Bench Speedometer

[KenS]

This could just as easily go under tools, but since I'm using it for DCC decoder programming, I'll put the topic here.

 

I bought a Bachrus MTS-DCC40 speedometer for my DCC decoder programming, to help me set up speed tables.  To be honest, I bought this some time ago, and it's been sitting on a shelf waiting for me to get to the decoder work. I bought it through traintekllc.com, and cost today would be around US$135 with the cradle.

 

They make two versions of this, one standalone and one for use with a computer running the JMRI software.  I bought the latter, in the N-scale size (which they call "40 series").  To use it, you also need some additional rollers, and I bought a set of three (part number 40SA-3). Photo 1 shows my DE10 sitting in the cradle.

 

I'm using a Mac, and all their documentation is for Windows.  They support Macs (and linux, but I shudder to think how hard that would be to get working given the state of their documentation). But the folder on the CDROM just contains JMRI, with no explanation of how it works.  I finally figured out that I needed to use DecoderPro's Preferences to select the /dev/tty.usbmodem1d11 driver.  At that point, it started working. And it coexists happily with the LocoNet USB interface I use to connect to the Digitrax Zephyr command station, so I can program the loco using JMRI, either in Ops mode right on the cradle, or move it briefly over to the programming track to program there, and back to the cradle.

 

Actually setting the loco into the cradle is a bit tricky, as the sensor unit is unbalanced and heavy enough to want to fall off the track (I eventually taped it in place).

 

But once in place, it seems to work fairly well. The JMRI speedometer display has both mph and kph ranges, and you can set it for any of the usual modelling scales, including "Japanese N 1:150". Photo two shows the screen with the DE10 doing 40.7 kph on the rollers with the throttle set to position 19 (that's using an out-of-the-box Digitrax DN163K1D decoder I haven't yet programmed). The only thing I haven't done yet is time the loco around the test track with a stopwatch at throttle 19 and see if it's really doing 40.7 scale kph. I think that's tomorrow's project.

 

I haven't done any programming with it yet either, but that's my next project. I'm not anticipating any problems as, aside from the driver confusion, it "just works".

[cteno4]

Ken

 

How do you like the cradle? I just ordered that last week?

 

Jeff

[KenS]

Ehh, not enthused.  The non-sensor cradle parts work well enough, but they keep moving out of position or falling off the track as I try to put the loco into them. I've seen some photos that indicate you can connect them together with some kind of rod (but they don't sell one for the 40, even though I've seen photos of 40s with them; perhaps I just need a bit of brass tubing).

 

I've only just started playing with it tonight, so I don't have a lot of experience yet.  As rolling roads go, it does seem to sit on the track better than the last one I had, which was even worse at falling off the rails.  I don't seem to have much luck with these things (or maybe it's my sausage-fingers  ).

[CaptOblivious]

I recently ordered just the Bachrus cradles themselves (helps debugging decoders when your debugging cable doesn't have to follow a moving train!), and feel meh about them as well. Very difficult to secure to the track, and very difficult to get my locos to sit squarely in them. Despite all this, the loco seems to be pretty secure even at high speeds, so I'm not so concerned about these issues. Been thinking about that speedo, thanks for the review!

 

You can run a bit of brass tubing through the sides—doing this is supposed to help ensure proper alignment when you've got a bunch of them. It would be nice if the brackets were long enough to permit running the tubing down both sides, rather than just down one. But even then, for me, the hardest part is getting them centered under the wheels, and then keeping them from slipping out of place.

 

Still, I'm glad I have them, and they do work.

[cteno4]

cool to hear the bacharus 40 can get the rod. was looking at the pictures and wondering if some shims of wood on the outside of the track to help them rest there as well might help with tipping/slipping also.

 

track is not the greatest thing to sit these things on, its cute, but seems like a simple base would be better to deliver power and give more positive lock on the rollers. but thats what some of the really expensive rollers have so perhaps these guys are just keeping the price the lowest with the fewest parts.

 

looking forward to mine, will be nice to be able to run trains while working on them on the bench easily. i have like 20 trains that need some overhauling after years or jrm running and i keep putting them aside to

 

jeff

[CaptOblivious]

track is not the greatest thing to sit these things on, its cute, but seems like a simple base would be better to deliver power and give more positive lock on the rollers. but thats what some of the really expensive rollers have so perhaps these guys are just keeping the price the lowest with the fewest parts.

 

Well, it does serve a functional purpose—it ensures that the rollers are at the exact same gauge as the track the train normally runs on.

[cteno4]

Yes There is always that competition between stability and adjustability...

 

Can't wait to get mine this week to play!

 

Jeff

[Hobby Dreamer]

So Ken...

 

How fast does your bench go?  ;)

[KenS]

I just thought I'd add a follow-up note on the Bachrus: I've stopped using it due to inaccuracy.

 

The problem isn't the Bachrus itself, but the poor mounting of the sensor unit to the track, causing the light wheels of an EMU to make intermittant contact, particularly at higher speeds. And I wanted to accurately characterize speed for given speed-table settings, including high speeds for Shinkansen, so this kind of inaccuracy is unacceptable.  I've gone back to using a stop watch to time trains around a loop, but will eventually investigate one of the LED-based trackside sensors that can time a moving train.

 

It might work better with larger locos, but I bought it for use primarily with EMU motor cars, and have also had problems with the relatively small Kato DE10.

 

It's possible I could fix this with a purpose-built sensor track that holds the sensor and cradles firmly in place, and I might try building one with some sheet styrene eventually. But I think the root problem is that there is enough resistance in the sensor wheel that the light EMU motor car justs lifts itself off the wheel.  It may only happen due to the positioning issues with the cradle, but I'm suspicious that I'd still have that problem even if I fixed the cradle mount.

 

Disappointing.  I'd really hoped to get out of the stopwatch timing of trains using this.

 

However, I can report for HD that I've carefully timed the bench with my stopwatch, and can report its speed as 0 mph +/- 1% (that's 0 kph if you use metric).  It's important to remember to correct benchtop measurements for this factor, although mainly if you're working in a motor home or boat. Or live in California. :)

[inobu]

Ken,

 

LOL, I knew you would come to the same conclusion. Time base is the only way to go for accurate readings for us.

 

MPH/KPH is distance per hour and that system does not calculate distance traveled. There are other factors which plays an important role mainly load. The MTS is a tread mill and tread mills cannot factor true load. This is why some people can run 10 miles on a tread mill and barely do 5 on the ground.

 

The wheels are free spinning and the revolution are counted but that means nothing. Case in point. On a sheet of ice your wheel can be spinning 20mph and the car would be moving only 3mph. The actual load on the motor is nil and the motor could spin the wheels to 100mph instantly. Place the car on pavement and it will take a few seconds or minutes to get to 100mph.  

 

The IR speed readers count distance traveled in a period of time and converts it to MPH.

 

A unit pulling 3 car at half throttle can travel 50mph add another 5 cars and it will not travel 50mph at half throttle because of the added load. The IR speed device it a stop watch.

 

Another issue is this unit gives theoretical speed not actual as it does not take into account the efforts of the rear drive train. You can make the comparison with front wheel drive verses all wheel drive. 4 wheels pulling verses only 2 effects the load which equates to speed traveled.

 

Inobu.

 

I think I had a post made a while back on speed measuring. I agree with the stop watch method. 

 

 

 

[KenS]

That's a good point.  I hadn't really considered the load/no-load aspect, and it's definitely going to affect the usefulness of the Bachrus numbers, even if I could get them to report correctly.

[inobu]

I was almost certain that I did a write up on this MRT Accutrack

 

One can be purchased MRT AccuTrack

 

 

 

It is pretty simple the only down fall is its look.

 

The way it works is the train enters from either side and breaks the IR beam. That starts the timer as the train exits it breaks the second beam stopping the counter and the speed is calculated. There is a switch that sets the scale from N to HO.  I was able to get it down to the tenth of a mph and up to a few hundred miles by breaking the beam manually. 

 

Mine failed after a few weeks. Instead of sending it back I dug into it myself (good excuse to tear it apart) It turned out to be that the transmit IR LED was bad. I got a hold of Tom Accutrack creator and he told me that there was a bad lot of IR LED he used in one release he has since changed them out. He said he would warranty the unit as long as there was no abuse to it. I have not had a problem with it since. 

 

A DIY By G.W. Robertson, March 2011

http://mysite.verizon.net/reszs0px/

Edited January 6, 2013 by inobu

[cteno4]

On the JRM layout we have used the speedometer system by a company who's name escapes me right now that uses two photo sensors mounted in the track. its worked very nicely and accurately. only issues we had was under some interior lighting situations (ie the cow warehouse at the fairgrounds) it would not operate. we got newer sensors that worked much better, but then the system got broken and no one has gotten on it to repair here lately. it was getting torn apart all the time on the old layout so it got a lot of abuse its not built for. newer layout the wires are in a more hard wired space, but something broke at some point and not sure exactly what. 

 

the accutrack is nice as it has its own light source. we thought of putting the sensors in the tunnel and providing overhead leds, but its on a curve and complicates the spacing a tiny bit also the ceiling is w/in the mountain which comes off so the wiring and power source need to get built into the mountain for this.

 

having good speed is important. folks always want to run things too fast! shinkansens let you get away with it a little, but even then we had folks running things at 500kph at times!

 

jeff

[Densha]

having good speed is important. folks always want to run things too fast! shinkansens let you get away with it a little, but even then we had folks running things at 500kph at times!

I always ask the question to myself why manufacturers actually make the motor to go so fast, I'd rather have they top around the maximum scaled down speed. Shunter locos don't go 300Kph either!

[cteno4]

Basically they make them so that usually around 1/2 - 2/3 of full voltage the are running around max prototypical speed, getting it closer would be hard. it would be very difficult to get them made so that they would run at their top prototypical speed at say 12 or 14v (but then different power packs have different max voltages too). You would need a voltage reducing circuit in each engine (similar to what is in the kato portram) tuned to limit the top speed resulting from the gearing and motor speeds in hat model. This is what dcc gets you with your power curve adjustment..

 

Otherwise it's figuring out what the top speed should and just developing an eye and checking it with a stopwatch now and then or set up a speedometer.

 

jeff

[Zeppelin driver]

I was almost certain that I did a write up on this MRT Accutrack

 

One can be purchased MRT AccuTrack

 

 

 

It is pretty simple the only down fall is its look.

 

The way it works is the train enters from either side and breaks the IR beam. That starts the timer as the train exits it breaks the second beam stopping the counter and the speed is calculated. There is a switch that sets the scale from N to HO.  I was able to get it down to the tenth of a mph and up to a few hundred miles by breaking the beam manually. 

 

Mine failed after a few weeks. Instead of sending it back I dug into it myself (good excuse to tear it apart) It turned out to be that the transmit IR LED was bad. I got a hold of Tom Accutrack creator and he told me that there was a bad lot of IR LED he used in one release he has since changed them out. He said he would warranty the unit as long as there was no abuse to it. I have not had a problem with it since. 

 

A DIY By G.W. Robertson, March 2011

http://mysite.verizon.net/reszs0px/

Went ahead and got an Accutrack. A fun side toy when playing with my trains ..err when 'engaging in model operations !' :)

Thanks for the lead on this device,

Howard