Freerails Home 
Home Search search Menu menu Not logged in - Login | Register

Freerails IS ACCEPTING new Members ... To join Freerails ... See how to Register as a Member in the 'Joining Freerails' Forum

Mynis Cule Magna - a BPRC adventure
 Moderated by: W C Greene Page:  First Page Previous Page  1  2   
New Topic Reply Printer Friendly
 Rating:  Rating
AuthorPost
 Posted: Wed Feb 18th, 2015 08:25 pm
  PMQuoteReply
11th Post
DavidT
Registered
 

Joined: Wed May 2nd, 2012
Location: United Kingdom
Posts: 474
Status: 
Offline
Valid point Herb. Using new stuff makes the challenges clearer and hopefully leads to improvements over time.
Regards, David.

Back To Top

 Posted: Wed Feb 18th, 2015 08:56 pm
  PMQuoteReply
12th Post
DavidT
Registered
 

Joined: Wed May 2nd, 2012
Location: United Kingdom
Posts: 474
Status: 
Offline
Geoff, I was thinking aloud too perhaps :-) I think you need a switch per action to make it easy to label/as intuitive as possible. I would want to lay them out with graphic and positions that match the layout.

In theory one Tx can handle up to 70 2-position switches each with completely independant action. That would need a microprocessor based circuit to measure all the switches, scrunch their positions into 7 channels of data, and squirt it out on a 'Tx3' type product (with the catchy name 'Tx1-K1'). It would need new 'Rx105' type receivers that can decode the scrunched data. With one Tx, fresh data every 22ms should be possible.

For turntables my background leads me to hobby servos to rotate them but that would need a new 'Rx105' type controller to handle the start and stops. Probably more feasible with 2-position turntables (do some have more positions?).

The challenge with all of this is that the 'requirement' is for any and every purmutation of anything and everything. It might need a bit more discussion to be able to do that without our heads hurting :-)
Regards, David.

Back To Top

 Posted: Fri Feb 20th, 2015 08:07 pm
  PMQuoteReply
13th Post
Geoff L
Registered
 

Joined: Tue Nov 4th, 2014
Location: Hemel Hempstead, United Kingdom
Posts: 44
Status: 
Offline
David

Thanks for that - it certainly looks promising.

Turntables can have any number of positions, limited only by the number of tracks that can be laid around the circumference and each could be accessed from either direction (forward or reverse). The use could be for switching tracks instead of using turnouts or the conventional means of reversing tender locos to operate in a forward direction.
The MERG driver I mentioned can service 63 tracks with a head/tail switch to reverse direction and the turntable takes the shortest route to achieve the objective. This uses a micro-controller. It may be that a servo could be used but having not found any details, I was uncertain whether more than two positions could be programmed. It was mentioned previously that satisfactory operation is dependent on accurate indexing and if that can be achieved, a conventionally driven motor would be feasible.

Turntables are a bit specialised and on a conventional panel could use bespoke applications. Similarly, with a conventional geographic panel the number of controls is immaterial and in many applications space is not a problem so the benefit may be in the cost of fewer transmitters (other than the saving on wiring).

Commercial turntables with controllers are available but are very expensive.

Having enthusiastically embraced the concept, I was (am) trying to minimise the layout control to a single hand-held controller to demonstrate advantages over a completely DCC electrically driven system for a similar sized layout.

Above a certain size and complexity and with multiple operators, one or more panels would be needed, preferably with a graphic representation of the settings and track occupancy.

It is possible that voicing my personal requirements might deflect you from the bigger picture, but if they can be served then the application should be easily transferable. I thought that trying to detail a practical example might help you by showing the approach from a model railway perspective from someone ignorant of the workings and possibilities of radio control - probably like many of the potential customers!

Going back to the Selecta principle, this uses a single throttle control, inertia setting, lights switch, etc to drive each of a number of locos chosen one at a time. I initially had hoped to achieve something similar with a set of various switches driving the turnouts, signals, magnets, etc, each category selected one at a time. The logic being that on a simple layout the settings would remain the same whichever loco was selected, or had been changed before the new loco was selected.

The new Tx looks as though it might go some way to doing this.

Regards

Geoff

Back To Top

 Posted: Sat Feb 21st, 2015 08:02 am
  PMQuoteReply
14th Post
DavidT
Registered
 

Joined: Wed May 2nd, 2012
Location: United Kingdom
Posts: 474
Status: 
Offline
Geoff, I find your worked example most helpful and I like to try to accomodate as many perspectivies as possible. I see the appeal of a hand-held unit. My range evolves from feedback and experimentation so please keep asking.

Some advantages of servos for turntables, points, semaphores etc are that they can be very cheap, available in many sizes, easy to mount, and have an integrated controller. The latter is helpful because people can use converted receivers, Arduino, etc. The number of positions is determined by the microprocessor. The resolution by all the electronics, gears and mechanics involved.

Hobby King and others sell 'robotics' servos which allow greater than normal or continuous rotation. A sail winch for a model yacht will be similar. And in the old days we used to replace the feedback pot of a standard with two fixed value resistors and remove the internal stops to allow 360° rotation.

The indexing challenges on a servo-based turntable will be similar to any other implementation. Not having done it I would expect that one sensor may not be accurate enough and I'd start by assuming two magnet/hall effect sensors per position. Magnets placed on a metal pad glued to the board would allow easy adjustment of positions, later fixed with a blob of epoxy.

I like converting cheap 2.4Ghz receivers into other things. The cheapest have 7 pins for plug 'n play connections. One would drive the servo. The other 6 would read sensors. So I'd expect 3 positions are easy, 6 if a single sensor is accurate enough. More starts getting into the cost/size challanges that other approaches must have. Radio control is helpful for this because the control signal does not need a wire.

The new Menu switch could help with simple layouts because each switch position can control one or two points. One if each trigger button selects left/right movement. Two if each trigger button uses a 'toggle' principle. Should be feasible with a new 'Rx105' type product.
Regards, David.

Back To Top

 Posted: Fri May 29th, 2015 03:40 pm
  PMQuoteReply
15th Post
Geoff L
Registered
 

Joined: Tue Nov 4th, 2014
Location: Hemel Hempstead, United Kingdom
Posts: 44
Status: 
Offline
Hi again

The last three months have seen progress concentrated on building and first fixing the track, installing the modules on the baseboards and finding a way to operate the uncoupling magnets. Some photographs showing the progress have been added to the album, although it should be noticed that the backscenes have been temporarily re-sited to the viewing side for these. There has also been a slight adjustment to the siding alignment as compared to the original plan.
The good news is that the four position sector plate, the turnout servos and the uncoupling magnets are now responding to radio control. The only drawback is that the LED on the sector plate driver will need to be in view for operation. In normal operation the LED changes from green to flashing red when the switch position is changed and pressing the stop/run button causes the sector plate to rotate; under radio control, operating the stop/run switch has no effect unless one waits for the flashing red LED to change to solid red. Unfortunately the length of this wait seems to be variable.
Solid state relays are used for operating the uncoupling magnets, the 3.1v from the Rx output switching an infrared LED acting on a photodiode array which in turn activates MOSFET switches to pass the 12v current.
On the rolling stock front, I have just taken delivery of some 2s lipos small enough to (tightly) fit into the small tank engines to be used on the line. These are MiniAviation Blade 130 450mAh measuring about 48mm x 28mm x10mm which tuck into the tanks of the pannier tanks and cab of the 0-4-2T. They will also fit into the skirted underfloor compartments of the diesel railcar instead of protruding into the passenger space as at present, although with reduced running time. However, the connectors need attention to make them compatible with both the motor leads and the battery charger.
So far as the original challenge is concerned there are now only two of the original twelve months left!

Back To Top

 Posted: Wed Aug 12th, 2015 12:28 am
  PMQuoteReply
16th Post
Geoff L
Registered
 

Joined: Tue Nov 4th, 2014
Location: Hemel Hempstead, United Kingdom
Posts: 44
Status: 
Offline
Hi

It's three months since I last posted, mainly because progress has centred on constructing the layout rather than on radio control.
The twelve month target is past and the layout is unfortunately far from complete and I will give a fuller update in due course., however, Mynis Cule Magna has been selected as one of the finalists to appear at ExpoEM North in Manchester in mid-September.
My reason for posting is an update on the problems I previously mentioned I was experiencing with (lack of) control of the sector plate.
I found I was having a problem with the route selector interfering with one of the turnout servos and checking this out found that the only reason the route selector was working was because it was broken!
I mentioned previously that the control devices were set to change between 0v and 3v but part of the copper track on the strip board had lifted and some devices were not getting 0v. When this was fixed the sector plate wouldn't work. When the selector rotary switch was wired on/off, it started to work again, so the moral is "if it's broke, don't fix it!"
The same fix on the run button had no effect, but created a further barely noticeable break in the continuity to the switch.
With the switch fixed a long train of investigation found that there was indeed an almost instantaneous hit on the sector plate control board when it was activated. The sector plate operated with 5v but not with 3v from the Rx.
 Further frustration ensued when the spare solid state relay from the uncoupling magnet installation would not switch 5v using the 3v from the Rx. Silly me. I followed the data sheet wiring diagram for dc operation which showed the two output switch pins connected together and was feeding constant 5v to the run button.
Seriously contemplating using the 3v to switch 12v (in the same as for the uncoupling magnets) to operate a mechanical relay to feed the 5v, I thought, "this is stupid".
A trip to the workbench with a couple of batteries soon worked things out and the sector plate now works correctly.
This sets the record straight for anyone following the previous details and also explains why people couldn't see why I should be having a problem.

Geoff

Back To Top

 Posted: Wed Nov 8th, 2017 08:04 pm
  PMQuoteReply
17th Post
Geoff L
Registered
 

Joined: Tue Nov 4th, 2014
Location: Hemel Hempstead, United Kingdom
Posts: 44
Status: 
Offline
Hi All,

Having arrived back from the Eighteen Millimetre Gauge Society (EMGS) AGM and 4mm scale Modeller’s (sic) Day in Wells, Somerset on Saturday I thought an update was in order.
It is over two years since the last post and, needless to say, the layout did not reach the final because it was not reliable enough to be left in caretaker hands whilst I swanned off to Portland. With no deadline to meet development has been sporadic, but most of the required structures have been built ready for decoration and ground cover.
On the r/c front I have a rebuilt Mark II controller. A larger project box was obtained and the controls simplified with on/off switches and bind buttons for the two transmitters, just 12 way selection and direction/speed knobs for the locos and a 4 way route selector and 20 switches for the accessory controls (13 on/on and 7 (on)/(on)).
The EMGS test track was at Wells and I took the opportunity to test a re-wheeled Bachmann Pannier tank with replacement Nigel Lawton motor and a re-wheeled Lima Class 73 with replacement drive unit, both powered by nano-tech or similar 2S 450mAh LiPo batteries. The runs were curtailed due to time restraints but each ran for 45 minutes on the ca 30ft circuit at about a lap a minute or a scale 25mph. The 1350ft covered should translate into some 350 shunting (switching) moves on the layout. I aborted a run with a similar HO gauge diesel yesterday evening after 1¼ hours so still don’t know the full potential.
Further photographs of the controller, locos and layout have been added at the previously posted link.

Geoff

Back To Top


 Current time is 11:37 am
Page:  First Page Previous Page  1  2   

Top



UltraBB 1.172 Copyright © 2007-2016 Data 1 Systems