I am curious about getting speed info from my locos and cabs.
I am using DCC Digitrax but have seen an ad for MRC Prodigy (back of MR August) showing speed “126”. I am assuming that this means speed step 126 and not actual scale speed.
Question - do any of the DCC cabs register scale speed? I think not, and thus #2, is there any other product available on the market besides the TDP Trainspeed units?
Cheers, Jake
[%-)]
Jake,
Actually, the QSI decoders themselves have a verbal or audible speed confirmation when you press F10(?) from your DCC throttle.
The quick and dirty way to measure speed is to measure a distance on your layout and create a time/speed comparison chart in whatever scale you are in. Time the locomotive from point A to point B, find the time on the chart and see what the corresponding speed is.
I believe Circuitron makes a car with a LCD digital readout for viewing speed.
Tom
I saw a club layout at a show. They were using radio control DCC cabs that showed scale speed, but I don’t remember what brand they were. Also, I’ve seen a circuit with IR LED’s that mounts under the track and shows the scale speed of trains passing over it. (again, no link)
How can this be any more than an indication of how fast the motor is going? And without knowing gear ratios and even wheel diameter is only proportional to the speed, and not the speed itself. Do they have a way to program in these factors? I guess it wouldn’t be too hard, but it would be an extra step. I have always thought that measuring the time between two fixed points, and getting used to the ‘feel’ at certain speeds would generally do the trick.
Jeff,
I believe the QSI decoders are configured for the individual “type” locomotive they are installed into. Therefore, a BLI 2-8-2 Mike running at speed step 060 will have a different SMPH (scale MPH) than a Proto 2000 0-8-0 running at speed step 060. It would be part of the specific decoder chip. That’s why the upgrades chips are not generic but specific to each locomotive type. When ordering, you have to specify which locomotive you want the upgrade chip for.
I agree with you, Jeff. Personally, I think the time/distance measurement is more fun and realistic than a verbal SMPH or having to running an MPH “car” amongst your rolling stock. But, that’s just me…
Tom
Tom-
That makes sense. I figured it would have to be loco-type specific, and they’d need to be able to know the motor speed, too. And lots of things get involved there, unless you just read it. Even back EMF won’t tell you how fast it is going, just how it compares to what you are commanding, I think. It seems like a lot of work for somehting that probably isn’t accurate anyway (I am prepared to eat crow on that, but for now it is my opinion).
This is an interesting topic and it also is a good segway to how I speed match my locomotives.
I have a ‘scale’ speed meter and I program all my locomotives’ speed curves so that if I have ‘50’ selected on my controller (50 not = to 50 percent, but rather the 50th speed step) it is equal to 50 scale KPH (I am Canadian). It isnt hard to do and it elimiates the ‘golden locomotive’ way of speed matching locomotives. Also, you can set speed limits to certain parts of the track and the throttle will allow you to set all your locos to the correct speed through these limits.
This also allows you to speed match steamers and modern Gevos…
JMRI is your friend.
David B
Most of the posts have identified the important points. To measure speed accurately in the locomotive with DCC, the following have to occur:
the decoder must use a tach or cam on a drive axle, and have a time base. Or if using BEMF or a flywheel tach, must have a time base, and “know” the gear ratio.
the decoder has to have a way of transmitting the speed info through the rails back to the DCC system
the DCC system must be capable of understanding the speed data packets and displaying them appropriately
AFAIK, this would all be outside the NMRA DCC spec, and thus would be a proprietary extension. This means the chances of different brand decoders having speed read by any given DCC system is not high.
Without setting up a speed measuring section of track, or pulling a speed measuring car, you can estimate scale speed within 3-4 MPH by counting cars passing a point in 5 seconds. This works much better if you cars are fairly consistent in length and some reasonable divisor of 12" in actual length.
60 smph = 10 6" long cars in 5 seconds (HO 40ft box cars are 6" long)
9 cars in 5 seconds = 54 SMPH
7 cars in 5 seconds = 42 SMPH
5 cars in 5 seconds = 30 SMPH
4 cars in 5 seconds = 24 SMPH
3 cars in 5 seconds = 18 SMPH
2 cars in 5 seconds = 12 SMPH
1 car in 5 seconds = 6 SMPH
Note how slow a realistic switching speed is - one to two 40ft cars passing the turnout frog in 5 seconds.
A steam drag freight over a grade would be 2-4 40ft cars passing the summit marker post in 5 seconds. A train of 20 ore (4" long) cars should take at least 30 seconds to pass that summit marker.
A loop on a 4x8 typically has about 20ft of track. A train at 30 SMPH should take 40 seconds to go around the loop.
I dare say very few of us consistently run at realistic scale speeds. We tend to abuse the excessive acceleration and deceleration rates our models are capab
Hey, davidmbedard, I’m planning on using your method when the layout is up and running, about another month or so. How are you compensating for the length of the train, or doesn’t it make a difference? I’ll be running single 4 axle locomotives with at most 7 cars.
As an addition to fwright’s post, something I learned from an employee timetable of the Staten Island Rapid Transit line, which did switching for the B&O on Staten Island in NYC. A 40’ car will pass a fixed point in 1 second at 28 mph, a 50’ car at 35mph. So your 40’ cars need to take longer than 7 sec. to pass a fixed spot for safe switching. That ought to stretch out the ol’ operating sessions. Ken
Actually, that’s a great way to do SMPH. I hadn’t even thought of that. Quick and easy. And, as long as you knew it’s particular length, you could do the same thing with just your locomotive.
Tom
My old layout used both methods for speed - I worked out charts (just happened my mainline “loop” was almost exactly 1 km around) and had a couple BLI engines with the built-in speed announcement. From what I could tell the BLI output was accurate.
BTW I generally run below prototype speed - 30-35 for passenger trains or fast freights, 18-25 MPH for freights, down to 12-15 MPH for iron ore trains. Keep in mind if you converted a 30" radius HO curve to real railroading, it would have a severe speed restriction on it due to the tightness of the curve - so I figure trains look better going slow around those curves.
I agree, we mostly all are guilty at one time or another of overdoing the speed for the scale. I have tried to be realistic, and now rarely raise my DT400 throttle setting over 40, even for limited passenger trains. You get to see them longer between tunnels and copses of woods and buildings, and they probably run better if trackage is less than perfect.
I would like to point out that in the Weekend Photo Fun thread two weekends ago, Sue and Larry (gearjammer) had a rather excellent video rendition of their mallet doing a highly prototypical struggle with a drag of freight. Having several DVDs with archival footage of these beasts shot with 16 mm cameras, I think they got it right. I make my PRR J1 do the same on my 3% grades.
Prototype speeds were checked via mile markers (mile posts)and number of poles passed (X number of poles in a mile)with a accurate watch. The engeeneer would note in the engine log, “Repair or replace spedo, is fast” and other repairs he noticed that needed attention. I do the same as my telephone poles or strike fence poles are spaced at one foot intervals. One foot at one sec is 60 MPH and 2 seconds is 30, and so on. After a time one can mentally count off the seconds fairey accuratley as you run the train. Works for me and enjoy doing it anywhere on the layout…John
To compute scale speed from the speed step set in to the decoder, you have to multply the speed step by a scale factor, something which micro processors (both the controller and the decoder have micro processors in them) can do with ease. If marketing thought that a scale speed readout on a DCC controller would sell, engineering would furnish one. But if they did, I’m not sure I’d beleive in it’s accuracy, because no one can tell how accurate the scale factor in side the micro processors really is. I’d want to check speed with a stop watch and a ruler.
Stop watch time measurement accuracy is limited by human reaction time. It takes about 3/4 of a second to see something and react to it. So hand stop watch time measurement has a error of a fraction of a second built into it. For good results you want to make the time fairly long so the error is small. One lap around the main line is long enough to make the time errors neglible.
I measure the length of my mainline with a carpenter’s steel tape and then measure the length of time it takes the train to do a lap with a stop watch. Then with a hand calculator and remembering 5280 foot per mile and 12 inches per foot and 60 seconds to the minute and 60 minutes to the hour and perhaps some highschool algebra and pencil and paper, I compute the locomotive speed. For example a stock Mantua Pacific had a to