Scratch Building A Model Railway Vertical Fiddle Yard! | ml1.

[@richardnew1215]

This would make for a fantastic how-to-do-article in the Model Railroad Hobbyist emag. Well done!

[@modeloco]

Thank You Richard. Must get an update up soon 👍

[@stephenpike3147]

Hi Andy, now that is very different, good on you. One possible way to give accurate positioning is to use a digital encoder on your lift motor the pulse output (if use quadrature pulses A and B you can get direction and thus never loose accumulated count) fed into a high speed counter, count the pulses and have trip points for each cassette position and one for each direction of approach to the exit track. As you approach from maximum speed you slow down to a crawl and then stop. All you need is a top and bottom overtravel limits and that is it. Used this technique with plcs many years ago in a number of industrial applications with superb accuracy. You can then easily change the trip point values in software if the wood say expands with humidity changes - a problem I have here in Canada. Best regards Stephen ps in our application I would suggest need a datum to reset to on power loss or otherwise.

[@modeloco]

Thanks very much Stephen for this superb information. Certainly interested in this way of stopping the lift motor if, you could pass on any links to these sort of items, you mention, online. I’m going to be using block detection and software to stop trains accurately and reed switches to tell the software when to power up the tracks, only when in line. This has just all been wired in so, it’s all being logged in next update video. I also came up with a design to hold the lift in a more level, sturdy state as it moves up and down, without any visible runners, etc against the wall. The biggest challenge was an accurate datum edge or two to work off and, suitable materials as, it’s all got to be spot on so, as not to bind or, become tight in travelling the 800mm. Fingers crossed, it seems a success.

[@stephenpike3147]

@@modeloco Hi, so pleased that was well received, you are most welcome. The industrial kit costs £1000s (Square D series 500 plc and SSD Link - all discontinued) but the principles can be applied. You can logic gate the phase shifted A and B pulse trains to produce separate direction and pulse train signals (at 4x the A or B frequency), after that you need something to count the pulses. For encoders try Litton Encoders, it sound like you have a high reduction gearbox motor to load so perhaps an old CD could be used as the basis for a crude but very coarse encoder that would still give good positional resolution - I am thinking out of the box here. I do not know if groups around raspberry pi, MERG etc do anything suitable to count and run suitable software for and then give the digital/ analogue outputs to control your motor… For holding position perhaps you could try making a simple disc brake attached to your motor (using your gearbox to full mechanical advantage) and sequenced from your decoder count and start/stop from/to positions. So many ways of doing things, if you do go down this path drawing up a timing and sequence diagram is a big help. Hope gives you something to think about! Best regards Stephen

[@Dr.Gunsmith]

What did you use as a lifting hydraulic system.

[@modeloco]

Now that would spoil the upcoming instalments, once I catch up with all the TT stuff and the winter has finally receded. 🤞

[@Dr.Gunsmith]

@@modeloco the suspense 😂 I’ll be looking forward to that.

[@alanmusicman3385]

I have been planning and experimenting - just very simply at small tabletop scale - with a train lift for a while now and am nearly ready to build something upscale now. Mine will use T-Slot Aluminium parts though - such as are used in 3D printers for moving the job platform and the print head around - that stuff is also used in robotics for some tasks. From my experiments, I think you are unlikely to get away with any simple scheme for the positioner. For the repeatable positioning precision that is required to align each shelf with the surrounding track, I think reed switches or hall effect switches alone will be unlikely to be satisfactory. They may work initially with very careful setup but there are a lot of variable factors - for example the exact point at which they trigger will vary a little as the setup ages, then the exact decelleration time of the elevator motor could vary quite a lot depending on how many trains are on the shelves and wear on the drive mech (whatever that is) will also be a point of long term variance. In short I think you will need some programmed in intelligence to get the precision you need for long term reliable operation. That would involve using a microcontroller unit (MCU) such as an arduino or an ESP8266 with an attached motor driver circuit which allows moving the lift, but also fine positioning it. You could then use magnets on the shelves to trigger hall effect switches on the entry/exit tracks for the coarse ("Almost there") position detection, but for the fine positioning I think you'd need to use optical sensors. The opticals could be an infra red LED and sensor pair (a single device) shining from the underside of the entry/exit track shelves. On the underside of each shelf you then put a small block with a narrow (2mm maybe) horizontal slot cut into it with a mirror behind it. When the coarse ("almost there") magnetic sensor is triggered, the microcontroller sets the motor to a slow crawl and hunts for the point where the target shelf's optical sensor returns the most reflected light - which will be the point at which the shelf tracks are level with the exit/entry track. Of course the MCU would need to have additional sensors to let it know when the lift has reached the absolute top or bottom of the lift ("limit sensors") which could just be microswitches, but it would also need to count how many shelves it passes (mag sensor detections) on its way to the target shelf so it knows when to enter fine position hunting mode. When I get my upscale prototype done I'd be happy to share details with you of the parts, circuitry and software. Best of luck. Alan

[@modeloco]

Thank you for the very much appreciated detailed response. I'm asking a lot yes, especially with twin tracks on each shelf 1-10 but, giving it a go. Developed machinery many moons ago as a trainee development Engineer using pneumatics with bearings mounted on chrome 25mm diameter runners. They had to be shimmed very accurately to enable it all to travel freely, end to end. The LocoLift is a lot bigger, in a vertical plane, not horizontal, has a much, much heavier moving platform bed and isn't made of steel but, is timber, thick enough to drill and not cause flex, so all in all I guess that would be a big undertaking indeed on this scale to use the robotic aids or car motorhome parts and, hoping to get it all nipped up to run absolutely square and smoothly, using my wall as a kind of rough datum surface, may have been a bit too extravagant for a hobby idea in my head. Already put a fence and shed up for this. That said, I'm relying on the dead weight of this unit, coupled with the natural positioning as it rises and falls to achieve my goal. Have developed and produced track adjustment for each double track on each end and so far am happy at the early test results as its not a production idea, just me using it... but, it is still early days. The 5" diameter bore hole was a gamble that paid off, as if it had been too high or low it wouldn't have enabled the lift to work in its full potential. The other factor was drilling it square. It ran off a bit but, probably due to fence and exterior wall being neither square with each other or the interior plastered wall. We drilled from inside out to save plaster bursting off, which would have happened, had we gone from outside to inner wall. Got it about bang on in the end. This bore then set the benchmark for everything else to be set to. Lots of measuring before drilling. Anyway... Spring is here... Back on it gain now the Winter music producing months have had a good run. See music seems to feature in your life too. Would love to hear and see your evolving efforts in due course. Kind Regards, Andy.

[@scottsound4711]

TopMan

[@modeloco]

If you can't go out, the only way is up!