Questions about lack of wheel conductivity and metals

[svedblen]

Oh.. wheel gauge.  Yes, it is correct.  I have checked it many times.  It's dead-center in the notches on an NMRA plate.

I'll just say this, then I'll shut up     and give up this line of thinking:
I did not doubt that your trucks are spot on. On the contrary. I was rather thinking that the track might be gauged at the tight side at the problem spot. And the same (tight gauge) for the Kato engine that ran well (or is that unit something other than a switcher and thus able to have at least one set of trucks outside of the problem area?).

[Maletrain]

Maletrain:
To address your concern about the flanges.  My home-made wheels have pretty fine-scale flanges.  They are smaller than Kato's and definitely do not bottom out on the ties.   But you are right about one thing, the Kato tender truck flanges do
graze Atlas code 55 spikeheads unless they are sanded down a little, which mine are.

Pushing down on a wheel with stick makes the engine move, but it's very hard to do that without physically jarring
or otherwise moving the tender, ever so slightly.    So even though the stick makes the engine move, it could easily be
making the axle points touch the insides of the cones, for example.  It wouldn't tell me that the problem is lost contact between wheel and rail.  Touching the wheel with a wire whisker, that is not connected to anything, and seeing the engine not move, and then connecting the other end of the whisker to the rail, then touching the wheel, and seeing it move,
and doing this same test on different wheels dozens of times, is awfully convincing evidence that the wheels
are not picking up from the rails.

I think you have proven that the rail and wheels are not making contact.  So, I am working from that.  Part of the problem with the tests, so far, is that the proof of the lost contact (electrified whisker touching wheel without jostling anything) destroys the condition that you are tying test.  How about testing as follows?  With the engine stalled on the track, do your unpowered whisker test to make sure that it is in the condition that you think it is in.  Then, disconnect the track power so that the engine will not move when and destroy your test condition when you get wheel-to-rail continuity.  Now, connect the whisker to a continuity tester, and the other lead of that tester to the rail.  Check for continuity again with the whisker.  If you still don't have continuity, then try to move the wheel slightly down to the rail to see if that creates continuity.  If so, does it lose continuity when you release the downward pressure?  If that does not show us anything, then try putting pressure on the wheel to force it away from the electrified rail.  If that doesn't show us anything, try pushing the wheelset from the other side to force it against the electified rail.

Another question that would help us understand better is whether this problem loco is running slower than the other locos that you say run well throgh the same tracks.

[jeffstri]

I assume the super-flex wire is silicone rubber insulated, correct? That's certainly the most flexible you could use but the way it's routed could still cause one side or the other of the tender to be lifted when the engine bounces. When this happens, the short wheel base of the tender makes complete loss of contact on one side more likely than with a longer tender. I think any tendency of the wires to lift one side of the tender when the engine bounces will be reduced if you eliminate the damping block and allow to wires to move freely up and down through the tender floor. Also, the routing of the wires looks nice (simulating air/water lines) but in the only steamer I had with low loops like that, going through the bottom of the tender, they interfered with the draw bar and caused intermittent loss of contact which was eliminated when I rerouted the wires higher, above the floor of the tender and straight across.

However, I think the main problem is the lack of electrical pickup from the engine drivers. All my early Kato mikados had the same slow speed intermittent stalling problem, that could not be fixed by anything I did to the tender, until I replaced the drivers with the "revised" ones.

[Cajonpassfan]

[quote author=mmagliaro link=topic=39602.msg484509#msg484509 date=1468702929

To answer some questions:
The draw bar is very free.  It is not lifting the tender.
I did try adding 30g of extra weight and that did not help.

Max, adding weight to the tender *should* help, but an ounce spread over eight wheels is only 7 grams each, not really that much. Have you tried loading up the tender externally to see if it helps? Without the need for a decoder or sound, you should be able to add a lot more weight? (In my experience, if tender is free-rolling, the extra weight shouldn't affect the pull as much as one would think).
Beautiful engine BTW...
Otto K.

Max, I just realized my math was off; do'h. Adding 30 grams over 8 wheels makes it only 3.75 grams each, close to nothing. So my rationale still stands, just double as much
Otto K.

[u18b]

As was previously suggested - I'd add a LOT of weight without the shell.

See if that makes a difference- it should.

[peteski]

Max,
Judging by the responses you are getting, several members do not know the history of that loco. Maybe you should explain that it is scratch-built in-progress loco and if I understand it correctly, you are currently only using tender (not the loco) for power pickup.

I also say to add more weight. If you don't think that all the wheels are supporting the loco's weight, if you add enough weight, maybe the truck side-frames will flex enough to get all the wheels on the track?  I also don't recall if you temporarily substituted Kato trucks and whether that solved the problem?  If no, maybe you should try that?  Or maybe substitute one truck at a time and see how the loco behaves?

[mmagliaro]

Max,
Judging by the responses you are getting, several members do not know the history of that loco. Maybe you should explain that it is scratch-built in-progress loco and if I understand it correctly, you are currently only using tender (not the loco) for power pickup.

I also say to add more weight. If you don't think that all the wheels are supporting the loco's weight, if you add enough weight, maybe the truck side-frames will flex enough to get all the wheels on the track?  I also don't recall if you temporarily substituted Kato trucks and whether that solved the problem?  If no, maybe you should try that?  Or maybe substitute one truck at a time and see how the loco behaves?

The current state is that is does get pickup from 2 of the drivers, but the pickup isn't good enough to run the loco without the tender.  I did try it with Kato trucks, and also with my trucks + Kato's wheelsets, and finally with my trucks and my wheelsets.
In all cases, the behavior is about the same.

I found a slight misalignment between the sideframes on one truck today - where the wheels on one side would be all shifted about .010" forward compared to the ones on the left.  i.e. the axles would actually be rolling along at a very slight non-perpendicular cant to the rails.  That never seemed to bother its free-rolling properties (i.e. just rolling the truck on the track by itself, it rolled very smoothly and rolled a long way without stopping, at least as good as a Kato truck).  But regardless, that misalignment isn't a good thing, so I corrected it.

Then I added 15g more weight stacked on top for a total of 45g.  So far, it has made it around the layout past all the places where it traditionally stalled before.  I will repeat the test a few times and then reduce the weight to see if the problem returns.

[Lemosteam]

Max, since you dfound a fore/aft shift in the sideframe, could it be that there is also a side view twist?  This could potentially lift one wheel on one side and the opposite corner wheel of on the other.

Also, have you tried to wire up only one truck (no chassis) to the loco at time with no weight, then you could add weight, etc.?

Since the screw holding the truck on takes the same axis as the tapped hole- could it be that the two bolster screw axes are not parallel in the front view of the tender frame?  The screw head could be lifting one side off ever so slightly.  This could also be true of the two planes that the truck bolster mates to.

[narrowminded]

What is the total weight, not just the added weight?  And while the trucks swivel and swing freely, do they make contact or derive their support from a point that may be off center to the wheel that could have all wheels contacting but due to lever fulcrum, not with equal weight distribution?  Possibly very light at some point?  If you remember back when, I had found in testing on my project that 5 grams per wheel seemed to be the minimum weight that would pretend to pick up power with reasonable reliability on clean track and that 7.5 per wheel was pretty much optimum, any extra being bonus.  That would suggest that a minimum weight of 40 grams total, evenly distributed, or as much as 60 grams total preferred, reliable pick up should be there.  I feel your frustration.

[u18b]

Max,

Having recently dealt with issues of placing bearings in Bachmann Amfleet cars, I encountered a problem with one of the cars derailing.

It turned out when I installed the bearings, I was not keeping the height of each bearing the same.
The end result was that the wheels were not all perfectly laying on the track.  And since the Kato wheels are fairly fine scale, the small flanges were not enough to keep it on track.

While your situation could not be nearly so bad..... I'm wondering since this IS scratchbuilt.

Are your bolsters perfectly even and identical?

It seems to me that if something were a little off, you might lose some contact with the rail.  The phosphor bronze springs would never be an issue since they are always touching.

This is a puzzler.

[mmagliaro]

I'm sure it is highly likely that the 4 wheels of each truck are *not* perfectly aligned with the same pressure on each one.
When the trucks sit on the track, they roll beautifully, and I certainly don't have enough error to see, nor do the trucks "rock" on their corners (like a table with one leg too short).

But yes, with all the tedium of making scratchbuilt brass/nickel/BeCU truck frames with cone point axle holes, I'm sure they are not perfect.

So.... 45 grams was the ticket.  With that much extra weight stacked on top, it runs perfectly - never stalls.  If I take the extra off and try it with 30g of extra weight, the problem comes back.  The problem is that there is no way on earth I can fit 45g of additional weight inside this tender - even using tungsten.  So I need to hone my trucks.   As you are suggesting,
it probably is indeed an issue of balancing the weight more evenly on all the wheels. 

I'll weigh the tender later and get back to you with a number on how heavy it is, in total.

[jdcolombo]

I'm sure it is highly likely that the 4 wheels of each truck are *not* perfectly aligned with the same pressure on each one.
When the trucks sit on the track, they roll beautifully, and I certainly don't have enough error to see, nor do the trucks "rock" on their corners (like a table with one leg too short).

But yes, with all the tedium of making scratchbuilt brass/nickel/BeCU truck frames with cone point axle holes, I'm sure they are not perfect.

So.... 45 grams was the ticket.  With that much extra weight stacked on top, it runs perfectly - never stalls.  If I take the extra off and try it with 30g of extra weight, the problem comes back.  The problem is that there is no way on earth I can fit 45g of additional weight inside this tender - even using tungsten.  So I need to hone my trucks.   As you are suggesting,
it probably is indeed an issue of balancing the weight more evenly on all the wheels. 

I'll weigh the tender later and get back to you with a number on how heavy it is, in total.

Hi again, Max.

This may sound crazy, but if you need weight, how about making the tender floor out of tungsten?  Not sure what the tender floor is for you now, but whatever it is (even if it's the Bachmann metal floor), tungsten almost certainly will add weight . . .

John C.

[mmagliaro]

Remake the tender floor?  From Tungsten?     

Yes, I supposed that's possible.  But you see that tender in the opening photo?  It was not exactly a short project to machine that thing out of brass, and making it from a plate of tungsten would be almost impossible.  Cutting the shape, the slots for the pickup strips, and such would be agony.
But you are right in that it would add a lot of weight to the tender.  Tungsten is more than double the density of brass.

-----
New data.   I slipped thin little pieces of paper in between the "thumbs" and wiper strips in the floor, for only one truck, and ran it.  That effectively runs it only only one truck's pickups without affecting the geometry.  I repeated the test for the other truck.  Somebody suggested trying to run it on only one truck, and I liked that idea so this is how I did it.

The idea was to find out if one truck was working significantly better than the other (i.e. maybe 90% of the time, it is running on one truck and the other one is lousy?)

But the tests showed that each truck is about the same.  The engine stalls more freuquently with either truck not conducting, and increase in failures isn't any different for either truck.  But this was a good idea and is useful data.
And yes, I realize this doesn't tell me about how the individual wheels in each truck are doing.

[peteski]

You know Max, I'm really puzzled and surprised with just how fussy and unreliable that all-wheel-pickup tender is.  It doesn't seem to make sense.  Do you think that a stock Kato Mikado tender (with its original weight) is as unreliable as your tender?  I don't recall if you tried using it for testing or not.

One solution (some can call "cheating") would be to use a DCC decoder with a simple keep alive circuit which would make this loco run for few seconds without reliable power pickup. They do work.    I know, it is not to be.

[mmagliaro]

I assume the super-flex wire is silicone rubber insulated, correct? That's certainly the most flexible you could use but the way it's routed could still cause one side or the other of the tender to be lifted when the engine bounces. When this happens, the short wheel base of the tender makes complete loss of contact on one side more likely than with a longer tender. I think any tendency of the wires to lift one side of the tender when the engine bounces will be reduced if you eliminate the damping block and allow to wires to move freely up and down through the tender floor. Also, the routing of the wires looks nice (simulating air/water lines) but in the only steamer I had with low loops like that, going through the bottom of the tender, they interfered with the draw bar and caused intermittent loss of contact which was eliminated when I rerouted the wires higher, above the floor of the tender and straight across.

However, I think the main problem is the lack of electrical pickup from the engine drivers. All my early Kato mikados had the same slow speed intermittent stalling problem, that could not be fixed by anything I did to the tender, until I replaced the drivers with the "revised" ones.

You win a blue star!
Shockingly, the flexible wires between engine and tender were a large part of the problem.   The photo I showed at the top of this thread has the wires with nice flexible loops.  But what I overlooked was that in the last few days, as I've fussed around with this thing, I broke off and had to resolder the wires twice.  Each time, I had to snip a little off the broken end before resoldering.

I *though* there was enough slack to allow for that, but I was wrong.  One wire was definitely getting too taut and was pulling on the corner of the tender.   I released some more slack from inside the tender, and it ran hugely better.  In fact, it was able to
make it through most of my trouble test points with no added weight at all.

Next up, I added smaller amounts of weight.  Somewhere between 15g and 30g, the magic happens, and it never stalls now.
That's way better than needing 45g to make it work.

narrowminded:  You asked about the actual weight of the tender without the added weight.   It is 31g.  So with 25g added,
that would get be 56, your experimentally determined number for 8 wheels (7g per wheel x 8 = 56) for very reliable pickup.
How interesting that I now find experimentally that my magic number is between a total of 31 and 60,
or between about 4 and 7.5g per wheel.

So now... I am going work on neatly adding weight to see if I can hit the 56g number.