Fox Valley wheels - with resistors

[Cajonpassfan]

 When I have read different articles on putting resister wheels sets in a cra it is usually recommended to put 2 wheels in them, one in each truck.

Richard

Well, I'm trying to avoid that account that would take cca 800 wheelsets to cca 1600
I'll experiment when we get home from the holidays...

[GaryHinshaw]

I finally got my first block detection circuit wired up yesterday and I was able to do some experimentation with sensitivity and reliability.  I'm using the RR-CirKits Watchman boards and started with the default sensitivity setting.  It detected a loco with no problem (I have to admit that I cast a joyous fist pump when my panel first lit up ).  But it was having trouble detecting two hoppers with newly-installed FVM 10K wheels, so I cranked up the sensitivity and the "debounce" time and was getting slightly better results.  I then cleaned the wheels with alcohol and ran them out and back around the layout (shockingly this is now an 11 scale mile round trip, counting the helix).  Upon returning to the block, the cars were detected very reliably while rolling, but as soon as they were parked, the detection would fail. 

So mixed success so far.  One of my intended applications is detection of occupied staging tracks with parked cars and no locos, so I hope this issue can be licked.  I'll report back after some more experimentation.

[Cajonpassfan]

....  I then cleaned the wheels with alcohol and ran them out and back around the layout (shockingly this is now an 11 scale mile round trip, counting the helix).  Upon returning to the block, the cars were detected very reliably while rolling, but as soon as they were parked, the detection would fail. 

So mixed success so far.  One of my intended applications is detection of occupied staging tracks with parked cars and no locos, so I hope this issue can be licked.  I'll report back after some more experimentation.

Yeah Gary, my results exactly. And weighing the cars (within reason) doesn't seem to make any difference. Now, my situation is different in that I don't plan to have CTC and short OS sections; my prototype was all TT&TO operation; all I need to have is occupancy detection for whole trains with cabooses and only a few blocks shorter than a standard train length. Makes me wonder whether I just rely on locos and cabooses for detection; equipping 800 cars seems like a lot of trouble to cover a handful of short blocks...
Otto

[jagged ben]

I created a handful of resistor wheelsets with FVM wheels using shaved graphite (pencil lead) and acrylic paint.  It's not a solution for every car (although Otto, it might work for you) because:

• it's too time consuming
• the results are too unreliable
• some of the resistances came out too low for a whole train e.g. 500 ohms

But, it has worked awesome for cabooses with two of these wheelsets.

Point is, either 10k is too high, which you can probably adjust sensitivity for on the rr circkits stuff at least, or there's something about the tread coating on the FVM resistor wheels that differs from previous runs of regular wheels.  I did have to scratch away the coating on the ones I did where I applied the paint.

[peteski]

I created a handful of resistor wheelsets with FVM wheels using shaved graphite (pencil lead) and acrylic paint.  It's not a solution for every car
...

Point is, either 10k is too high, which you can probably adjust sensitivity for on the rr circkits stuff at least, or there's something about the tread coating on the FVM resistor wheels that differs from previous runs of regular wheels.  I did have to scratch away the coating on the ones I did where I applied the paint.

You seem to be familiar with electronics: did you happen to measure the resistance of your home-brew resistor wheels?  Less than 10K?

[jagged ben]

You seem to be familiar with electronics: did you happen to measure the resistance of your home-brew resistor wheels?  Less than 10K?

Of course.   I had to know if they would work.   Duh. 

As I said, it was pretty difficult to get consistent results.  If I recall, most of them came in between around 2k and 4k after I futzed with them enough.   It was really difficult to get above 5k or so and actually still have a resistive connection.     Sometimes I'd get nothin' (open circuit) and have to chip the paint off and try again.  Sometimes they'd come in too low, e.g. under 500 ohms, in which case I worried about the heat dissipation and I'd redo those, too.

What's much harder to know is how stable they've been over time.  I figure if they heat up too much I'm much more likely to loose the connection entirely than to develop a less resistive short.   But none of them have had more than a few hours run-time, so I can't tell you much about long-term reliability, except that as far as I know they all still work.   I did about a dozen total. 

[peteski]

Of course.   I had to know if they would work.   Duh. 

As I said, it was pretty difficult to get consistent results.  If I recall, most of them came in between around 2k and 4k after I futzed with them enough.   It was really difficult to get above 5k or so and actually still have a resistive connection.     Sometimes I'd get nothin' (open circuit) and have to chip the paint off and try again.  Sometimes they'd come in too low, e.g. under 500 ohms, in which case I worried about the heat dissipation and I'd redo those, too.

What's much harder to know is how stable they've been over time.  I figure if they heat up too much I'm much more likely to loose the connection entirely than to develop a less resistive short.   But none of them have had more than a few hours run-time, so I can't tell you much about long-term reliability, except that as far as I know they all still work.   I did about a dozen total.

...

Point is, either 10k is too high, which you can probably adjust sensitivity for on the rr circkits stuff at least, or there's something about the tread coating on the FVM resistor wheels that differs from previous runs of regular wheels.  I did have to scratch away the coating on the ones I did where I applied the paint.

Thanks!
You might be onto something with the non-conductive wheel darkening layer.  Kato's American Freedom Train GS-4's drivers used unusually dark coating on the drivers, and whatever that was, it was non-conductive.  I had to remove the darkening layer in order for the wheels to pickup electricity.  I had a thread about the GS-4 issue on the A-board.

[GaryHinshaw]

I'll report back after some more experimentation.

I got the ohmmeter out over the weekend and took some measurements to see if I could tell what was going on with my car detection issues.  If I put the ohmmeter probes directly on the treads of a factory-installed FVM wheel with 10K resistor I got a reading of 10K ohms to within a few percent, so that suggests that the tread is a pretty decent conductor.

Next I equipped 10 cars with a single resistive axle per car and put them on an isolated test track and put the ohmmeter probes on the two rails of the test track so that the circuit was completed by the 10 resistive wheel sets across the rails.  Ideally, I should measure a resistance of 10K/10 = 1000 ohms but I was getting results that varied between open and 12K ohms.  To be more specific, I pushed the cut of cars along the track and let them come to rest where they may and then took a reading.  I would either get a reading of open, or a stable reading of 12 K ohms.  At first I was getting 12K ohms in about half the trials and open in the other half.  Next I cleaned the rails with alcohol and the fraction of good readings went up to maybe 75% (but the resistance itself did not drop much below 12K ohms).

So far, I conclude that the wheel-rail contact must have a pretty high intrinsic resistance.  (That result was probably obvious, but I was surprised by the repeatability of the 12K ohm result even after cleaning the rails.)

Next tests I want to try:
* Measure N identical cars at a time and see if the measured resistance goes like 1/N
* Measure a single car with N weights placed on the car.
* Hook up my RRampmeter (when it gets here) so I can measure DCC current reliably.  I'll repeat these tests with the track power and detection circuitry hooked up so I can see what the detector sees.

[peteski]

Gary, you you have a Dremel tool with a wire brush?  If yes, could you use it to remove the blackening from the wheel treads on one of the wheelsets (like I did in the GS-4 thread I liked to in my last post)?

I suspect that the layer of blackening might be causing the problem. When you place a multimeter probe on the tread that probably punctures through the layer of blacking and you get reliable resistance reading.  But when the wheel rests on track, the blackening is not penetrated reliably cussing intermittent resistance reading.

[Cajonpassfan]

Guys, I'm not one with much electrical knowhow, so correct me if I'm wrong, but it shouldn't matter whether there is one wheelset or ten across the rails; current will flow through the path of least resistance and that's going to be close to 10K Ohms, no? So the reading Gary's getting (about 12K) seems about right, except when it's zero and we are back to the original "lack of conductivity" problem. One would think that with ten cars, at least one axle at a time would provide the necessary connection, but apparently it doesn't. And these are not even blackened wheel sets... odd...
Otto K.

[GaryHinshaw]

I suspect that the layer of blackening might be causing the problem.

Bingo!  Good call peteski.  (It really had to be that, but it's nice to see it verified.)

I didn't have a wire wheel, but I did have an emory impregnated stone wheel for the Dremel and that worked like a champ: take a few quick passes on the tread while the wheel spins in the truck.  It was amazing how much blackening there really was when you see the bare metal  by comparison.  I put the first car back on the track and I get 10K across the rails.    The second car came in about the same, and the two together came in at 5K.  Physics works, and so do these wheel sets!

So polish the treads for the best reliability.  As a fringe benefit, the polished treads look great!

Guys, I'm not one with much electrical knowhow, so correct me if I'm wrong, but it shouldn't matter whether there is one wheelset or ten across the rails; current will flow through the path of least resistance and that's going to be close to 10K Ohms, no?.

That is not quite right Otto.  If there are two paths of equal resistance (like two wheel sets) the current will split evenly between them and the effective resistance will be half of either path.   Think of it like having two lanes available for traffic - half the resistance to traffic flow because the electrons (cars) naturally find the best lane.

-gfh

[peteski]

Bingo!  Good call peteski.  (It really had to be that, but it's nice to see it verified.)

I didn't have a wire wheel, but I did have an emory impregnated stone wheel for the Dremel and that worked like a champ: take a few quick passes on the tread while the wheel spins in the truck.  It was amazing how much blackening there really was when you see the bare metal  by comparison.  I put the first car back on the track and I get 10K across the rails.    The second car came in about the same, and the two together came in at 5K.  Physics works, and so do these wheel sets!

So polish the treads for the best reliability.  As a fringe benefit, the polished treads look great!

Excellent!

Polishing 1200 wheel treads can't be much fun.  I'm pretty sure that the blackening will also wear away after some running time on the wheelsets (I've seen it worn down on locomotive wheels).  But I don't know how much running time that will take.  I'm also not sure if a layer of (possibly non-conductive) metal oxide will develop even on cleaned treads, if the car wasn't ran for days or weeks.

[Cajonpassfan]

That is not quite right Otto.  If there are two paths of equal resistance (like two wheel sets) the current will split evenly between them and the effective resistance will be half of either path.   Think of it like having two lanes available for traffic - half the resistance to traffic flow because the electrons (cars) naturally find the best lane.

-gfh

Thanks for the enlightenment.... and for finding the problem. The threads *look like* they are not blackened, but obviously there is some residue on them...
Otto

[peteski]

The threads *look like* they are not blackened, but obviously there is some residue on them...
Otto

I have never encountered any metal model wheel (from any manufacturer) selectively blackened. All the chemically-blackened (tinted) metal wheels I have dealt with (including FVM) have the blackening on all the metal surfaces.  I suspect that they simply dip all the metal parts in some sort of blackening solution (before assembly). It would be pretty much impossible to just blacken the face and the back without blackening the tread area.  But it does wear off in use, so higher-mileage wheels will have bright colored tread area.

[jagged ben]

Considering that the wheels are actually made from brass, that tells you how much of a difference the blackening makes.

It's unfortunate they seem to have overdone the blackening with the resistor wheels as compared to the earlier runs I used to make my homebrew wheelsets.