Interested in a signal building group?

[Dave Noble]

Keith,

Yes, they are the ones with the strands wound round the main wire in turn, on the operating wires. I know how they are done in full size, but the wire we have is very springy stainless, only 0.5mm diameter, so unlike full size, it doesn't stay where it's put. And, the individual strands are tiny. But, there is plenty to practice with.

Testing is a good idea. I've been in touch with Jonny today who assures me that all is well, and various parts have been made to talk to each other, despite me trying to suggest short cuts. As I simply don't understand electronics, I leave it in his capable hands, and hope that if he and Richard would like some help, they will ask.

Dave.

[Keith1500]

I feared you might say the wire used is springy stainless steel. I am happy to try and devise a technique if you want to send me some bits.

The other thing I can do is post a drawing and some info on the method used for full size , which might be useful for anyone else willing to try.
I'll try and do this when I am back in the office tomorrow.

[Richard Phillips]

Here is a prototype setup (small images below) - it's got three servos attached to prove we can drive more than one arm. I did actually solder it to veroboard as I think a custom PCB is an unnecessary expense. I am re-thinking the veroboard approach though, just a small one to hold the voltage regulator, although I have some concerns about heat, we've barely got enough depth to use a small heatsink.

I'll make up a daughtboard to hold the servo connectors and lamp power take-offs. Something that will be quick an easy to knock up for each signal.

Dave, ideally if you can get me a base, some servos and the 7pin DIN plugs I can check this lot all fits.

IMG_2703.JPG (37.2 KiB) Viewed 24232 times

IMG_2703.JPG (37.2 KiB) Viewed 24232 times

I realise this is only one board, I have tested with two! I have a large video (87 MB) if anyone wants to see it "in action". I still have some ideas to try like sending a degree setting rather than an on/off instruction so we can adjust the travel without re-flashing the boards,

[Keith1500]

Can you not use the base to act as the heat sink? I think it's the same as 0 volt if using one on the 78 or 79 series. Just try to minimise the amount of work the regulator has to do, this will keep the heat down.

I would not mind seeing the video but it's way over 10 meg not sure it would email? You could try.

[Richard Phillips]

Can you not use the base to act as the heat sink? I think it's the same as 0 volt if using one on the 78 or 79 series. Just try to minimise the amount of work the regulator has to do, this will keep the heat down.

I would not mind seeing the video but it's way over 10 meg not sure it would email? You could try.

Just been on the 'phone to Ben about this, if the base bottom lid(?) was metal we could well do this and mount the electronics on it so there were no screws sticking out of the top - and I guess we'd be okay on a concrete floor - not sure about the tabletop at home after running continuously - I do wonder if we can trust the voltage regulator on the arduinos to handle it - but without soaking these things on continuous test for 3 days - I can't really say. I was wondering about cable and power draw too - CAT 5 network cable (which would seem cheap and plentiful) will handle 0.577 Amps and 125 Volts DC - what our combined power draw is I don't know yet! On the day we only need to run 4 cores. Jonny sent me a 7805 in a T0220 package, seems fairly standard.

The video size is beyond most email servers rejection for email body size - I'd bung it on a webserver if you really want to see it, but what format?

Dave (if you're reading this) I could do with a shackle or two if you have one for my "test" ground disc signal, it's all a bit heath Robinson at the moment. Any idea what angles we'll want to drive the servo/arm through?

[Keith1500]

If you are worried about table top operation is scorching due to the heat then simply mount on a nice polished wooden base that can be removed.

Don't forget the heat from the 7805 will be due to a) the current drawn by the 5volt circuits connected to it and b) the difference in voltage, that is the supply voltage. So for example it will be better to run from a 12 volt supply than 18 volts this would be 1/3rd less heat I think.

However, you might be up against volt drop in the CAT5 cable. Don't forget the volt drop is the cables resistance x the current it is carrying, and resistance increases with length. 93ohms per 1000m for CAT5 . So say for 100m there will be 9.3 ohms if the cable was carrying 0.5amps the volt drop would be 0.5 x 9.3= 4.65 volts. Which at 12volts would be ok for the 7805 (5v circuits) but not for anything intending to use 12v direct. Also as the current varies so does the volt drop so again anything using the 12v circuit might become dimmer or sluggish as the signal arm moves. Does that makes sense? Try it with a realistic length of cat5.

7805 are quite robust but they do get hot. For a short soak test just load it up and leave it for a few hours. You'll soon know the temperature. Mind you don't burn your fingers !

[Dave Noble]

Hi all, to comment on a few comments.

"Bearing in mind time constraints the easiest thing is to have the lights come on when the signal is powered". Can't see a problem with that.

Arduino Micro. If you're sure that's what we need, I can get them, just let me know, but, the Farnell site points out specifically if using more than 12V the voltage regulator may overheat and damage the board. Not sure how this ties in for your requirement for more umph, a 12 volt battery should already be giving more than 12 volts, and I am expecting the whole thing to work off batteries, particularly for home use.

Richard, I believe you were wondering about space available in the base, there is a dimensioned drawing in a previous post, page 6.

Heat sinks, is it a good idea to mount parts on lids, like a steel bottom cover, always seems to me like a recipe for damaged joints. There should be a fairly large block of lead in each base, that could be cast with a little pocket in it to take the regulator perhaps? Acrylic for the bases is with Chris for laser cutting.

The motor I used last year in the catch point was rated to take 70ma at maximum efficiency. I don't think it's up to reliably moving two blades, so may need a larger motor, 944D, which draws 200ma at max eff. Pure guesswork, but might it take two or three times that when working hard, say half an amp? Two motors per crossover, so do we need at least 1 amp (at 12 volts) to drive point motors, plus electronics? Putting this into your volts drop equasion Keith, would make me wonder if CAT5 is up to it. The DIN sockets are rated at 34v, 4amps.

[Richard Phillips]

Hi all, to comment on a few comments.

Arduino Micro. If you're sure that's what we need, I can get them, just let me know, but, the Farnell site points out specifically if using more than 12V the voltage regulator may overheat and damage the board. Not sure how this ties in for your requirement for more umph, a 12 volt battery should already be giving more than 12 volts, and I am expecting the whole thing to work off batteries, particularly for home use.

I was using a 9Volt battery in that case, so 12v will likely be okay, I just didn't have a 12 volt to hand. Operating voltage of arduinos are 7 to 12 volts, but with 7 not giving a guaranteed 5v logic from the board, so 9 to 12 being the ideal range.

The Micro certainly works, but seems quite pricey as there's no cheap copies available yet. I am having thoughts about this however as the area in the base seems adequate to fit the obsoleted (and plenty of cheap copies/"compatible" boards available) Leonardo module which is working as a master and will likely be fine as a slave. Depth being the aforementioned premium dimension. I'll see if I can adapt the program.

Richard, I believe you were wondering about space available in the base, there is a dimensioned drawing in a previous post, page 6.

Yes, just easier when you've got all the physical bits in hand, should fit

Heat sinks, is it a good idea to mount parts on lids, like a steel bottom cover, always seems to me like a recipe for damaged joints. There should be a fairly large block of lead in each base, that could be cast with a little pocket in it to take the regulator perhaps? Acrylic for the bases is with Chris for laser cutting.

...umm, pass, since we have a block of lead it's not a bad idea, otherwise the regulator is going to be up against some form of plastic.

The motor I used last year in the catch point was rated to take 70ma at maximum efficiency. I don't think it's up to reliably moving two blades, so may need a larger motor, 944D, which draws 200ma at max eff. Pure guesswork, but might it take two or three times that when working hard, say half an amp? Two motors per crossover, so do we need at least 1 amp (at 12 volts) to drive point motors, plus electronics? Putting this into your volts drop equasion Keith, would make me wonder if CAT5 is up to it. The DIN sockets are rated at 34v, 4amps.

Thicker power cable than what I suggested will be required, I was just thinking of the signals.

[steamysteve]

Dave,

Another 26 x 'U' bolts for the 15mm posts are on their way to you in the post. That should satisfy immediate demands plus some spares.

What about 'U' bolts for the 1/2" posts - are any needed or has someone else made some?

[Richard Phillips]

As Keith asked, here is a link to a large video (about 84 meg) converted to MPEG 4 (no sound) from my phone
http://people.bath.ac.uk/ccsrip/gl5/modbustest.mp4

It shows only one slave being controlled by a two-wire bus, commanding three servos independently or together. The slave is powered via a 9v power supply, the master is plugged via a USB into the computer. The voltage regulator is stepping the 9 volts down to 5 volts power the servos, the modbus and the signal lamp. You can see the signal lamp dim at the end when all three servos are moved at once. The principle is the same with more slaves (ie signal posts). Just add levers (ahem, switches)...

[Keith1500]

I think I need to better understand the system has anyone done a schematic. Is the power problem down to having a central battery? If so could this be more than 12 volts, say 24 volts. This would then half the current required which will lessen the effect of volt drop but it means the voltage regulators will work a bit harder. It also means a 7812 12 volt regulator would be required for the arduinos and the local 5v circuit. So, where to house this? Well how about a line side cabinet as used for track circuits or the like?

[Ben Lyons]

Hi regard bases and heat have today sent Richard a PVC base to install and thus test out . PVC will take some 80C with out to much trouble. If base had metal bottom would this act as heat transfer sheet? Dave are there parts to drill machine ect have at least one volunteer for the 16th at wotton may 2/3

[Dave Noble]

Trial etch for the arms arrived today.

P1030244, red.jpg (72.7 KiB) Viewed 24122 times

Folding is a bit tricky as the flange is so small. I'm happy with them so will send for another etch which should be all we need. Anyone got a little folder they would like to use to do these on? Anyone got any good ideas for painting these neatly?

P1030246, red.jpg (88.03 KiB) Viewed 24122 times

[Ben Lyons]

Hi Dave, send them to me I will sort them, if you can get them for next Saturday I will get the team me and ?John at present on them, Unless any one else want to do them.
Ben

[Richard Phillips]

The paint spec for the arms was quite thick, so as a suggestion (and I don't have access to such!) would be to powder coat? I guess it would have to be a white base coat and then the stripes would have to be stencilled on somehow..and the red/yellow fronts painted.... spraying it all would probably be easier!

Anyway, what time is the Wootton meeting? I'd like to get something demo-able from me if I can make it across - lots of wiring bits have been arriving by post, I still don't have a full set, but I'll lash something together so I can demo a multi-post multi-arm setup.

[Dave Noble]

Received some castings last week. They have been 3D printed then lost wax cast directly in bronze. One set to size for trial, one set slightly oversize as patterns. First ones to try were the lamp bracket saddles.

Lamp bracket reduced.jpg (54.59 KiB) Viewed 24051 times

All seems fine, so the other ones have gone off to the casters for copies. The lamp bracket itself is one which Ben had done a while ago, another one for a pattern should be with us next week.

THE LIST, with more details of parts, has been moved to the first post on this thread so that it is easier to find. I will try and update it regularly.

[Keith1500]

I like the look of those brackets. I have to say these signal parts are looking really good. Any more signals required let me know and I'll happy build one.

I received a length of signal wire yesterday. I have an idea for replicating the thimble termination and will report accordingly. With a bit of luck I hope to bring the termination along on Saturday to Wooton. I aim to be there for about 09:00.

Keith.

[Dave Noble]

HELP. I'm having a bit of trouble with the miniature spectacle plate as I don't seem to have a drawing. I've drawn it out best I can, using the main spectacle plate and LMS drawing as guides. I just can't get the curves quite right to properly clear the top of the arm, but I think we'll manage.

Min spec.jpg (23.85 KiB) Viewed 24009 times

The only decent photo I've found of one on the net shows it made from a flat plate with flange from the top of the arm, round the top and part way down the back. The On spectacle hole seems to have small flange round it. Does anyone know if this was normal, or is this a home made job. If it was normal, then a laser cut one would be much simpler.

Miniature arm.jpg (65.3 KiB) Viewed 24009 times

They do usually have weights added. Oddly, the hole in the screen for the On aspect (when fitted) seems to vary in position. I'm confused.

[Richard Phillips]

May not be much help but Dorman long list the following:
http://www.unipartdorman.co.uk/assets/s ... signal.pdf

BRS-SM 28/4
BRS-SM 28/5
BRS-SM 28/6

So dimensions are given in http://www.unipartdorman.co.uk/assets/086_009029.pdf and http://www.unipartdorman.co.uk/assets/086_009031.pdf for the circular lens.

[Dave Noble]

Thanks Richard, we have details of the glasses, it's the big steel plate I'm struggling with. It looks like Dorman only do the replacement lamps, or have I missed something?

Hopefully you received some bits in the post from me last week.

Dave.