Suzie

5 hours ago, Izzy said:

The requirement for the OP, is for DC ( it is in the DC section), just using the DPDT slide switches that work the individual point tie-bars via tube/wire if possible, which it is, to keep it simple and easily produced. 

 

Izzy

 

added - actually this general question has been asked here before, in 2015, and clearly and simply answered by Clive Mortimore. The details are here: 

https://www.rmweb.co.uk/community/index.php?/topic/86612-home-made-scissor-crossing-and-control/

 

read his post, about halfway down the first page. Pretty much what I will be posting with specific details about the Templot configuration Ian will be building.

 

 

Clive ignored the OP's requirement to keep the frog triangles intact (only four isolating breaks in the whole scissors) and provided the same solution that AndyID has already proposed above (which has twelve isolating breaks).

1 hour ago, Pete the Elaner said:

3 states?

All points set to straight. The crossing can be dead for this. As long as it is isolated, does it matter which way it is powered?

All points set to crossing. Top rail needs to be the same polarity as top rail & bottom needs to be same polarity as bottom rail.

1 crossover straight & the other to crossing. Why bother? Every usable scenario has just been covered.

The problem with your scenario is the crossing frogs - the polarity reverses between crossing one way and crossing the other way. There is no way round it other than switching in conjunction with crossing one way or the other - or separately switching those frogs.

1 minute ago, Pete the Elaner said:

I fully agree that only 2 states need to be considered: All straight & all crossed. Any others involve running in to a point not set.

With all points set to straight, the crossing needs no power at all, so I would isolate all rails. YOu only then have to consider how the crossing needs to be wired when the points are set to use it.

You do not need to take power from any switched frogs. If you only consider the crossing with the points thrown, you can take the feed directly from one of the main line feeds. If you have cab control, you could choose which circuit to include the crossing in.

So I would ignore the feeds from the frogs to the crossing, which are wrong anyway.

The polarity of the crossing frogs has to be changed depending on which crossover is being used. It is not possible to set 'all crossing' when using a live frog diamond in the middle. There are three states!

The feeds from the point frogs to the crossing frogs are quite correct when using Andy's scheme which requires twelve isolating breaks on the six frogs.

2-pole switches will require breaks around the four point frogs (eight of them) so that you can just link the crossing frogs to the point frogs and have one pole per point frog.

I tried very hard to rework it but DPDT + TPDT is the best I could get either way while maintaining the structural integrity as far as possible with breaks just on the crossing frogs.

1 hour ago, AndyID said:

 

And you are absolutely correct!

 

You might getaway with two states for the points but the diamond's frogs must be set according to the route.

Your diagram will work, it just needs a few more insulated gaps than my method. The discipline of only setting one crossover at a time still applies.

1 hour ago, AndyID said:

 

If Ian promises to never foul the points he can make it even simpler and go with a two-state (all straight/all crossing) version in which case he won't need any additional switches

The issue for this method is that the diamond frogs will need to be set depending on which way you are going through the crossover - that is why there are three states and not two.

1 hour ago, AndyID said:

Ian's scissor crossing is in a goods yard. I'm not sure he's going to need that much complexity

If it is 'one engine in steam' then all four feeds will be the same and that will simplify things a lot. Context is all!

Back in the 'Olden Days' when DAB first arrived we had real choice like The Storm and The Arrow - but it has all been dumbed down now with the closure of The Storm and The Arrow becoming just music streaming without DJs.

Planet Rock was OK to start with but now it is in mono with bandwidth that makes Medium Wave sound like good quality with pretty useless DJs and much reduced coverage and being on a multiplex that my car radio can't receive it is just pointless. Back to playing MP3s now for me I think.

I suspect that most of the stations that have closed or economised "Because no one listens" like 3C country and the Saga stations have just failed to find out how big their audience actually was - my dad's radio is getting very dusty now through lack of use.

I wish you luck in finding a broadcast station but I suspect you will be disappointed until there is a bit of a shakeup in the industry. Someone has obviously done some market research to find out what people like to listen to and has failed to contact anyone who was not completely vacuous.

The type shown are quite widely used in industry and can be very reliable when crimped properly. In many industrial applications ferules are crimped on to the ends of stranded wires before being inserted in to screw terminals to improve performance of the screw connection! 

There are three states in practice which makes the wiring a bit more intuitive:-

• All straight
• Crossing bottom left to top right
• Crossing top left to bottom right

I am assuming that you will feed both routes from one end and not require power routing and that you can operate the two points of each crossover with a single switch. DCC would have made this a lot easier, but you should be OK.

Note that the frogs of points 1 and 3 are connected together as are the frogs of points 2 and 4 to make life easier.

I could not do it all with DPDT switches, so you might have to add a microswitch to point 1 to get the extra pole.

If you need power routing you will need extra poles for that too. Wiring is common return with the black feed being the common.

Once you have made a crimp connection you need to give the wire a substantial pull to make sure it really has crimped correctly. If it comes out you need to do better! It is mostly about using the correct size crimp for the wire. With practice you get a feel for how well you have done but always do the pull test - if you can't pull it out it will not fall out.

These crimp connectors work well for their intended use when they are crimped correctly. You need to choose the correct size connector for the wire you are using and use a suitable crimping tool to fit them. When used correctly they do a very good job of making a circuit that is not disconnected very often.

The colour of the insulation usually denotes the size of wire that will fit. Yellow ones for big wire and red ones for small wire, with blue ones (as in your picture) for middle size wire. The correct crimping tool is not expensive and will usually come with a range of crimping places which if you are lucky will be colour coded red/blue/yellow so that it is obvious which bit of the jaws to put the crimp in to. You just squeeze to make a connection.

A good crimp will be better than a solder joint or choc block, but it is horses for courses.

There will be either cutting tracks and jumpering to the F1 and F2 pins, or just moving the wires for one end. Pads might be provided if you are lucky.

8 hours ago, melmerby said:

Surely they aren't saving anything because instead of using the bottom of the addressing range (in hex) there is a significant offset, e.g. instead of decimal 10239 being 27FF (14bits: 100111 11111111) it is actually E7FF (16bits: 11100111 11111111)?

 

That's what my original query was based on. Why use a 16bit address but only use 14bits?

The two most significant bits are "redundant" but have to be set.

The top two bits code the packet, DCC sends different types of packets not just mobile decoders - there are accessory decoders too.

You will need to do a bit of rewiring since F0f is headlight at one end and tail light at the other, with F0r being the reverse. I am not familiar with how to reprogram the ESU decoders, but I suspect you should be able to move the tail lights to F1 and F2 (green and purple) and get the decoder to do as you wish. I know that a Zimo will do it...

Short address 0 is the broadcast address, so it is quite sensible that you can't directly control it - it would drive all locos at once! Manufacturers of command stations often hijack short address zero so you can control a DC loco instead of the broadcast address.

Long address 0 can be used though (wiki linked above suggests it can't!).

I suspect that the NMRA did not expect anyone to have more than ten thousand locos on a layout at a time and therefore it was not worth gumming everything up by sending an extra byte for long address when 14 bits could be accommodated in the existing plan.

The Krauss-Maffei ML4000 was a C-C 'double Hymek' with a pair of 16 cylinder engines but had an axle weight of 26t so I think D-D will be more appropriate for the weight restrictions likely to be encountered in the west giving just less than 20t axle weight more in line with B-B Hymeks.

After all a C-C Hydraulic with a pair of V12s would be a Western!

1.5mm speaker wire will be fine for DC.

6 hours ago, 34theletterbetweenB&D said:

There's an imaginary loco project for someone, design a robust high output DH for BR use.

Here is my 7000HP suggestion from page 52:-

Should be quite hard to overload that pair I think. Still trying to get hold of some old Hymek bodies and peak chassis to have a go at making one up.

The biggest disruption connecting loads such as this to the DCC track bus will do is stop Railcom from working. If you use Railcom you need to feed the lamp circuit via a pair of inverse parallel diodes.

You will create asymmetry on DCC if you half wave rectify the track supply for large loads. 

There is a trend to move away from the ring main anyway which is not required when fatter wire is used, and unshuttered European sockets which don't require fused plugs are allowed (probably not for long), but I think the polarised Cenelec ones used in France would be better than the reversible Shuko type used in Germany when legacy UK items are likely to be used with fuses and switches in the live only. Perhaps the blue 16A sockets might become commonplace for high power items like tumble dryers going forward.

I rather feel that with the socket in that state the circuit needs to be isolated until the socket is replaced. Sticky tape is not good enough to ensure safety.

Cause of the failure is going to be hard to determine, but in the mean time you might like to run the tumble dryer through a plug-in RCD if you do not have one fitted since you cannot rule out a possible fault in the tumble dryer that may still be present. Fit a good quality (MK or similar) plug to minimise that possibility of the plug failing. A loose fuse is likely to result in a similar situation.

This is the sort of thing that happens when sockets are worn due to too much plugging in and out, and the unnecessary use of 'Socket protectors'.

The meter in current mode needs to go in series with the motor! Connecting it across the motor is applying a short circuit and the decoder will not last very long, especially when drawing 5A.

No is the simple answer. The Seep motors take around 4A each, so even just using one will overload the LS110 output unless you keep the voltage below 12V, at which point operation will probably be unreliable.

Peco make solenoids that are suitable for use with the LS110, the PL10W and PL10EW and you should be able to use a pair of these OK. Power the LS110 from 24V DC from a suitable switch mode PSU for maximum reliability of operation.