In Train Surgery we have been discussing gearing and scale speeds,
and I cobbled together a small program to measure what our little trains are doing.
For those who dont do surgery and gears etc. I thought I'd talk about it here with an improved cleaned up Mk2 version.
What you do is choose two markers 120mm apart (60ft in real life ~= 1 mk1 coach lengh) on the layout,
and go to my web page (link below) where there is a little calculator with appropriate buttons.
there just press the green button once when the front of the loco passes the first marker and then again at the second.
And prepare to be surprised :) !
once over your surprise press the red reset button and go again to check it , simples :)
I dont know how accurate we need to be with this sort of thing, not very I suspect, however it would be more accurate and perhaps easier to use if I made the measure twice as big ?
Anyway, try it out if you wish and let me know ( useless or otherwise :) all comments welcome :) ) what you think.
Go here to my web page in my Dropbox (https://dl.dropboxusercontent.com/u/81786082/NG/Speedo/SpedoMk2.htm) to try it out.
Hi malc.... Good bit of work! When I use the trainspeed program my markers are on my longest straight section of track which is about 13ft....this gives better accuracy as you suggest as with a very short distance you have to be lightning quick on the start and stop buttons !
Quote from: sparky on June 29, 2015, 07:27:52 PM
lightning quick on the start and stop buttons !
Thanks Sparky.
Indeed !
That is where the surprise comes in - when one finds that one's dearly beloved shunter has been doing 90mph for too long lol!
I think a Mk3 version is needed where a selection of marker distances can be chosen first, I see some 'radio' buttons in the offing,
or even a Mk4 in which any ol distance one fancies can be input,
then of course there is the Mk5, a real circular dial mit zee traditional Morris Minor pointer - I have seen some java code on the interwebby for a graphic like that, dont understand all the code yet tho'. This could, as Cpt Oats once said, take some time !!
Thanks for looking and your input, very interesting.
Quote from: MalcolmAL on June 29, 2015, 07:13:57 PM
In Train Surgery we have been discussing gearing and scale speeds,
and I cobbled together a small program to measure what our little trains are doing.
For those who dont do surgery and gears etc. I thought I'd talk about it here with an improved cleaned up Mk2 version.
What you do is choose two markers 120mm apart (60ft in real life ~= 1 mk1 coach lengh) on the layout,
and go to my web page (link below) where there is a little calculator with appropriate buttons.
there just press the green button once when the front of the loco passes the first marker and then again at the second.
And prepare to be surprised :) !
once over your surprise press the red reset button and go again to check it , simples :)
I dont know how accurate we need to be with this sort of thing, not very I suspect, however it would be more accurate and perhaps easier to use if I made the measure twice as big ?
Anyway, try it out if you wish and let me know ( useless or otherwise :) all comments welcome :) ) what you think.
Go here to my web page in my Dropbox (https://dl.dropboxusercontent.com/u/81786082/NG/Speedo/SpedoMk2.htm) to try it out.
I'm gonna give that a try just out of curiosity. You're full of surprises Malcolm!!
:beers:
Quote from: austinbob on June 29, 2015, 08:02:37 PMYou're full of surprises Malcolm!!
:beers:
:laugh:
as the actress said ;)
Those of a nervous technical phobia disposition look away now : - -
I did have a bit of bother with a div by 0 as it started its run throwing up an infinity, seem to have got over that now,
but there is still an oddity that if the speed calculation happens to be dead-on a whole number then the decimal '.0' is not being shown, no great deal but a bit odd.
I wonder if it is a "feature" of the Math.round() library function ?
Hmmm, this is going to bother me !!
Quote from: MalcolmAL on June 29, 2015, 08:21:45 PM
Quote from: austinbob on June 29, 2015, 08:02:37 PMYou're full of surprises Malcolm!!
:beers:
:laugh:
as the actress said ;)
Those of a nervous technical phobia disposition look away now : - -
I did have a bit of bother with a div by 0 as it started its run throwing up an infinity, seem to have got over that now,
but there is still an oddity that if the speed calculation happens to be dead-on a whole number then the decimal '.0' is not being shown, no great deal but a bit odd.
I wonder if it is a "feature" of the Math.round() library function ?
Hmmm, this is going to bother me !!
You mean you didn't do a proper QC on your product release before you sent it out to us poor non-tech users. - I am sooo! disappointed.
:angel:
Quote from: austinbob on June 29, 2015, 08:26:11 PM
You mean you didn't do a proper QC on your product release before you sent it out to us poor non-tech users. - I am sooo! disappointed.
:angel:
Oh sorry :(
Did you not get my tweet in which I promoted you to Chief of my QC dept. ?
Tut, Twitter is geting to be a waste of time
:laugh:
Quote from: MalcolmAL on June 29, 2015, 08:30:58 PM
Quote from: austinbob on June 29, 2015, 08:26:11 PM
You mean you didn't do a proper QC on your product release before you sent it out to us poor non-tech users. - I am sooo! disappointed.
:angel:
Oh sorry :(
Did you not get my tweet in which I promoted you to Chief of my QC dept. ?
Tut, Twitter is geting to be a waste of time
:laugh:
How much does it pay??? Malcolm?
:D
A slightly lower tech idea....
See how far your train goes in 8 seconds. The distance it has covered in inches = the scale speed in MPH.
(Or in metric, time 3 sec and the distance it has gone in cm is the scale speed in MPH). Rough and ready, but maybe good enough for government work. And the beauty of this is that you can use it when you are a punter at exhibitions to estimate the speed of trains on other people's layouts and give them helpful advice! :worried:
Cheers Jon :)
QuoteHow much does it pay???
I dunno, youngsters these days , , :doh:
you pay me for the privileged position it gives in society >:D
Quoteand give them helpful advice!
:laughabovepost:
Yep, we done the coach lenght magic rope trick and other methods in Surgery, but they wuz arksin for 'proper' electronic methods, so I did this as a stocking filler till someone made with the Arduinos uc ;)
Quote from: PostModN66 on June 29, 2015, 08:39:59 PM
See how far your train goes in 8 seconds. The distance it has covered in inches = the scale speed in MPH.
Very handy :thankyousign:
Paul
Hmm... interesting approach. I tried an alternate way of solving the problem, making a downloadable (Mac & Windows) application (http://brackishfaq.webspace.virginmedia.com/Programs/scalespeedcalcul.html) instead of a web app.
I pretty much left the user to do the timing aspect (something you implement quite nicely in the web app). On the other hand, I think letting users select the scale they're using provides some extra utility. One challenge is that not all users want to use miles per hour, so I fiddled the output a bit to provide km/h as well as mph.
Cheers, NeMo
(http://brackishfaq.webspace.virginmedia.com/Programs/Resources/Scale_Speed_Calculator_mac.jpg)
Quote from: MalcolmAL on June 29, 2015, 08:21:45 PM
I did have a bit of bother with a div by 0 as it started its run throwing up an infinity, seem to have got over that now,
but there is still an oddity that if the speed calculation happens to be dead-on a whole number then the decimal '.0' is not being shown, no great deal but a bit odd.
I wonder if it is a "feature" of the Math.round() library function ?
Hmmm, this is going to bother me !!
I presume you have used Excel to make this speedometer?
If you change the format of the cell to a "General Number" it will show two decimal places :thumbsup:
cheers John.
Quote from: NeMo on June 29, 2015, 08:59:53 PM
I pretty much left the user to do the timing aspect
,
,
so I fiddled the output a bit to provide km/h as well as mph.
Thanks NeMo, interesting, downloaded will look later.
Is the Windows freestanding ie.portable .exe or does it mess with my registry, I not used Realbasic, just many of the others Yabasic Liberty etc. but moved on to 'C'.
It was the timing that they wanted to be done for them in Surgery with trackside gizmos etc :) I thought pushing a button was next best worsest :) !
You are telepathic, I toyed with kph and was going to make some remark about newfangled metrics lol !
Would be easy to do if there is a general
wail demand.
QuoteI presume you have used Excel
No Excel, just raw javascript in Notepad text ed. then stuffed into the HTML webpage (which was also notepaded )
Quote from: MalcolmAL on June 29, 2015, 09:28:08 PM
Thanks NeMo, interesting, downloaded will look later.
Is the Windows freestanding ie.portable .exe or does it mess with my registry, I not used Realbasic, just many of the others Yabasic Liberty etc. but moved on to 'C'.
Should be totally freestanding. Causes no problems on my Windows XP installation.
Quote from: MalcolmAL on June 29, 2015, 09:28:08 PM
You are telepathic, I toyed with kph and was going to make some remark about newfangled metrics lol!
But then you'd have to worry about European N scale, Japanese N scale... where do you draw the line???
Cheers, NeMo
Quote from: NeMo on June 29, 2015, 09:34:20 PM
Should be totally freestanding. Causes no problems on my Windows XP installation.
,
where do you draw the line???
Thanks
,
yep, exactly so, then it gets so big that it needs a manual as well and peeps complain they cant understand it all :laugh3:
I think I'll leave it simple like this a while, then maybe do a version with radio buttons to choose between maybe 4 length options.
I dunno if anyone noticed ! but last night I removed my magnum opus,
intending to replace with a different version to take account of the length vs. accuracy issue.
Well I suffered a little computational difficulty ;) aka incompetence ! but I think I am now proud to announce a new version ( we are only up to ver. 8g !!! )
In which you can choose either 120mm, 240mm or 480mm.
(there may be a variable one along later, but I am allergic to variable variables in experiments, maybe I am (borderline?) OCD ? lol! )
play here (https://dl.dropboxusercontent.com/u/81786082/NG/Speedo/st8g.htm)
Quote from: austinbob on June 29, 2015, 08:02:37 PM
I'm gonna give that a try just out of curiosity.
Did you ? I think you may be a rare breed, dunno if anyone else did, loadsa views tho' :)
Find any holes ? any faux pas ?
Wots the plural of "faux pas" ?
ok Zebedee >>>
Quote from: NeMo on June 29, 2015, 08:59:53 PM
I pretty much left the user to do the timing aspect..
Cheers, NeMo
Now I will have to dig out my stopwatch, or count the seconds - (one kangaroo, two kangaroo, three Kangaroo (or elephants)) etc.
Who remembers doing that?
Quote from: MalcolmAL on June 30, 2015, 11:02:34 PM
Did you ? I think you may be a rare breed, dunno if anyone else did, loadsa views tho' :)
Find any holes ? any faux pas ?
Wots the plural of "faux pas" ?
ok Zebedee >>>
Unfortunately Malcolm my grand daughter broke my laptop and my desktop is downstairs with test track upstairs so I'm stuffed for the moment.
:beers:
In one of his very rare 'silly comments' about model railways (well, it might work with O gauge or larger but in N or smaller ::)) Cyril J Freezer recommended in his book "1001 model railway questions and answers" (number 950/ page 210) that if you count the revolutions of the wheels and they exactly match the prototype then the speed will match.
Try doing that across the speed range in N :no:
Rather more helpfully in his excellent publication "Railway Modelling" Norman Simmons gives a table on page 250 which I quote only for N ( - other scales are available -) as the one which most interests us here.
In N gauge the following applies:
Feet per minute = SCALE MPH
6 = 10
12 = 20
18 = 30
24 = 40
30 = 50
36 = 60
42 = 70
48 = 80
54 = 90
and so on. From this table it can be seen that you can work out quite easily to the nearest 5 mph (eg. 15 mph would be 9 feet per minute) and I imagine that's quite a close enough estimation for most people.
Of course, very few of us are lucky enough to have layouts 54 feet long (if only :drool:) but we can always shorten the distance. For example, a scale 80 mph means a train moves forward 12 feet every 15 seconds or 6 feet every 7.5 seconds.
I have long worked to these times and feel they are quite realistic looking.
I hope this helps.
Greg.
Quote from: austinbob on July 01, 2015, 08:19:02 AM
Unfortunately Malcolm my grand daughter broke my laptop and my desktop is downstairs with test track upstairs so I'm stuffed for the moment.
Argh ! I feel your pain :(
We really do need a bod with an arduino and two sensors then !
I've just had a thought - I've still got some BeebBs, now they were proper computers :) dead easy to interface with real world sensors via the UserPort, hmmmmmmm,
hands-up, how many of you would find that interesting if I were to whizz up a prog for that, if I can still remember 6502hex :)
Quote from: colpatben on July 01, 2015, 07:38:00 AM
Now I will have to dig out my stopwatch, or count the seconds - (one kangaroo, two kangaroo, three Kangaroo (or elephants)) etc.
Who remembers doing that?
!! :) Not me, I was taught "one thousand and 1" "one thousand and 2"
T-shirt :)
(https://dl.dropboxusercontent.com/u/81786082/NG/50671Stop.jpg)
Really fun thread this one....just to reiterate I have used the trainspeed software to fix the scale speeds to my NCE Procab settings....it takes quite a bit of time to set the speed curve cv values but I have got my locos running within +/- 1 mph of the cab settings.... the hardest part is setting up the first curve for a new loco type say a Farish Class 47...but once one is done then another loco of the same type and make with the same motors etc takes a lot less effort as you just need to tweak the odd setting to reflect the slight differences between motors...I set the speed curves up and test in both forward and reverse directions as some locos definately run a little faster or slower for the same power setting when running in different directions....to be honest it is a good bit of fun getting them all speed matched and then when running double headed they all run nicely together irrespective of speed settings so getting the scale speed accurate is important to avoid straining the locos when double heading....all the settings are stored on my pc and backed up.
Quote from: sparky on July 01, 2015, 11:04:26 AM
Really fun thread this one
Well done that man, well spotted, that is in fact what I had in mind ;D :thankyousign:
Quote from: Only Me on July 01, 2015, 11:15:01 AM
I use the rule
I cant resist :-
I was wondering how long before someone brought their slide-rule to the party :)
computers, bah! who needs 'em
"roughly" and a decimal place, I luv it :)
Time for anuther cuppa >>
Hmmmm is this code freeware? I'm asking because I could translate this to python and run it in the Raspberry PI to create a real speedometer with two hall sensors and a digital display :hmmm:
Quote from: sparky on July 01, 2015, 11:04:26 AM
Really fun thread this one....just to reiterate I have used the trainspeed software to fix the scale speeds to my NCE Procab settings....it takes quite a bit of time to set the speed curve cv values but I have got my locos running within +/- 1 mph of the cab settings.... the hardest part is setting up the first curve for a new loco type say a Farish Class 47...but once one is done then another loco of the same type and make with the same motors etc takes a lot less effort as you just need to tweak the odd setting to reflect the slight differences between motors...I set the speed curves up and test in both forward and reverse directions as some locos definately run a little faster or slower for the same power setting when running in different directions....to be honest it is a good bit of fun getting them all speed matched and then when running double headed they all run nicely together irrespective of speed settings so getting the scale speed accurate is important to avoid straining the locos when double heading....all the settings are stored on my pc and backed up.
I hate to mention this but your models are probably better calibrated re speedos than many a real train in Britain where an allowance of up to 5 mph is (at least was) made in the cases of 'overspeeding' by drivers for speedometers giving false readings.
Quote from: MalcolmAL on July 01, 2015, 11:33:25 AM
Quote from: Only Me on July 01, 2015, 11:15:01 AM
I use the rule
I cant resist :-
I was wondering how long before someone brought their slide-rule to the party :)
computers, bah! who needs 'em
"roughly" and a decimal place, I luv it :)
Time for anuther cuppa >>
:thankyousign: I totally agree Malcolm. Next we'll be having someone work it out to the 0.0000001% of accuracy!
OK, I exaggerate but it does seem to be, in the immortal words of "Monty Python", "All getting a bit silly" now.
On the other hand, what is life without a bit of light relief? :bounce:
Quote from: deibid on July 01, 2015, 11:51:47 AM
Hmmmm is this code freeware? I'm asking because I could translate this to python and run it in the Raspberry PI to create a real speedometer with two hall sensors and a digital display :hmmm:
Why bother with a RasPi when you could put it onto a PIC ? Smaller, neater & allows the RasPi to be used for more important things :D
Quote from: MikeDunn on July 01, 2015, 01:40:33 PM
Quote from: deibid on July 01, 2015, 11:51:47 AM
Hmmmm is this code freeware? I'm asking because I could translate this to python and run it in the Raspberry PI to create a real speedometer with two hall sensors and a digital display :hmmm:
Why bother with a RasPi when you could put it onto a PIC ? Smaller, neater & allows the RasPi to be used for more important things :D
Clear answer: because the PI makes things (important and unimportant) way easier !
Quote from: deibid on July 01, 2015, 01:53:27 PM
Quote from: MikeDunn on July 01, 2015, 01:40:33 PM
Quote from: deibid on July 01, 2015, 11:51:47 AM
Hmmmm is this code freeware? I'm asking because I could translate this to python and run it in the Raspberry PI to create a real speedometer with two hall sensors and a digital display :hmmm:
Why bother with a RasPi when you could put it onto a PIC ? Smaller, neater & allows the RasPi to be used for more important things :D
Clear answer: because the PI makes things (important and unimportant) way easier !
and because it is meant to introduce kids to programing just like the good ol BeeB did ;)
On the other hand a PicAxe could be used and Basic ( the fact that it is Interpreted shouldnt worry our levels of timing accuracy ??) also, way easier to program than the naked Pic, but I still have a Pic assembler round here somewhere when I got tired of writing them in hex, gosh that was tedius :)
Deibid > sorry I didnt reply earlier, I've been
playing demo-ing N gauge to/with my daughter, (she has 00 ) we may have a new convert to the fold ;)
where was I ?
ah freeware,
as far as I am concerned yes,
some I have cut and choped from the web as one does,
some I have adapted from stuff on the web,
and some ( the messy bits of code obviously !! ) is my own.
So no worries, good luck.
Quote from: D1042 Western Princess on July 01, 2015, 12:22:55 PM
Quote from: MalcolmAL on July 01, 2015, 11:33:25 AM
Quote from: Only Me on July 01, 2015, 11:15:01 AM
I use the rule
I cant resist :-
I was wondering how long before someone brought their slide-rule to the party :)
computers, bah! who needs 'em
"roughly" and a decimal place, I luv it :)
Time for anuther cuppa >>
:thankyousign: I totally agree Malcolm. Next we'll be having someone work it out to the 0.0000001% of accuracy!
OK, I exaggerate but it does seem to be, in the immortal words of "Monty Python", "All getting a bit silly" now.
On the other hand, what is life without a bit of light relief? :bounce:
:laughabovepost: :thumbsup:
which is why I asked, somewhere in this tangled web we have wove, if I needed to round it to 1 deciplace, may have been on my original speedo page, no one answered tho' :( !
Quote from: MalcolmAL on July 01, 2015, 04:51:16 PM
On the other hand a PicAxe could be used
I agree - I've sometimes felt like taking a pickaxe to my computer too. ::)
:D
Interesting varying methods. I think I'll just measure round my oval in mm, convert that to miles, and time the locos round the loop with a stop watch with various speed settings on the controller.
Should be able to work out speeds in miles per hour.
There's quite a lot of variation in the locos I have. The Pannier tank is very low geared. I think 90 on the gaugemaster would be a similar real life speed for the tank. The Drummond M7 goes like a greyhound. 90 on the controller for that looks like over 200 in real life. That's my perception. I'll let you know the actual figures for all my locos.
Quote from: Firefox on November 24, 2015, 11:04:14 PM
Interesting varying methods. I think I'll just measure round my oval in mm, convert that to miles, and time the locos round the loop with a stop watch with various speed settings on the controller.
Should be able to work out speeds in miles per hour.
There's quite a lot of variation in the locos I have. The Pannier tank is very low geared. I think 90 on the gaugemaster would be a similar real life speed for the tank. The Drummond M7 goes like a greyhound. 90 on the controller for that looks like over 200 in real life. That's my perception. I'll let you know the actual figures for all my locos.
OK, work in mm, then convert if you like 'difficult', but why not just go for simple and use REAL measures? Feet and inches? Much more simple: 9 feet = 1/4 mile, 36 feet = 1 mile (OK, not exact, but near enough for 'realistic speeds') so if one circuit is (say) 18 feet and a train completes one circuit in (say) 30 seconds it will be travelling at 60 MPH.
Simple.
I'm a chartered engineer and we work in mm and m all the time. We don't use feet and inches so what comes natural to some is different for others. At the final step I convert to mph as that's a unit everyone is familiar with. I'd say use what works for you - at the end of the day we only want to know how far to turn the knob to get a realistic speed. There's loads of ways to get to the answer, it doesn't really matter how :D
Oh dear! not again! Imperial vs Metric.... :doh: :dunce:
LOL, is that a big debate here!?
I'm happy with whatever works for the individual, but I always found it strange that scales were often expressed in units like 2mm to the foot, 4mm to the foot etc. Seems quite a strange practice :hmmm:
Oh yeah old debate... I had a hard time trying to explain that 5% is not metric nor imperial... is a PROPORTION.
I don't want to open another can of worms though...
That's why I prefer a statement of scale like 1 : 150 or 1:75. Works for whatever unit you want to use :laugh3: