Nigelcliffe

Thanks guys for your advice. I'll see if I can get the wheels reprofiled first then if that fails, have a bash at a scratchbuilt chassis. However I have never done this before so will be a challenge for me.

Before you start, you may want to check the availability of spare parts. If it reprofiling goes wrong, you have wheels which don't work with N or 2FS.

You've picked a difficult starting place for a conversion or re-chassis. Outside valve gear and lots of wheels usually makes things complicated.

- Nigel

As Chris says, theory tells us about thrust bearings or other couplings between worm and motor. Practise tells me that in small locos it doesn't matter. My DY1 has done a fair amount of running in the ten or so years since it was built, though it doesn't do Copenhagen Fields mileages. The DY1 is on its original Faulhaber motor, worm direct on motor shaft. Other locos have similar arrangement but fewer years running.

- Nigel

 

 

Once back together things were slightly better. Testing various combinations with the rods, it seems everything is fine with the left hand set on as intended, but things go stiff whenever the right hand rod is introduced.

 

I imagine this means the central hole on the right hand rod needs a bit of reaming? Or there might be a problem with the wheel itself?

Could be anything on the RH side. So, do it methodically, turn the RH rod over and fit the rear crank hole over centre axle, centre hole on rear axle. Study what happens. Now put front crank hole on centre axle, centre hole on front axle. Study again. Then decide where the tight spot lies. It really is worth getting the article mentioned !

Before you cut anything check the crankpins very carefully for perpendicular to wheel and, if you can, crank throw.

- Nigel

However, putting the intended connecting rods on results in it rolling smoothly with finger pressure applied from above, but not so freely that it will roll under its own weight. I reamed the holes slightly, but the fit is exact. I used the Association quartering jig so I assume the quartering must be spot on. The back to backs seem correct too - bobbin gauge is a snug fit. Should I ream the rods much further? Or is there anything else to check?

 

Its time for systematic checking to find where the tight spot may be.

The cause could be rod holes needing slightly enlarging, or it might be a bent (or mis-aligned) crankpin in a wheel. The latter isn't common, but it is also not totally unknown in 2mm wheel production.

Rod setting is covered in an article in the April 1995 magazine (backnumber archive!). I think its repeated in one of the booklets, possible the split-frame chassis booklet. The basics are to remove the rods, turn them over and refit so the loco is a 2-4-0, study whether there is a tight spot with the 2-4-0. Now repeat again as a 0-4-2.

I'd expect a small enlargement of a rod hole is what's needed. But, which direction it needs enlarging is important.

- Nigel

They are. If we're being pedantic they're helical or skew cut, which makes them suited to use with the worm, but they are still spur gears and the description indicates they are for use with the worm. I'm sure Nigel would listen to any input regarding how things might be amended to improve clarity? (time for webmaster duties permitting, of course)

Descriptions aren't my bit. The text on the listing comes from the chief shop keeper as part of the pricelist file (product code, description, price, etc..). The spaghetti then displays the file on the website, finding any photographs or information files associated with that product code.

- Nigel

Andy's description is correct. The pictures need moving.

Yes, Andy's correct on the details.

- Nigel

I need a 21:1 worm and gear set - which is the correct association part number? The picture seems to be with the wrong product on the website, unless I'm wrong?

21:1 is the 3-358. Picture looks correct to me, the highlighter on the label indicates dimensions of worm shaft and wheel bore.

- Nigel

Knowing Julia it will have been fitted with DCC couplings, lights, and separate handrails by 4pm.

Can't see any point rodding on that track though....

There already are "finescale Brio" couplings; there pictured on the older of the banners which accompany the PR stand, towards the bottom. Invented by one of the East Anglian Wheel Elves.

- Nigel

Dave,

glad to see its working with the shuffle. I might try to tweak the shuffle speed down a little further, but its a personal preference thing.

Also, if you want the shuffle to work further back in the train (eg. detaching a guard's van) its necessary to increase the distances that the loco shuffles back and forwards to allow for the compression of couplings along the entire train.

Final issue, anyone using a PowerCab has to either stop the train with the throttle at the end of shuffling, or alter the PowerCab's setting for when it sends stop-bits. Otherwise the Zimo loco will keep running after it has finished the shuffle. I've only seen this issue on NCE controllers. The underlying issue is complex, but if anyone really wants to know, send a PM.

For under-track magnets, it needs something much stronger than the standard magnets. Magnetic fields decay at the square of distance. At track level, the gap is perhaps 2mm at most, whereas below the track, the gap will be nearer 4mm. So, that would be 1/4 the strength by doubling the distance.

When I used Microtrain (Kadee) couplers on a US N-scale layout, I used Kadee O-gauge magnets, cut in half, under the track. They'd work through a thin (perhaps 1mm) wood or card base, plus Peco code-55 sleepers to the rail.

Were I doing it today, I'd try a pair of modern high powered "super" magnets below the track, approximately below each rail at a slight diagonal line. One would need the N-pole upwards, the other the S-pole upwards. I'd try different sizes until I got the one which worked best. I'm not sure if a metal magnetic shunt between the two magnets (bottom faces) would assist or not, so further experiments required.

- Nigel

 

I've also used the 14.2 axle length in the past but having recently acquired a little Unimat I'm thinking of turning the wheels down - they are very fine for N and require very little taking off.

 

Someone has suggested just pulling the Mathieson wheels off their axles and then pushing on 2mm ones.

- Nigel

Dave, for it's next trick, program the "Kadee Shuffle" into the Zimo and a single function key press will perform the backup and pull forwards to uncouple move for the Dapol couplings.

Hi Justin,

 

If you use a Lenz silver you will be able to adjust the lights to whatever level you need to.

 

Peter

Or, there is the much cheaper, superior running qualities and adjustments in a Zimo MX621.

(Zimo is £27 vs Lenz at £35 at my usual supplier. )

With any chip there is a limit to light adjustments, and this comes in two places....

First, the lights are wired in pairs. So, turning down the "no 2 end" red lights will also turn down the "no 1 end" white lights. This can only be fixed by rewiring the loco so the lights are independently controlled by different function outputs on the decoder, and then plug-in decoders won't do the job, so its a wires and solder connection. (Situation may improve if makers move to PluX connectors rather than the 6-pin).

Second, eventually at very low output levels you risk seeing the pulsing of the LEDs from the pulse wave modulation used in the decoder to dim the lights. It may be better to replace the resistors in the loco with higher values to get the worst of the light levels down.

- Nigel

Ask Roco about their CAN bus. Very recently they had a survey as to which CAN to support. Their survey did not include the MERG option.

Nigel

Hi

Is there a quick way to straighten bullhaed rail from a coil?

Many thanks

Ray

Yes there is a quick way, but it involves a fiendish machine built by a chap from the Darkest corner of Darkest Essex (I understand its called Darkest Essex because the power goes off regularly!). The machine probably breaks a few electrical safety regulations, so don't try it at home without asking a responsible person.....

The machine is the hollow heating bar from an electric fire, about 2ft long, attached to a strong metal box structure. Thread rail down centre of heating element. At one end of the element is a fixed vice, fit one end of rail to this. At the other end is a vice attached to a strong spring which keeps the rail under tension, fit other end of rail to this. Pull spring (there is a lever to help with this, and a latch to hold it) so that rail is under tension. Turn on power to the heating element for a few seconds until the sprung vice is just seen to start to move. Power off as soon as any movement is seen, and the residual heat allows the rail to be stretched a little further. Release rail from both ends and use. Repeat for next length of rail.

Without such a machine, then straightening is a slow process of running rail between fingers, etc.. For solder constructed track the rail needs to only be relieved of some of its curve, if the rails are soldered with the curves in opposite directions then the tension in the track is evened out. I suspect that curved rail doesn't work well in plastic track bases.

- Nigel

See my comment no 338 here http://www.rmweb.co....325#entry725126 where I've listed a very simple alteration that makes them perfectly usable.

 

Andi

In this thread there are the odd useful bits of information, such as Andi's above.

A MERG member (on the MERG forums) has come up with an alternative method for fitting to DCC accessory decoders which deliver a pulse (typically for solenoid motors such as Peco or Seep). The earlier thread entries suggested a relay, which does work, but at modest cost for the relay. Like using a relay, this method does not invalidate the Dapol warranty by opening the signal, the electronics is simply closing the two switch contact wires, exactly as if wired for manual push button operation.

The alternative to a relay is an Opto-Isolator (also known as a "photo coupler"), which is a solid state device costing around 20p. Wire the Opto-Isolator and a 1kOhm resistor in series to the accessory decoder's outputs ( the resistor is important, the Opto-Isolator contains a couple of LEDs inside the package, and the resistor is needed to control the current in the LEDs, just like using LEDs for loco lights!).

The other two pins of the Opto-Isolator then connect to the control wires on the Dapol signal.

When the accessory decoder is operated (for a brief pulse, as set for operating solenoid motors), the Opto-Isolator "closes" the contacts on the signal, and thus moves the signal arm.

One suggested Opto-Isolator is a KB814, available from MERG (for its members) or from Rapid Electronics (mail order), and no doubt numerous other electronics suppliers. Ebay may be the cheapest place if just wanting one or two, rather than the fairly high minimum order and post charges from most commercial electronics suppliers.

Note, the Opto-Isolator solution does NOT fix the lack of deterministic movement in the Dapol signal; only a redesign by Dapol or opening the case and modifiying the circuit (as Andi suggested in the link above) will fix that design mistake.

- Nigel

Thanks Ron.

 

Can anyone talk me through or point me in the direction of how this is done? Essentially I'd like to know the procedure for finding out if a loco has a Lok 3.5 or 4 fitted.

 

Thanks

Martin

There isn't an easy way to determine whether there is a 3.5 or 4.0 from CV reading.

Unfortunately, ESU don't put version numbers on their decoder CV's which could make the distinction straightforward. Other decoder makers do provide such information, and can be very easy to identify by CV reading.

Best strategy I know is to adjust the primary volume control. Try values around 60-63 and the LokSound 3.5 will be at its maximum loudness, increasing the value further won't raise the volume. Whereas a LokSound 4.0 reaches maximum volume around 190-192, and a value of around 60-63 will be only a fraction of the possible maximum volume. Decoder read-back is not required for this approach.

( For what its worth, I think that buying a ProBox purely to read back CV values is very poor financial value for money. A Sprog will do this with a computer interface for about £60 (leaving enough change from the ProBox to buy a sound decoder!). Or, if you want a DCC system keypad rather than computer to do this, an NCE PowerCab or Digitrax Zephyr Xtra purely for the bench costs less than a ProBox. There may be other features of the ProBox which makes it a good value item for some, but I struggle to identify them. ).

- Nigel

And when you sqeeze up the drivers to correct BtoB how do you keep muff central to stop is pinching the wire (between muuff and bearing) thus preventing it functioning as intended?

Using the shim pieces supplied with the quartering jig :-) Wedge those in before tightening up, then pull out when finished.

The lathe method has a honourable tradition; Bill Blackburn has a set of tools to do just the same, and the quartering jig came about by working out how to replicate Bill's jigs in as a useful tool.

- Nigel

Well Chris I'm no engineer so you can perfectly well say I'm mad or stupid and I won't mind but according to the gear meshing tables the centres for gears are quoted as follows:

 

64DP 14T to 18T = 6.35mm

M0.4 14T to 18T = 6.40mm

100DP 18T to 31T = 6.35mm

100DP 20T to 30T = 6.35mm

 

I tried the 3rd and 4th options and they both work perfectly well with apparantly good meshing but decided on 18/31 and it has survived testing under load for three days continuous running on a shuttle circuit.

 

My rationale is that 100DP gives you a greater number of ratio options and ability to have a bigger ratio thus enabling the loco to run more slowly and with better control for shunting. Which is what I prefer.

 

Tony

Tony,

I don't know which tables you were following. The 100DP tooth counts you list are not standard shop items.

These centres come from the 2mm website "gear mesh calculator", which I wrote and am reasonably certain is accurate.

64DP 14T to 18T, theoretical perfect mesh 6.35mm, with running clearnances = 6.48mm

M0.4 14T to 18T, theoretical perfect mesh 6.40mm, with running clearance = 6.53mm

100DP 18T to 31T, theoretical perfect mesh = 6.22mm, with running clearance = 6.35mm

100DP 20T to 30T, theoretical perfect mesh = 6.35mm with running clearance = 6.48mm

Gears, gear mesh calculations and related items are discussed in a pair of articles by Denys Brownlee and Henk Oversloot in a 1990's 2mm Magazine, which is in the archive CD set.

- Nigel

For an identification on an analogue control panel of the way the signal is set a point lever or similar monentary switch could be used with both outputs wired together such as those sold by Hornby or Peco for point motor operation.

Nice idea, but doesn't work, for the reasons explained by Andi (Dagworth) a few posts above. (27 June 2012 - 23:37).

As the Dapol signal can adopt a random direction on first control pulse, the lever cannot indicate signal position.

- Nigel

I've just stumbled over this, it is superb running, well done.

A couple of suppliers of bits if Giles hasn't tripped over them yet -

MicroAntrieb in Poland. They do various motorising bits and pieces for RC control of small model road vehicles (some for HO scale), there may be something of interest in their range. Some of the documentation is in German.

http://www.kkpmo.com/homeENG.html

Plantraco Microflight in Canada. Makers of very small RC aircraft bits and pieces, the receivers and servos might be useful.

http://www.microflight.com/

- Nigel

Thanks Andy. I didn't realise Judith Edge did the O5 in 2mm - I was aware of the 4mm version. I was actually going to try and scratchbuild the body! For the Hudswell Clarke I saw in the N Gauge Journal that Parkwood are doing a resin body for the later series (D2510-2519) - the picture is actually mounted on a Farsh O4 chassis.

I saw the N Gauge Journal article on the resin bodies, I was very underwhelmed. If that is the only "kit", I would advise scratchbuilding.

The Judith Edge 05 etch is the 4mm one shot down, so its a set of etched parts, to Michael Edge's usual standards. But, no work done to the etch to make it compatible with 2mm parts and no castings/mouldings. In its favour, its a fairly big shunter, the bonnet area is quite large compared to other small BR shunters, so it should not be particularly difficult to get a good mechanism and lots of weight inside one. I covered making the skirted 4-wheeled DY1 version (Ipswich docks, later Goole docks) in a 2mm magazine article some years ago.

- Nigel

 

Nigel have you made your coil etches available yet as I would like to try your dcc uncoupler?

I have hundreds of them, so drop me a line off list.

They are primarily aimed at the 4mm scale modeller, but some of the parts on the etch may be useful in 2mm adaptions.

Coils come from Plantraco Microflight in Canada; they sell "nano servo kit" which is a coil and two magnets. Its clearly not a proper "kit" as there are no instructions!

Nigel

I have a feeling that any claim to the design principles would be hard to establish as a lot of the principles are shared with other designs ( loops, buffing plates, etc) however the artwork would definately be subject to copyright. If I were to take my etches obtained 30 years ago from Nick and copy them even in part I would be in breach of copyright unless I had permission. However if I was to redraw the couplings and produce my own artwork it would be a different matter.

There is no copyright possible of an idea or mechanism, so the principles of operation are not covered by copyright law. They could be covered by patent, but that would not be possible once the invention has been disclosed to the world.

Copyright would not be breached if the same item were drawn afresh. As someone is not likely to produce an etch drawing by scanning an old etch, a breach of copyright is not likely to occur. In the event of a breach of copyright, the next issue is to show the amount of financial loss, which in very rough terms is the lost sales. But as a breach is unlikely to have occured, then no case will occur.

Its a different matter about politeness: most designers of model gubbins I've met like to hear about improvements or enhancements. So one can gain permission, information, friendly relationships and a constructive environment by consulting people.

- Nigel

Cheapest decoder option is probably Paul Harman's DIY design; it was originally done for coach lighting to reduce the cost of controlling the lights in a rake of coaches. Depending how its built, it can have up to eight function outputs per decoder.

But, it does mean assembling a circuit on vero-board. Around £3.00 per decoder if you have a PIC programmer to load the code into PICs bought from Rapid or Farnell.

Nigel