I could hear the ‘click’ happen at about the point when the loco was on the Atlas turnout. Why would it trip the AR unit there, unless there is a short? So I may pull and check that turnout before yanking up the more complicated 3WT. If there IS a short in the Atlas turnout or the small curve between it and the 3WT, it might explain the failures when the engine was stopped on or straddling the insulated gap, regardless of the status of the AR unit. I think I was pretty good about sweeping away construction debris before installing these sections, but it’s possible I missed a small wire or rail clipping—or it could be a manufacturing defect .
2. Power imbalance between the hardwired feed to the whole spur/yard 3WT section, and the feed to the reversing section. Just a wild hunch, but: As best as I can remember, in all three failures the loco was on or straddling the gap. In the first, it stopped with one wheel-set right on the gap, and there was a hefty spark as the engine halted & died. In the second and third failures, one truck was on one side of the gap, and the trailing truck on the other side, AND the train was physically halted by the derailed box car—meaning the engine was pulling/pushing under load, with effort limited only by the friction of the wheels on the rails. In the first and second events, power was fed to the future reversing section by jumpers and alligator clips. SO: If the contacts were not solid, perhaps it led to an imbalance in voltage/amperage that, given the power draw of the halted but still running engine, caused an arc in the PS3 board circuitry? And perhaps in the third event (with the AR unit in place) the circuitry of the AR unit also changed/limited the power draw on that side of the gap, again leading to an imbalance? Seems a bit out there, but I am grasping at straws…
3. Proximity short near the 3WT frogs. I saw an entry on a wiring website that noted some shorts occur near insulated frogs if metal wheels touch both the through and diverging rails where they almost meet, downstream from the frog. This doesn’t seem likely unless I am mis-remembering where the locos were when the problems occurred—but I can’t entirely rule that out, as I could be mis-remembering.
So thanks again Rich, and everyone else who chimed in, for your helpful thoughts. I’m going to check each of these out before putting engine to track again.
He had looked at it, as of a day or two ago, and he seconded the suggestion (which I think was yours) about pulling up the 3-Way and inserting a single piece of track in its place and then using a lighted caboose to confirm AR function and determine if the 3WT was the cause. Beyond that, he too is stumped.
You ought to ask that layout designer to review this thread and then give you his thoughts on the entire matter.
richhotrain [...] I have not read anything that suggests that DCS engines do not like loops or that you should alternate the gaps and feeders such as you have done. Can you point us to any articles that discuss this issue? Rich
I have not read anything that suggests that DCS engines do not like loops or that you should alternate the gaps and feeders such as you have done.
Can you point us to any articles that discuss this issue?
That information came from a layout designer who frequently works with MTH DCS in HO as well as the bigger scales. I contacted him directly because I'd read his posts on another forum/thread about DCS in HO and auto-reverse. He advised that DCS does not like complete electrical loops, and that a limited number of feeders--perhaps as few as one for the main layout and one for the AR section--would suffice for DCS, as the control signal is handled differently than in DCS, where it rides on the power fluctuations. He also told me that his company had better experiences (on DCS layouts) with the Bachmann AR unit than with most of the others. Later he recommended a more complicated AR setup involving sensors, relays, and stall-motor switch machines so as to anticipate the necessity of a polarity swap rather than react quickly to a short, but for a little layout like mine -- and my lunch-money budget -- I stuck with the recommended Bachmann unit.
The alternating gaps and feeders on the inner and outer loops were my own compromise, as the gaps in the W/S DXO at bottom center are built in, and I didn't want to cut the outer rail at the same point. Electrically I believe they should be no different staggered the way they are as long as it's on a section that is NOT auto-reversing.
-- Rick
Agreed about the (eventually) repaired MTH locos--I'm not letting them go anywhere near the death zone until I am sure I have found and eliminated the problem.
I'm thinking of picking up a pair of cheap standard DC engines at the LHS and doing any further on-track testing with them. The DCS Commander can output a standard DC signal with voltage controlled throttle.
richhotrain [...] Are you sure that the stopped loco exiting the NE DXO is a stall and not a short?
Are you sure that the stopped loco exiting the NE DXO is a stall and not a short?
I'm not experienced enough to be 100% sure. After the second engine failure, I tended to operate with one finger hovering over the DCS Commander's emergency 'Stop' button, so when an engine halted unexpectedly I was pretty quick to push the button to kill power to the track. I'm fairly sure I did so each of the three times the engine stalled traversing the NW DXO from the figure 8 to the outer loop. What I can say for certain is that I did not hear or see a spark, that neither the Z-1000's nor the DCS Commander's breaker tripped, and that the PS3 board did not fry.
richhotrain I keep going back to Problem #1 when there was no AR installed, no completed reversing section, and all matched polarities. Yet, the loco shorted and the PS3 board fried. So, what does that indicate? A bad 3-way turnout? A faulty Atlas snap turnout (perhaps the one below the 3-way)? A bad DCS controller? A faulty Z-1000? Is the TIU acting up?
I keep going back to Problem #1 when there was no AR installed, no completed reversing section, and all matched polarities. Yet, the loco shorted and the PS3 board fried.
So, what does that indicate? A bad 3-way turnout? A faulty Atlas snap turnout (perhaps the one below the 3-way)? A bad DCS controller? A faulty Z-1000? Is the TIU acting up?
In that order:
Not sure what it indicates; yes, I haven't yet ruled out a problem with the 3WT or trapped metal debris underneath it causing an intermittent short across the jumpers on the underside; I am starting to think that a problem with the Atlas turnout 'beneath' the 3WT is a more likely problem*; yes, it could be a bad DCS controller, but then again it works fine on the rest of the layout at all times; I don't think the power supply is goofy, as it too works fine when operating on the rest of the layout; but I am 100% certain that it is not a TIU problem (because the DCS Commander is a simplified 'all-in-one' DCS unilt -- there IS no TIU)!
richhotrain Crossed feeders (sorry, Rick, gotta ask)?
Crossed feeders (sorry, Rick, gotta ask)?
No need to apologize--at first that seemed the most likely cause to me too. But as mentioned upthread, I checked and re-checked polarity--against the plan, and by following color-coded wire, and by using the voltmeter. I am quite sure that (NORMALLY*) polarity is good.
richhotrain A partially closed gap?
A partially closed gap?
If there's a problem at the insulated gaps, then I'm going to get stinkin' rich for figuring out how to get plastic to conduct electricity. I used Atlas #552 joiners on all the indicated gaps.
richhotrain Is something amiss on that NW DXO?
Is something amiss on that NW DXO?
The NW and NE DXOs are built the same, other than being mirror images. The NW DXO is even insulated the same way, but has hard-wired matching polarity; the NE DXO has the Bachmann AR feeding power to the lower half. The NW DXO works in all cases, the NE DXO works in every case except when crossing from the Figure 8 to the outer loop. Very odd indeed.
richhotrain Is it a bad batch of PS3 boards in Rick's locos?
Is it a bad batch of PS3 boards in Rick's locos?
This is also possible, but again, they operated just fine on every other part of the layout.
I think we're all stumped. I WISH it had been a simple/stupid mistake, easily diagnosed and corrected. I'd rather be embarrased than still uncertain...
--Rick
*I have a couple of other possibilities that have occured to mewhile chewing on this whole thread... I'll discuss them in a later post.
rallison --The two gaps at center bottom are built in to the Wathers/Shinohara DXO, and the two odd ones at the top mirror those gaps on the opposite rail at the top. Granted, it's an odd arrangement; but, along with the built-in ones in the DXO at bottom, it prevents each of the outer and inner tracks from forming a complete uninterrupted electrical loop because I've been told that DCS engines do not like loops.
rallison Rich -- Yes; I expect once I get the engines repaired I'll replicate the 'stall' to confirm it's a lack of power at particular parts of the NE DXO. Then add temporary juice as needed to correct, as you suggest, & then make permanent whatever power addition fixes it. I just didn't mention it because I can't do much about it 'til I have engines to run over the DXO...
Rich -- Yes; I expect once I get the engines repaired I'll replicate the 'stall' to confirm it's a lack of power at particular parts of the NE DXO. Then add temporary juice as needed to correct, as you suggest, & then make permanent whatever power addition fixes it. I just didn't mention it because I can't do much about it 'til I have engines to run over the DXO...
Thanks, Rick, for posting that pic.
I'm stumped. It all seems right, everything on your layout. Yet, there are so many things that could be wrong. (see my questions below).
I am disappointed that no one else has jumped in here with some thoughts or suggestions.
So, what does that indicate? A bad 3-way turnout? A faulty Atlas snap turnout (perhaps the one below the 3-way)?
A bad DCS controller? A faulty Z-1000? Is the TIU acting up?
As I say, it all seems so right, yet, so many things could be wrong.
...and just for fun here's a pic of the Walthers/Shinohara 3-Way Turnout portion of the layout, showing polarity. The joins at the three routes are soldered, and the join at the bottom is insulated; the section below that is now connected to the Auto-Reverse unit, but used to be hard-wired via jumpers to match the polarity of the 3WT.
Sure, Rich. The pic is posted here. I only ask that no-one make derisive comments about my sloppy newbie track-work--I'll clean up the ties later...
Yes, the DXO consists of two Atlas #541 at upper and lower left, one Atlas #540 at lower right, one Walthers/Shinohara #6.5 LCT at upper right, all connected to an Atlas #575 45-deg. crossing, and with short bits of Atlas #500 flex track thrown in as needed.
All joints (except the insulated gaps between the #575 and the #541 and #540 turnouts at bottom) are soldered. Frogs are unmodified/insulated. The roadbed in this section is foamcore board and is not yet ballasted. There is an Atlas #65 under-table switch machine controlling the W/S #6.5 LCT.
As mentioned earlier, this is (structurally) a mirror image of the DXO at upper left/NW on the layout, which works correctly in all directions... the only real difference being the reversing section below the insulated gaps on this one.