Resistance wheelsets

[CRL]

Since I will be running cabooses on all trains, I plan on only putting resistors on the caboose wheels… probably on 2 axles. I think this should be sufficient as train length will be limited to about 6’ maximum.

[GaryHinshaw]

I use CTC on my layout, and the Control Point turnouts are separately detected.   In that scenario, it's useful to have every car detectable so the dispatcher can't inadvertently throw a turnout under the train.  (Occupied turnouts are locked.  You can guess how I learned this lesson.)  A 10 K resistor on one axle per car generally works fine for this, as long as the wheel treads are clean.  I've also configured a 3 second delay in the transition from occupied to clear as a 'keep alive' buffer.  Very helpful for short blocks like this.

[C855B]

... A different brand of detector may need different resistance to work reliably. ...

As mentioned, 10K has been the accepted standard for detection design since transistorized block detection was invented in the 1960s. It has become a design point. It may be possible that some detectors will function with more resistance (more Ω = less parasitic mA), but 10K per car should work for everybody's commercially-sold circuit. What has significantly improved detection in the past 30 years has been DCC with its constant voltage across the rails; DC was hit-or-miss since voltage was frequently too low at slow throttle settings to generate enough current flow.

This discussion has given me considerable motivation to change my thinking to Digitrax' 22KΩ recommendation where there are two axles per car with resistors. The 22K on one axle is probably just enough to trigger recent detector designs, which then mitigates momentary continuity drops from one wheelset on the car while still meeting the design objective of 10K per car. Win-win, just a little more effort in the number of "custom" wheelsets. I'd only have, oh, about 2000 axles to modify this way. 

[C855B]

Mike:
OK, 10K it is. Any thoughts on 1/8 or 1/4 watt?

Like Pete said, as small as you want to go, wattage is a "don't care" with this extra-low-current application. 1206 SMD "chip" resistors (0.12" x 0.06") can range from 1/20W to 1/2W, all of which are fine for axle shunts. That's what I use. If you're a masochist, 0805 (0.08" x 0.05") will work just as well.

[Cajonpassfan]

Since I will be running cabooses on all trains, I plan on only putting resistors on the caboose wheels… probably on 2 axles. I think this should be sufficient as train length will be limited to about 6’ maximum.

That's my plan also. My trains are much longer, but still shorter than the shortest block, which is what matters. In my pre- CTC era that's all that's needed, no short "control points".
Otto K.

[CRL]

I use CTC on my layout, and the Control Point turnouts are separately detected.   In that scenario, it's useful to have every car detectable so the dispatcher can't inadvertently throw a turnout under the train.  (Occupied turnouts are locked.  You can guess how I learned this lesson.)  A 10 K resistor on one axle per car generally works fine for this, as long as the wheel treads are clean.  I've also configured a 3 second delay in the transition from occupied to clear as a 'keep alive' buffer.  Very helpful for short blocks like this.

Thanks for your comment about turnout detection. The RRC Traincontroller program is designed to work with only the non-turnout blocks being detected, but they use a train length calculation of some kind… if I’m understanding it correctly.

[John]

Thanks for your comment about turnout detection. The RRC Traincontroller program is designed to work with only the non-turnout blocks being detected, but they use a train length calculation of some kind… if I’m understanding it correctly.

Not familiar with the Traincontroller program.  I'm using CATS at this time -- but saw the same issue with regular JMRI panels -- when I didn't have detectors on the control points, the logic would go bonkers and loose track of the train ..  so I went back and retrofitted the layout with the additional mainline detectors ..

[nkalanaga]

A question for the experts:  Why not use conductive plastic for the insulators in the wheels.  The stuff we use at work is barely conductive, in the megohms range, but it can be found in different resistances, and the resistance would also depend on the design.  Ours is a good-sized chunk, and the conductivity is for static dissipation.  The plastic molds like any other, and is tough enough for diesel fuel filter modules, so should work for wheel insulators.

It wouldn't raise the cost enough to affect the list price of the wheelsets, and would eliminate external resistors.

[C855B]

Better yet, formulate a conductive UV resin for our Photons, etc., and we can roll our own.

[Maletrain]

Better yet, formulate a conductive UV resin for our Photons, etc., and we can roll our own.

Not so sure that a 3D printed wheel would roll very well.  You would probably need to at least turn each wheel tread surface in a lathe, or at least a drill, to make them smooth enough to roll well.

[jagged ben]

A question for the experts:  Why not use conductive plastic for the insulators in the wheels.  The stuff we use at work is barely conductive, in the megohms range, but it can be found in different resistances, and the resistance would also depend on the design.  Ours is a good-sized chunk, and the conductivity is for static dissipation.  The plastic molds like any other, and is tough enough for diesel fuel filter modules, so should work for wheel insulators.

It wouldn't raise the cost enough to affect the list price of the wheelsets, and would eliminate external resistors.

This.  I've said this before. 

[jagged ben]

Not so sure that a 3D printed wheel would roll very well.  You would probably need to at least turn each wheel tread surface in a lathe, or at least a drill, to make them smooth enough to roll well.

He's only talking about the insulators, not the wheel or axle.   Although making the whole axle the resistor might also be an idea.

[peteski]

Using conductive plastic for the wheel insulator is IMO probably the most realistic (simplest and easiest to implement) idea, even as a retrofit for existing wheelsets.

[GaryHinshaw]

This is a really nice idea.  I would probably aim for 20K per axle (or less), but there would be many advantages (assuming the cost is really not significant):

* More reliable - every axle being live,
* More realistic - no more unsightly resistors behind the wheels, especially noticeable on tank cars,
* Less inventory - no need to maintain a separate product line (like FVM does).

Also, imagine if we could convince manufacturers to stick to a single (0.540") axle length.  Then the only inventory that would be needed is wheel size, and maybe tread width.  Maybe we could convince ESM to pursue this ( @bbussey ).  Ahh to dream. 

BTW, I think 3d printing these would be mechanically feasible, but the electrical properties would probably be challenging to dial in and control.

[jagged ben]

Just putting this out there as well...

I've made my own resistive paint just by mixing pencil graphite and modeling acrylic.   On the one hand, it's been difficult with this method to dial in the resistivity level.  I've gotten anywhere from 500 to 2000 ohm, which is too low for a whole train of cars.  On the other hand, I've been far from scientific about it.  I bet if I bought commercial graphite powder or lubricant, and was deliberate about mixing ratios of paint to grahpite at scale, I could dial in a formula.    As much as I like the insulator idea, a good paint formula would be teaching men to fish instead of bringing them fish, for most of us.