Dapol working signals review

[MRDEEDIESEL]

Well Dave as an N gauger.... im delighted with programme of new products from Dapol. especially the semaphores so please dont get disheartened.I hope that you progress your range and im sure they will get better and better.I can also say that about Farish as their Locos,coaches and wagons are quite frankly marvellous(eg The new 20)

[The Stationmaster]

Because you can't always see the signal - sadly a reply like that will do nothing to demonstrate understand of the subject and therefore some of the concerns outlined.

It is now well over 50 years since I first tried to pull off a signal - and the repeater (of a North Eastern Railway slotted post distant) fell to 'wrong', but it had a repeater because it was around a mile away. Since then, having pulled more than a few signal levers I've only ever needed to look at a repeater for signals I couldn't see (I've never scored another 'wrong' either come to that) because they were too far away to be seen with the Mk1 eyeball.

 

Slightly under 50 years ago I had a loft layout which was around 15 feet long, the running lines were equipped with multiple aspect colour lights controlled by a home designed and constructed NX panel with full interlocking between points and signals but no signal repeaters to tell me if the relays had done what they needed to - if I was in any doubt I looked at the signals and points.

 

The moral of this tale - model railways mostly aren't like real railways to the extent of needing things like signal repeaters and full interlocking and so on. For some folk these things are 'nice to have' - for a very large majority having a working signal in, hopefully, the right place is a huge step forward and they probably want to operate it in the simplest way possible. I doubt if more than a handful of them have ever heard of a signal repeater or know what it does and why it does it - and they don't really need to because on the vast majority of layouts they can see the signals from the control position.

 

Anyone who needs to fit a repeater to a Dapol signal could probably manage it but otherwise it is just like a full size signal in sight of the controlling signalbox - the only way you will know that it has worked correctly is by looking at it (as the Rule Book requires on the big railway).

[Stubby47]

If the layout needs repeaters, you could always buy two of each signal and place the second one next to the control panel, with both fired off the same switch...

I'm sure Dave wouldn't mind

[Dagworth]

The problem with all this talk of repeaters etc. is that it completely misses the point. I can build repeaters till the cows come home, but my signalling is interlocked. That means that the points can only be changed when the signals protecting them are at danger. My drivers drive to signals, not to some control panel. They KNOW that if the signal is off then the points are set and the line is clear for them to proceed. Telling me to look at the signal doesn't help, the signals are worked by computer. Without some means of either feedback to tell the computer the state of the signal - or a control system that I can set up such that electronic state 'A' has the effect of putting the signal to danger (or point to normal) while electronic state 'B' puts the signal to clear (point to reverse) - then the computer has no way of knowing what 'aspect' the signal is showing

With a system as designed into the Dapol signals I have no state 'B', I only have the option of contact open (stay as it is) or contact pulse closed (change position). Worse than that, sometimes the first pulse after powering up the signal will try to drive it to danger (when it's already at danger) and sometimes will drive it to clear. I have a very real possibility of the signal showing clear when it should be at danger. My drivers will take the signal as clear meaning the route is clear, the track circuits are vacant, the points are set and locked. I have a WRONG SIDE FAILURE. Any railwayman will know that this is the worst type of failure you can possibly have in signalling.

 

If you don't understand why this is a problem then come and see how my layout is operated at Members day next weekend

 

Ironically it would be vaguely correct as a method of working hand points, where pulling the lever (pressing the button) puts the switches to one side, releasing the lever puts the lever ready to be pulled again to move the switches to the other position. Signals DON'T work like that. You pull the lever to move the signal from On to Off, and put the lever back to put the signal back to On. (Yes I know about expansion causing mis-shown signals, hence the need the repeaters shown up the thread - and the 'wrong' position in the middle - but for our purposes that can be ignored.)

 

Andi

[beast66606]

The problem with all this talk of repeaters etc. is that it completely misses the point. I can build repeaters till the cows come home, but my signalling is interlocked.

 

Interlock off the repeaters then ??

 

They KNOW that if the signal is off then the points are set and the line is clear for them to proceed. Telling me to look at the signal doesn't help,

 

Do you have full detection on both blades ? otherwise they don't KNOW that the points are set, they only KNOW the points have been called by the route.

 

However

 

Without some means of either feedback to tell the computer the state of the signal - or a control system that I can set up such that electronic state 'A' has the effect of putting the signal to danger (or point to normal) while electronic state 'B' puts the signal to clear (point to reverse) - then the computer has no way of knowing what 'aspect' the signal is showing

 

Agreed - the control system could have been a lot better, HOWEVER, this (hopefully) is the start of a series of ready to plant signals, at least they are available and operable, even if a bit strange in the method. Maybe Dapol will change things when (if) Version 2 appears ?

 

If you don't understand why this is a problem then come and see how my layout is operated at Members day next weekend

 

There are some of us who understand very well, Widnes works exactly the same however we don't rely on line of sight, we also have a software repeater, in case the drivers are blocked by other operators, plus,imho, operators wandering around just looks silly.

 

Signals DON'T work like that. You pull the lever to move the signal from On to Off, and put the lever back to put the signal back to On. (Yes I know about expansion causing mis-shown signals, hence the need the repeaters shown up the thread - and the 'wrong' position in the middle - but for our purposes that can be ignored.)

 

Not correct - although your argument about the control system still holds - there are some signals which stand off with the lever normal, PULLING the lever puts them to danger and replacing the lever puts them to off.

 

No-one - that I've seen - is denying things could be better, but a start has to be made somewhere, that start has been made, the issues with the control system have been highlighted (too many times, which is the point that's wearing thin) and perhaps future models will reflect these comments.

 

PS - I'm not even sure why you're too worried at this stage, your layout doesn't need any main semaphores does it ?

[Dagworth]

Dave, the layout does require semaphores out of the NCB.

Interlock off the repeaters then ??

 

 

Any repeaters can't guarantee that they accurately repeat what the signal is displaying, the signal can be in either state.

 

However, I'm pleased to say that I have FOUND THE SOLUTION!

 

(photo borrowed from further up the thread)

 

Solder a very fine wire to the pad marked with the red arrow. This wire will go negative when the signal is 'Off' relative to the red supply wire (I'm running mine on the test bench off a 9v battery) sufficient to light an LED connected with a 1k series resistor between the additional wire and the positive terminal of the battery. I therefore have the repeater that I need that is always linked to what the signal is actually showing.

 

YOU WILL VOID ANY WARRANTEE by carrying out this mod, proceed at your own risk.

 

Andi

Edited June 27, 2012 by Dagworth

[beast66606]

An elegant solution

 

 

I'm still surprised you need semaphore main arms but that's wandering OT.

[Edwin_m]

That's a good one Andi. I was wondering whether something could be done with the limit switches but put off suggesting it by the general flak that seemed to be aimed at anyone suggesting things were not perfect.

 

Dave sums it up well:

 

Agreed - the control system could have been a lot better, HOWEVER, this (hopefully) is the start of a series of ready to plant signals, at least they are available and operable, even if a bit strange in the method. Maybe Dapol will change things when (if) Version 2 appears ?

[coachmann]

Speaking as one who pushes points over with his finger (!), I fully expected to do the same with signals, so any kind of electrical working is a bonus. I'll be more concerned with adding detail or changing arms on the Dapol's and I wondered how far in this direction other modellers on here had got. One consideration is fitting LNWR arms to the GWR lower quadrant signals.

[Gordon H]

Solder a very fine wire to the pad marked with the red arrow. This wire will go negative when the signal is 'Off' relative to the red supply wire (I'm running mine on the test bench off a 9v battery) sufficient to light an LED connected with a 1k series resistor between the additional wire and the positive terminal of the battery. I therefore have the repeater that I need that is always linked to what the signal is actually showing.

 

YOU WILL VOID ANY WARRANTEE by carrying out this mod, proceed at your own risk.

 

 

Simply using a 9V battery has already voided your warranty, I suspect.

Will the mod work with the recommended (insisted?) supply of 16V AC?

Edited June 28, 2012 by Gordon H

[Dagworth]

Simply using a 9V battery has already voided your warranty, I suspect.

Will the mod work with the recommended (insisted?) supply of 16V AC?

 

I don't know, I don't have AC round the layout, so I'll be running the signal from one of the DC supplies, probably 5v, once I install it, so I haven't tried it on AC.

 

When I first tried the signal on the battery I got no life in it at all with the supply leads connected with black to positive and red to negative, but it works fine with the two wires connected as one would expect for DC wiring, red positive and black negative. I'm sure that an earlier posting said that the unit contained a full wave rectifier but evidently not.

 

Andi

[Grovenor]

When I first tried the signal on the battery I got no life in it at all with the supply leads connected with black to positive and red to negative, but it works fine with the two wires connected as one would expect for DC wiring, red positive and black negative. I'm sure that an earlier posting said that the unit contained a full wave rectifier but evidently not.

Well now, mine most definitely has a full wave rectifier and it works equally well whichever way round I connect the battery, suggest you try again.

Simply using a 9V battery has already voided your warranty, I suspect.

And how would Dapol know you had used a battery, its not going to show, but taking it apart and soldering wires in, that's rather obvious.

Regards

Keith

[Butler Henderson]

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.

[Nigelcliffe]

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

[Dagworth]

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.

Potentially that would make the problem even worse, by linking both outputs you would be sending a double pulse and risk the signal reading them as change - change (go back to where you were*) in one movement of the lever...

*wherever that was in the first place

 

Andi

[Free At Last]

A Peco passing contact switch only pulses once per throw, unlike the Hornby one which does as mentioned.

[Dagworth]

A Peco passing contact switch only pulses once per throw, unlike the Hornby one which does as mentioned.

Ah cheers, I didn't know that, I've never used one. (Or taken one to pieces...)

 

Andi

[The Stationmaster]

Further Impressions - and an Interesting Discovery

 

Well I've now done some power tests on a 3 signals in my possession and all work as Dapol intended but definitely a need to avoid finger trouble because as discussed above the slightest pulse on the two yellow wires makes the arm move. However that apart the mechanism does what it says on the tin and one pulse makes the arm move and it stays where it is until it gets the next one.

 

The only potential problem when the signal operates is the angle of the arm at 'off' and it is - in my view - excessive, it goes too far so I wonder if it can somehow be adjusted? (don't forget I am talking about the GW signals in 4mm scale - see earlier posts 245 & 261 in this thread)

 

And so to getting down to work - well that will be in a different thread when I write it up - but during some preliminary fiddling about today I decided to measure the arm in readiness for work on the white band (and of course the black one on the back of the arm) and I made an interesting discovery - the arm is too long. I mentioned previously (post 245) the seemingly odd relationship between the red & white on the stop signal arm and I have wondered a bit about the overall proportions. Today I checked the proportions of a 4 foot arm on several drawings contained in Vaughan's book and then began to assess them on the signal itself - the arm is very slightly lacking in depth but the red/white coloured portion (on the model) which represents the timber part on the prototype is a tad over 1 millimetre too long (at 17mm and a bit) when it should be 16mm to equate to the 4ft of the real thing. This sounds very little but the extra length emphasises the lack of depth and makes the arm look unduly 'skinny' in my view.

 

So, as part of the detailing process, I resorted to the simple process of using an X-Acto chisel pattern blade to cut the arm back to the correct 16mm length. The result was well worth it although it doesn't show up all that noticeably in the pic below so I'd better add that the shortened arm is the one on the right and of course the relative position of the white band has also moved - for the better - although it still isn't correct. The observant will also note that the right hand signal post is part way through attention from the chisel blade, also revealing that underneath a pretty tough original paint job it's moulded from black plastic.

 

[James]

Interesting to see work on the Dapol signals Mike; would it be feasible to fit etched arms? I feel that might have a huge visual impact.

[The Stationmaster]

Interesting to see work on the Dapol signals Mike; would it be feasible to fit etched arms? I feel that might have a huge visual impact.

 

I've no doubt that it would make a big impact James but alas I've none to 'try for size'. What I do have - when I can find them (usual story of course) are some Ratio arms which are almost exactly to scale apart from a beefed up spectacle plate area and they will make an interesting comparison. Incidentally one thing I haven't done is measure the arm thickness and while it is clearly overscale it isn't, I think, too bad for a timber arm - perhaps I ought to pop out to the shed and measure a real one (a Reading original, albeit centre pivot pattern) for direct comparison?

 

But - and I still need to investigate this as I get to grips with work on the signal - is the question of what can or does go where in terms of pivot points and lamp positions. The reason for this is that the Dapol drive is arranged completely differently from the real thing and would require almost a start from scratch rebuild to alter. In summary the Dapol drive rod is at he back of the signal - probably mainly for appearance reasons (in which case the plan works admirably Dapol Dave) but this has resulted in the arm being driven via the back blinder, an unusual approach but it works. In theory this makes the arm potentially easy to change in some respects but that comes back to the geometry and lamp position etc; it might be possible, it might not.

 

In the meanwhile the simpler course - which is my aim at ptresent - is to alter the appearance of the front of the signal by adding a dummy down rod and balance weight lever which, with a paint job, should change the appearance considerably. It's thus much more in the realm of simple r-t -r detailing rather than, at this stage, a very major hack and reconstruct job; but alternative arm possiblities, including LNWR, are not forgotten. And as I might happen to need a few former LNWR signals at a future date doing something in that direction wouldn't be wasted

[steam in 50s]

My first Dapol LMS Signal arrived today. hole drilled and installed in under 10 minutes. Well done Dapol i can see where there could be issues with some people , but not me because a little weathering and its great. Its been along time coming but hey in my book well worth it. i am a modeller and i think i will fit a dummy counter weight at its base. The base is not all that big and with a bit of grass and or ballast round the edges it will be alot smaller. Brilliant Dapol.

[Stubby47]

In general, should there be an optimum angle to which the signal arm raises or falls away from the horizontal, to show the 'proceed' aspect ?

 

I had assumed an angle of 45 deg would be about right - my GWR lower quadrant drops about 65-70 deg.

 

[The Stationmaster]

In general, should there be an optimum angle to which the signal arm raises or falls away from the horizontal, to show the 'proceed' aspect ?

I had assumed an angle of 45 deg would be about right - my GWR lower quadrant drops about 65-70 deg.

The maximum should be 60 degrees according to official WR (Region not Railway) publications. But take heart as signals were sometimes way in excess of that with this type of arm.

 

I've noticed the same problem Stu and I'm not at all sure what the answer is unless the limit 'switch' inside the mechanism can be adjusted. I might try some sort of stop to see if it works without causing motoring problems but I've not yet looked inside the mechanism.

[Guest Moria]

The base is not all that big and with a bit of grass and or ballast round the edges it will be alot smaller.

 

Greetings. in N gauge the base is quite large, but I found that with 15 minutes work with a sharp 1/4" chisel to make a recess in the baseboard even that vanishes The one in the pic has the full base on it and the only mod is to pick out the base of the ladder in white so far.

 

 

Regards

 

Graham

Edited July 12, 2012 by Moria

[steam in 50s]

Yep looks good Graham, I dont think the base is going to be much trouble atall.