Mickoo's BR modelling

[mickoo]

You will want to double-check this as its been a long time since I was taught this stuff but IIRC DC motors all have the same kind of power/rpm/efficiency curves so the same basic rules apply to all of them. The specs always show the no-load rpm for their rated voltage, and sometimes the rated.

Max torque is at 0rpm, max power is at 50% no-load rpm, max efficiency varies, but somewhere 80-90% no-load rpm. This may be given as the rated rpm and is the highest you can run it under load, to be on the safe side I was told to use 75% of the no-load rpm if you dont know the rated.

A quick google shows the 1833 at 12V is no-load at ~9,400rpm and rated at ~8,500, so 8,500 is the absolute max speed it should run at under load. 8,500/93 =~ 90:1 for 15mph, or 8,500/124 =~ 70:1 for 20mph. That should crawl along pretty nicely

Jon

Jon,

Likewise it's a long time since I did detail motor theory, these days if it goes in the direction its supposed to, doesn't smoke, spark or throw its com all over its insides and plunge half the town into darkness I'm a happy bunny But then were dealing with much much smaller stuff here and I'd based my value on 9,000rpm as a rough yard stick, I can of course check the rpm as we have a rpm meter at work, just needs a flywheel with the mark required for the meter and the speed can be determined quite accurately.....if I were so inclined LOL.

I'm still yet to be convinced that a motor at 9,000rpm on a 70:1 ratio is better than a 9,000rpm motor at half speed on a 45:1, unless, big caveat here, it's a multi stage gearbox, mind I doubt you can get a 70:1 ratio as a single stage as the final gear would be quite large I fear.

 

[mickoo]

Hi Mick

I would argue that the ex Ron Chaplin gearboxes now sold by MSC are better, for the simple reason that the gearbox frame is machined from solid rather than fabricated (same sort of gear train etc). I would also argue that very few of the ABC gearboxes are square.

It doesn't stop me using them, as they still seem to run very smoothly and are DCC friendly.

I use Portescap (RG7) motors in some shunting locos, and the controllability is superb. They whine at higher speeds, and I would not put them in MT locos or diesels.

Regards

Richard

Richard,

Thanks for the input, most interesting. I do concur that squareness is essential as is rigidity, I also feel that ball races on intermediate and final shafts are important in reducing drag and thus gearbox noise.

A few of us at WM this last meet were discussing gearboxes and their construction and basically came to the same conclusions, rigidity, accurate shaft alignment, ball races on all shafts and good choice of gears made a good quiet gearbox.

Is the Portescap a can or coreless, I'm guessing a can and it'd be quite large sort of like my Buhlers so big magnets in there to give good control and power. Is it the motor that's whining or the gearbox because it's travelling so fast?.

 

[ZiderHead]

I'm still yet to be convinced that a motor at 9,000rpm on a 70:1 ratio is better than a 9,000rpm motor at half speed on a 45:1, unless, big caveat here, it's a multi stage gearbox, mind I doubt you can get a 70:1 ratio as a single stage as the final gear would be quite large I fear.

For a shunter I'd say you're probably right, 50% gives you max power at top speed and the reduction is so big you still get smooth low speed running anyway (and less noise if the primary stage is a worm). For a mainline loco 75% would be good so you can run express passenger trains at full speed, and freight at 50% where you need all the power.

In that case, 15mph = 9,400/2/135 = 35:1, 20mph = 9,400/2/180 = 26:1.


* edited for bad maths

 

[mickoo]

I'd like to experiment with both to be honest with the same motor and see which is quietest and easiest to run, I also need to explore to see if DCC can map a power curve to only give 50% power at 100% throttle input.

The problem now is finding a supplier of decent primary gears, helical, bevel or at worst contrate to try and eliminate some of the noise, there are some suppliers (Daval, HPC)who do very nice helicals but none appear to fit a 2mm motor shaft Not so keen on bevel or contrate as when they turn the drive 90° they expand the gearbox width more than helicals, not that should be too much of an issue in O gauge.

Slaters used to do a very low ratio final drive (FD01) with what looked like helicals so there must be a source somewhere out there LOL. FD03 looks nice as well but probably not so good as a primary gear into a drive train. Their GBG3 also looks to have one of those 3:1 helicals as a primary gear, which is exactly what I'm looking for as it'll allow roll on like the ABCs do.

And what happened to Branchlines, who apparently do 2mm bevel and helical gears, do they still trade, getting info on their stock is nigh on impossible.

 

[Steph Dale]

Is the Portescap a can or coreless, I'm guessing a can and it'd be quite large sort of like my Buhlers so big magnets in there to give good control and power. Is it the motor that's whining or the gearbox because it's travelling so fast?.

Mick,

It's coreless, as all Escap motors are. The whine comes from the straight-cut gears in the 'box.
Advantages: efficient so low current consumption, great torque due to the coreless motor, ready-built unit sop drop in and go. Good value.
Disadvantages: at speed a bit of noise in some. At 20-30mph though they're all but silent. Cost

Mashima 1833 no load at 12v is around 8500rpm.

You can map speed curves to give you what you want (someone I know has a Mercedes Unimog that correctly holds speed at gear changes as it accelerates). Even a three-point curve will allow you to limit top speed.

Steph

 

[mickoo]

You can map speed curves to give you what you want (someone I know has a Mercedes Unimog that correctly holds speed at gear changes as it accelerates). Even a three-point curve will allow you to limit top speed.

Steph

Guessed you might be able to, but your confirmation is greatly appreciated.

Thanks for the info on Portescap, would angled cut reduce the gear nose then?

 

[Steph Dale]

Guessed you might be able to, but your confirmation is greatly appreciated.

Thanks for the info on Portescap, would angled cut reduce the gear nose then?

Gear nose? Nose? Aaah - noise...

Hmm, it depends; but generally yes - helical gears are quieter and more efficient.

Steph

 

[mickoo]

Gear nose? Nose? Aaah - noise...

Hmm, it depends; but generally yes - helical gears are quieter and more efficient.

Steph

That's my cue to get off this research train LOL, the SD gearbox has been run both ways for quite some time and is quietish, there's a little click now and again so might strip it down and look for something in there just catching or not quite meshing, I'd read somewhere that a light covering of Brasso to help smooth the gears in is often benificial during early stages?

 

[SteveO]

That's odd as the Sentinel has tiny little wheels so I'd guess 40:1 should be ok?

Sounds also like you were running DC and not DCC?, I think (guessing here) you can adjust the speed curve in DCC so that 100% input only gives 50% output I.E digitally cap the top speed, if that is so then you can adjust the whole speed curve to make the first 50% of the controller control the first 20% of the speed and the other 80% in the last 50% of the controller, where such close speed control is not generally required.

We do the same at work with the crane joystick, the first 50% controls the lower 10% of the speed range where you need fine control for landing boxes on trailers 120 feet below you, the rest of the 50 controls the other 90% of the speed range, well it doesn't actually, it controls the next 40% up to half speed, what we call base speed. the last and highest speeds 50%+ are controlled by the load dependent system which measures the load and field weakens accordingly.

So with an empty crane the driver controls the lower 50% and the LD controls the top end and gives a total of 100% speed, with 40t on the driver controls the lower 50% and the LD calculates the load and applies no field weakening and thus the crane runs at 50% with the joystick hard over at 100%. Takes a while to get your head around but does allow for very fine control and safe heavy load high speed operation /end] OT speed controllers LOL

That's pretty interesting info on the crane - I had no idea they were so technical.

The Sentinel is DC at the mo but the plan is to throw in DCC and sound at some point. I don't have a layout, or even any permanently wired track at home so this is the first time I've run it since I bought it and it was surprising how fast it went. I used the club's controller which is an old H&M normally used on our oval test track but this time it was hooked up to our S&D layout-in-progress (by that I mean a couple baseboards and some laid track still being wired) and it was 'lively'! I'm not sure of the voltage or amperage but it is a fairly standard affair which powers everything except our DCC members.

I'm not too worried about it at the moment but I'll have another look when I get some time to concentrate on it.

 

[mickoo]

That's pretty interesting info on the crane - I had no idea they were so technical.

LOL, thats the Peter and Jane description, deep down they are very complex, especially the AC motor control versions which vary the frequency and modulation to control speed and even skip certain frequencies that are known to cause mechanical damage to the motor or drive (harmionics). On top of that we return power to the mains when lowering off as we then use the motor as an electric brake, two cranes lowering 60t is enough to let another hoist with 60t.

Back to the Sentinel, if it's DC and from a H&M then max voltage will be 12v and if your using a 40:1 gearbox then you really don't need more than 6v to run at max speed, I.E. half controller power.

As an aside, some one just said to me "why are you worrying about gearbox noise when your going to fit a DCC sound chip?" I have to conceed they may have a point.

 

[ZiderHead]

I was thinking that - for diesels the gear whine is just another layer to the sound. It could be said to represent the turbochargers, blowers or traction motors

 

[adrian]

Hi Mick
I would argue that the ex Ron Chaplin gearboxes now sold by MSC are better, for the simple reason that the gearbox frame is machined from solid rather than fabricated (same sort of gear train etc). I would also argue that very few of the ABC gearboxes are square.

The Wyvern gearbox from Bob Moore is also a fully machined gearbox and very well engineered. I remember Bob spending many an hour at the Manchester show talking with Sid Stubbs when he was designing it, so it's safe to say where the inspiration came from.

 

[alcazar]

I saw on at Telford an it looked the part.

Question, though: The ex-Ron Chaplin ones seem to fit the bill, why do we search for others?

 

[Dikitriki]

I saw on at Telford an it looked the part.

Question, though: The ex-Ron Chaplin ones seem to fit the bill, why do we search for others?

The ABC range has many more options for different spaces, where the ex-Chaplin one won't fit.

 

[mickoo]

I saw on at Telford an it looked the part.

Question, though: The ex-Ron Chaplin ones seem to fit the bill, why do we search for others?

Two reasons,

They're even more expensive than ABC

There's limited ratios to choose from.

 

[mickoo]

Not much to report on the 08, some new tips for the iron arrived so hopefully I'll have more success with the smaller fiddly bits, rather than struggling with the massive current tip!

Thoughts constantly turn to other projects in moments of idleness and one has been working up a decent of dimensions and notes for a class 66 model. Most of the stuff can be measured from the ground but the roof and cab fronts present some technical difficulties for the human race, it's a bloody long way up there.

Fortunately a word with the shunter who'd left a 66 parked up to some spine liners on a siding solved a few issues, spine liners have a small platform above the buffer beam, allowing access to measure cab windows etc, so a quick hop up and a couple of measurements in the bag was all that was needed. Granted many do not have this access so please don't try this at home folks

Having gained the minimum two measurements, one vertical and one horizontal, I took four to cross check everything, it is time to proceed to step two, the procurement of high detail close up photos.

To minimise the risk of parallax errors caused by shorter lenses it is recommend you use 150mm or above, the bigger the better really, I used 300 and grabbed a couple of cab front close ups.

Step three can be accomplished several ways, print out your pictures and use a rule and then work out your scaling factor, or if working digitally do the same but count pixels to get your scale, both require a modicum of hard labour so I decided to try something new.

Adobe photoshop has the ability to lay a grid over the image and the grid is scalable, so with some jiggery pokery we end up with the below images, each major grid is 1"each sub grid ½", from this it is very easy to work out major dimensions and then transpose them to the cad art work for etching or even for just making up a decent drawing.

I do need to get a better head on shot with out barrier clutter in the foreground but the principle is sound and I've some side profiles (but need better) I can do the same to cross check my ground measurements.

As a word of warning, class 66 cabs are tapered so the roof profile is not quite the same aft of the cabs.


Knowing the full loco width and thus the cab width, it is possible to workout the roof profile on the body from the front shot, as well as the radius for the brow and windscreen corners.







The last is quite good as it shows nose to nose cabs 1 (left) and 2 (right), which are different lengths, but same taper, for those interested the cab taper starts on the centre line of the jacking pad and can more easily be seen by the change in roof profile.

It doesn't have to be a class 66 (I was just lucky to have one or two kicking around), any loco will do but the principle is the same, one or two accurate dimensions and a couple of high definition photos from as far away as possible with as big a lens as you can get should get you well on your way to accurate dimensions for model making. It also helps if your chosen loco still exists LOL.

Addendum, after more research jacking pads are not in the same location at each end, see below.

 

[mickoo]

Managed to get some, in poor light side profiles to apply the 1" grid too, quite impressed with the accuracy, certainly less than ¼" which is more than enough for large parts like cabs and engine covers, exhaust silencers. More detailed parts can be taken in close up like the horn cover above.

Class 66 side profile, far from simple to model accurately due to the tilted prow which will throw off the height of that face plate if viewed head on, windscreen too, both will need a small height compensation when doing the CAD artwork which is flat. The detailed side profile will allow the angles to be measured accurately and thus the exact height calculated.

Warning large file, grid is 1"sq


The shot shows quite clearly the change in roof profile above the cab, due to the 2½" taper which as noted above starts mid jack point, you can see the weld join just aft of the TOPS panel. What I didn't know was that the jacking points are not in exactly the same place, #2 end is further back yet the cab taper is still the same length at 65". I'm guessing it's to do with the weight disposition as #1 end might weigh a little more due to the cooler group, or it could be that #2 end is correct and #1 end pushed forward to clear the cab door? Either way they are in different locations.

One thing I've begun to notice, these are far from simple boxes on wheels, the utilitarian construction leaves welds and seam plates all over the place and the stamped body sides as well as removable sections with their associated gaps and weather seals makes for quite a lumpy model. I also like how some of the older locos are now beginning to rust and paint bubble making for some nice detail effects.

Below an exercise in how far can you go with detail and a grid, again a 1" grid overlaid on a whole bogie, to retain definition the image is much much larger so click large size at your own risk LOL, it's small...ish bandwidth size but physically very large, but posted here to show how much accuracy and detail you can get, certainly enough to work a 3D print from. I do need to shuffle the grid a little to align it with the axle centres and it may not be scaled just so at 163" wheelbase but again, no more than ¼" out by my quick calculations. Im also not dead centre as can be seen by the parallax error creeping in on the rear axle, about 8" to the right would have been perfect.


Note also the clean rubbers around the axle bearings, a sure sign of new wheel sets having been fitted, confirmed but the 'fat' tyres on the wheel and shiny steel effect. The rim can be measured at +1¾" off the witness mark, unlike 556 below which will be due for new wheel sets very shortly as there is virtually no tyre left above the witness marks.


What's next? Well I have a 4mm model and a Gauge 1 one in bits in the loft which reminds me I must put it all back together LOL, it's in bits because I wanted to super detail it but the scale of 1:29 puts me off as I know all that work will not be right for 1:32 (no that hobby horse will not go away) and.....there's too much out of scale even in 1:29 to satisfy my uber detail wishes for a class 66.

So a nice accurate S7 model (no 59/2 hybrid work around) would be nice, I can see my Euro bubble (they do paint theirs such pretty colours) looming large on its orbital path, so with care an etch that would suit both sides of the channel could be accommodated.

Then there is the very large model itch that keeps nagging away at me, 5" gauge would give 70mm dia wheels which I can turn in the lathe and that class 40 I saw at Warley keeps nagging my conscience to follow suit with something equally as big, which of course means breaking out the boat building tools and materials for the sub frames.

And then theres' all the other projects to tend too...sigh.

Of course I do understand I'm in a class of 1 when it comes to shed foaming so will excuse y'all asleep at the back of the class by now LOL.

 

[ZiderHead]

To minimise the risk of parallax errors caused by shorter lenses it is recommend you use 150mm or above, the bigger the better really, I used 300 and grabbed a couple of cab front close ups.

This is well worth repeating, almost every shot I see modellers of all flavours take for reference are taken from wide-standard. I understand its impossible if you're standing in the 6' with your back to another loco or in a packed museum but if at all possible stand waaay back from the subject, and if you dont have a long tele just crop the image. Portrait photographers shoot at ~100mm because its far enough away to avoid giving people big noses, but close enough so they dont have to shout at their subject

 

[alcazar]

Interesting.

I laughed at the thought of people trying to get measurements of loco windows.

Imagine the driver's face when some guy appears on his front buffers with measuring tape in hand........

 

[Pugsley]

Of course I do understand I'm in a class of 1 when it comes to shed foaming so will excuse y'all asleep at the back of the class by now LOL.

Not at all, carry on, I'm finding this very interesting and I'm sure others are too.