Axle Hung Motor Gearbox (AHMG)

[IanT]

I'm in the process of gathering various bits of useful information before I start-in doing some CAD 'doodles' for an axle hung motor-gearbox (AHMG) design, initially using the high quality 7-pole motor (HQ7P?) of which I'm very pleased I recently purchased a small quantity.

My target 'prototype' locomotive has 4ft diameter wheels (twin bogies & four powered axles). However, as I'd like to try and make my design useful to others, I was wondering if anyone had any views/ideas on the minimum wheel diameter I should target, as this will impact the gear PCDs used (e.g. gear sizes/ratios).

Perhaps I should clarify "minimum".

I'm not after the smallest wheel-size ever used in full sized bogie/locomotive practice. I'm asking what would be a 'useful' minimum full-sized wheel diameter to work to? I'm sure your reasons (e.g. choice of prototype) would also be of interest to others. So please don't just state a wheel-size, give the reasons for your choice too.

Other data - such as distance between (powered) axle centres, inside or outside frames etc. - would also be useful

Regards,

IanT

[cabbage]

We have been through this before...

The std BR wheel was either 3 feet 9 inches or 3 feet 7 inches. According to BR regs all wheels on a bogie had to be within 1 inch dia of each other. I proposed a wheel size of 3 feet 8 inches as this would satisfy the THIRTY SEVEN classes of BR loco that could use this wheel size.

So, the std wheel can be turned from a 50mm slice of steel bar and you would need a gearing of around 6:1 or 8:1. If you use a 0.7MOD would give a 64 tooth drive gear and a 12 tooth spur gear which would be exactly a 6:1 ratio...

This could simply be used to provide a power module by adapting the same axle slung system from the Chris Barron "Peak" design.

regards

ralph

[Doddy]

I would argue for a 3' 9" wheel with enough extra tyre material on it, deep enough for the fastidious to turn down to 3' 7" where needed. Alternatively provide both! 

When discussing this offline, I tabled the following information.

3' 0" (36" / 40.5mm) Disc & Boxpok Pony Wheels
40, 44, 45, 46 & 10201-3

3' 1" (37" / 41.63mm) Disc
1st Generation DMU's & 18000 Gas Turbine Idler (3' 2")

3' 3-1/2" (39 1/2" / 44.44mm) Disc
15, 17, 31 (middle idler), 42, 43, 201-296 DEMU Driving Bogie (40"), HST Power Cars

3' 6" (42" / 47.25mm Disc
10000-1, Blue Pullman, 201-6 DEMU's Trailer Bogie

3' 7" (43" / 48.375 Disc, Spoked* & Boxpok**
20, 22*, 23 ,26, 27, 31, 33, 37, 50, 52, 53, 55, 60, DP2, EM2, D600-4, 10201-3**

3' 9" (45" / 50.625mm) Disc
24, 25, 35, 40, 44, 45, 46, 47, 56, 57, 86, 87 (3 9 1/4"), HS4000, D0260 Lion

4' 0" (48" / 54mm) Disc and Spoked*
08*, 14*, 71*, 74*, 81, 82, 83, 84*, 85, 18000 Driving Wheel*

4' 2" (50" / 55.77mm) Spoked
EM1 (Class 76)

Question is who is currently building what locomotives?
What locomotive is likely to be built in the future?
What will sell?

[cabbage]

The next British locomotives that I intend to build is the NBL Gas Turbine and the SR 10203. (Yes I am that lazy!) The next sequence will be the NS 1500 "Pandora" in NS livery. We do need a source of std wheels be they 3'9" or 3'7" and a std gear and axle design to make our power bogies from... Wether the Society can provide these as laser cut parts or a supplier takes up the challenge. It is something that we need.

regards

ralph

[Geoff Nicholls]

I'm building a Derby Lightweight DMU, so I I need 3' 0". This appears to be standard for most first generation DMUs
Second generation DMUs have smaller wheels, say 800mm or 35.5 in G3 including the class 150/2 I'd like to build.

But my priority is 3'

[keith Bristol]

Hi

For my Class 58's 3'7'' and 3'9'' will be fime

Cheers

[keith Bristol]

I should add that that the DMU wheels of Geoffs will be very useful to me..

[Gavin_B]

Am planning a metropolitan camel back, working on using brushless motors with no gearbox.  Have got the motors turning very slowly with plenty of touque,  just need to sneak a lathe past the boss to turn down an axle to fit the motor to.

[Doddy]

working on using brushless motors with no gearbox

What controller are you using?

[Gavin_B]

What controller are you using?

Am using an arduino uno to output pwm signals in a sine wave to an l6234 triple half bridge amplifier from here
http://www.drotek.fr/shop/en/home/212-brushless-gimbal-controller-l6234.html

Sounds far more complex than it really is, the l6234 board was cheaper last week as well

[Peaky 556]

I am building a Peak (Class 44) with 3'9" drivers on 8' axle spacing and outside frames.
Tim

[IanT]

OK - I'll work to a minimum wheel diameter of 3ft 7" as it seems to suit most needs (and a bit more if possible) - all depends on what the track clearances etc. look like.

Just to be clear - I'll be doing this at my own pace (which is typically slow) and mainly for my own use. When (if) I get to the stage where I'm going to order any 'parts' - then if anyone wants to join me, they will be very welcome to do so. I have no plans to sell bogie kits (or anything similar) on an on-going basis. It would be like the HQ7P motors - a group purchase approach.

Gavin - a very interesting approach to a drive unit - what brushless motors are you using?

I've been playing with an Ardunio Uno for a little while myself - it's a very interesting piece of technology. However, I've been looking at possible alternatives to the Arduino/Atmel solution, mainly because if I invest my (limited) time in learning a particular architecture, then I want to make sure it's got some 'headroom' and I won't be forced to change to something better/faster/cheaper etc for some reason later..

So, I've just decided to standardise on Microchip 32-bit (PIC32) based devices and for 'on-board' use, set them up with an Arduino-like (Chipkit) environment using the MPIDE. I'm currently waiting for my PICkit 3 to arrive, so I can load a custom serial boot loader. If you are interested, Microchip already sell a PIC32MX250F128B with a USB/Chipkit boot loader installed. This is a 28 pin SPDIP device - so it's easy to breadboard. They sell them for just under £3.00 each (plus P&P) and they will deliver a lot more power, memory and peripheral choice than the Atmel CPUs on current Arduino's but are pretty much s/w compatible. You only need to add a few external components to have a very powerful (and affordable) controller, programmable via a laptop with a USB connection (just like the Uno). The Microchip part No. is TCHIP-USB-MX250F128B. Just a suggestion.  :-)

I can sense Mike W nodding off - so I'd better leave it there.

Regards,

IanT

[Geoff Nicholls]

for my DMU, I was assuming I'd use the GRS 3 hole steel 3' wagon wheels with a 1/4" axle. The axle hung motor/gearbox, would then be compatible with Slaters and Mark Wood wheelsets.  the use of the steel wheel would mean the spur gear could be bolted on to it, and the bolts would be concealed by a cosmetic disc on the outside of the wheel.
I would power both axles of one bogie per car, thus four axles in all, the motors connected in series to a Revolution receiver and a 14.2v battery.
Of course the DMU would only need to pull itself. Would the arduino be used to vary the power to cope with heavy trains etc? 

[Gavin_B]

Ian,

Will take a look at the chipset you suggested, I do have a nagging doubt that sooner or later will get to the limit of the arduinos, saying that not got close so far.

I am using the D2208 Gimbal motors, they are designed for slow speed use

[IanT]

Hi Geoff,

Given a sufficient hardware base, an 'Arduino' controller can do just about anything you want it to do but of course you have to write the code. One of the attractions of the Arduino approach is that there are libraries of pre-written code to drive various 'devices' (DC Motors, Servos, Steppers, Sensors etc). So these can be used to produce custom solutions. You can also create your own libraries - so it would be possible to write routines to meet our specific needs in any particular area and then share them.

The key limits to any embedded controller are the processor speed, available memory and (effectively) the peripheral 'pin-out' (e.g. how much access it has to the outside world). In terms of more sophisticated DC (Brushed) motor control the two main feedback mechanisms are probably an 'encoder' (something that measures rotational speed) and a current sensor (to monitor the current being drawn). So without going into too much detail, you probably need to drive your 'power' unit with a PWM source and measure the speed/current using some form of ADC (analogue digital converter). All this is possible with an 'Ardunio' type architecture.

The few 'commercial' solutions in this area that I've looked at appear to be based on older 'PIC' technology but more to the point are generally 'Black Boxes' as far as the user is concerned. As I've already mentioned, a very powerful 32 bit micro controller in a 28 pin DIP is available for under £3 and even on perf. board it would have a very small form factor. The 'pin-out' on this device is programmable - so you can chose what kind of I/O you implement (it has a very extensive peripheral set available - including multiple PWM and ADC channels) so we could have a simple & affordable common 'hardware' platform which offers enormous flexibility in terms of custom solutions via the easily (USB + PC) installed software. Some might view this as overkill but in my view it's far better to have too much capability, rather than potentially too little - and at >£3 per unit why do I care?

If others want to discuss 'Ardunio' further - perhaps we should start some new threads in this area? I'll open a 'bookmark' topic elsewhere.

Back to physical things....

With respect to 3ft wheel diameters, one of the obvious difficulties when trying to design any A/H gearbox is to find a sufficiently powerful motor that will fit comfortably within the wheel back-to-back dimension. The reason I've been trying to establish (a range) of wheel sizes is that the HQ7P (+ Spur Gears) will not fit inside the 58mm G3 wheel B2B dimension. So it needs a somewhat different approach and I probably have one but I'm already pretty sure it will only suit bogies with outside frames - which hopefully will not be a problem in most cases. How 'flexible' the design can be  - remains to be seen.

So I'm still thinking about the general approach required but I will then check it with CAD (currently getting into 3D and it's not easy). But that's all my news for now - back to my current project (I'm trying very hard to stay focused on it at the moment - which I'm not great at).



Regards,

IanT