An Asymetrical double outside slip in 2mm

[StuartM]

It is always important to draw up your wiring into a diagram so that if a fault occurs sometime in the future you will save yourself a lot of time by being able to refer to a pictorial representation that you can understand.

 

This is going to be my first foray into DCC and that's reflected in the way I've wired this track, although it is currently working on DC, but when the down main gets wired then I will have to change controllers.

DCC= 2 wires my foot

[Nigelcliffe]

I am rather tempted to go down the servo route but i am not sure if it will be as simple as the tortoise ( or similar models) with the tortoise there is only a rough set up beyond that its self adjusting and can cope with temperature changes. How do the servos fare with temperature. Presumably as long as it is close to the tie bar so any linkage is short, it is not much of an issue.

 

Servos will have no issues provided some planning is done first. If you expect a servo to twitch its arm less than five degrees to move the turnout, then I expect there *will* be problems. If, instead, you use the majority of the movement in the servo, and reduce the movement by mechanical linkage design, including a light springing action, then it will be fine. The Tortoise operating wire is a spring to ensure the blades are fully home and not to apply too much pressure. The same arrangement can be made with a servo, including adjusting the effective throw by moving the pivot centre.

The other item the servo will require is a means to change crossing polarities. This can be either a microswitch attached to the mechanical movement, or it can be electrics/electronics before the servo operating mechanism.

 

- Nigel

[StuartM]

In response to Nigel's comments about servo twitch, he is absolutely right, servos do twitch from time to time,

In order to over come this with my circuit/mechanism through experimentation I've built in a number of features....

1. The servo arm is NOT physically attached to the turnout operating mechanism (tou), instead it is centred in the middle of an aperture wide enough to allow the odd twitch without causing any movement to the tou mech. The arm moves left or right and presses up against the sides of the aperture and this causes the tou to move the switch blades.

2. You will notice there is a "stop bar" on the rear of the inner slide and this stop bar limits the amount of travel in either direction

3. The servo will stay pressed up against the slide bar, thus holding the switch blades in position, should at this point the servo 'twitch', because of the stop bar it can only move the bar in one direction and therefore not over extend the limit of travel.

4. In reality I have found this works well and even if the servo arm does come off the slide bar, there is enough resistance in the mechanism to keep the switch blades in place.

5. By use of decoupling capacitors within the servo circuit and the way the code is written, twitch can all but be eliminated

6. My circuit has a built in relay to switch track polarity.

 

I guess one could buy the proprietary switching units but at £10-15 a go plus the amount of space they take up, this becomes (for me) prohibitive. My units cost me around £5 to make and I can program them to suit my needs. This works for me.

[Izzy]

It would be interesting to know if your servo circuit has the capacity to overcome both EMI issues and twiching on power-up. I recently discovered that it isn't possible to use elecro-magnets with servos, as each time they are energized the servos twitch and loose position. This also seems to arise with some high magnetic field motors i.e old design/low efficiency/high current draw types. Twitching also occurs on power-up if the power supply isn't sufficient. These discoveries have been made using MERG servo boards coupled with Towerpro SG90's.

 

An additional problem is that twitching seems to take the servo beyond the normal 87 degree range to near the maximum the servo gearing can move, which is much more, near 190 degrees. Given their gearing and power output they can therefore easily rip themselves out of any restraints that are imposed and wreck whatever they are connected to, as I have discovered to my cost.

 

I like the small size and power of servos, but considering the problems in using them that I have encountered I would now say that Tortoise/Cobalt stall action units are far easier if you have the space under the baseboard and can afford them. In 2mm I would say finding the space is the major issue, and especially with pointowork, and that is where servos score.

 

Izzy

[StuartM]

Izzy,

servo's will always twitch on startup and there is nothing you can do about it

I agree using servo's is not an easy option but I'm not making these units as a commercial concern, I'm doing it for my own interest.

And by adhering to the solutions I've outlined above the issues are overcome.

Rgds,

StuarM

[Nigelcliffe]

It would be interesting to know if your servo circuit has the capacity to overcome both EMI issues and twiching on power-up. I recently discovered that it isn't possible to use elecro-magnets with servos, as each time they are energized the servos twitch and loose position. This also seems to arise with some high magnetic field motors i.e old design/low efficiency/high current draw types. Twitching also occurs on power-up if the power supply isn't sufficient. These discoveries have been made using MERG servo boards coupled with Towerpro SG90's.

 

 

Interesting problem, I'd like to know the details of your power arrangements.

I've got some servos (cheap SG90's) on the bench hooked up to a LocoServo board. I connected a Gaugemaster (Seep) electromagnet to a 12v DC power supply, flicking it on/off with sparking at the contact as I connected/disconnected the wires, trying to make as much electromagnetic noise as I could. I couldn't get a twitch out of any servo, no matter how close I put things to either the servo or the leads connecting the servo to their power board. It is possible that the Servo4 board is unusually noise sensitive compared to the LocoServo, but both designs use similar hardware.

 

 

An additional problem is that twitching seems to take the servo beyond the normal 87 degree range to near the maximum the servo gearing can move, which is much more, near 190 degrees. Given their gearing and power output they can therefore easily rip themselves out of any restraints that are imposed and wreck whatever they are connected to, as I have discovered to my cost.

 

Any mechanism using servos has to have enough springing or over-range allowances to cope with this. Otherwise it does break things. Its not difficult to arrange with springy wire or other common methods. Motors such as Tortoise have such springs attached at manufacture (the operating wire), so the same could be copied for servo motors.

 

 

- Nigel

[Izzy]

Interesting problem, I'd like to know the details of your power arrangements.

I've got some servos (cheap SG90's) on the bench hooked up to a LocoServo board. I connected a Gaugemaster (Seep) electromagnet to a 12v DC power supply, flicking it on/off with sparking at the contact as I connected/disconnected the wires, trying to make as much electromagnetic noise as I could. I couldn't get a twitch out of any servo, no matter how close I put things to either the servo or the leads connecting the servo to their power board. It is possible that the Servo4 board is unusually noise sensitive compared to the LocoServo, but both designs use similar hardware.

 

 

- Nigel

 

 

 

I tried using both 16vac and 12vdc along with several different set-ups couldn't prevent servo twitching on use of the Seep EM's. As others have had similar issues with old motor types I have concluded that it is basically EMI problems that affect users at random dependent on each particular set-up i.e. proximity of wire runs etc, but having little knowledge of electronics I was hoping that perhaps Stuart's electronic servo arrangement might possibly offset these issues.

 

Izzy

[StuartM]

Izzy,

I found that servo twitch in my circuit was down to the way I had written the code.

Basically a servo requires a regular pulse from the microchip around every 20ms, this is supposed to happen in the background, but what that means is the chip will stop operating the program, pulse the servo and then continue operating the program and it will repeat this for the operation of the chip/program, it just happens so fast we don't notice. However if the code is written in such a way that a continuous loop is in operation with regard to an input/output pin of the chip then the timing can get a bit muddled as the chip try's to do two things at the same time, ie: operate the servo and look for an input, and this can cause a loss of pulse to the servo and when the servo receives the next pulse it 'twitches' into position. I found that by adding pulseout commands into the input loops the twitch disappeared.

I'm not an electronics expert, and rely on help from the Picaxe forum until I achieve what I'm after.

if you want to follow my learning curve re servos follow the link http://www.picaxefor...php?20883-Servo

[Izzy]

Thanks Stuart, that's very interesting.

 

I have now learned that the problems that I and others using the MERG servo boards have been experiencing is quite probably EMI issues as I suspected, and resulting from the PIC variant used, which it now seems is particularly prone to being affected by electro-magnetic fields.

 

Izzy

[StuartM]

Try housing your merg servo board in a tobacco tin or some other metal case that is earthed.

Another way to damp down emi especially when switch's are used is to use small capacitors on any long lead to/from the board, the positive leg goes to the wire and the negative leg goes to a common earth, this can help decouple the wire which will act as an ariel and pick up all that interference.

 

This photo shows another circuit I made for controlling the signalling on my layout, it had around 14 wires leaving the board going to switch's on the control panel, every time I flicked a switch the circuit would behave in the most irrational way, until I added the decoupling capacitors and placed the circuit in a tin.

[Izzy]

Ah, thanks again, seems like a fairly simple and easy way of overcoming such issues.

 

Showing my complete ignorance over the finer details of such matters, could you possibly advise what size capacitors to use for wire runs i.e. does run length have an impact on capacitor size?

 

I have built a small layout with a remote/plug-in/tethered control panel, so 30 wires, 2mtrs long, plus the wiring on the layout, 50cm runs from servo boards. I have overcome the EMI issues on this by removing the electro-magnets and replacing them with servo-operated permanent ones along the lines devised by Ian Morgan, (cheaper and can span multiple tracks at once so a better method in the long run), but the knowledge would be handy to have for the future.

 

Izzy

[StuartM]

Ah, thanks again, seems like a fairly simple and easy way of overcoming such issues.

 

Showing my complete ignorance over the finer details of such matters, could you possibly advise what size capacitors to use for wire runs i.e. does run length have an impact on capacitor size?

 

I have built a small layout with a remote/plug-in/tethered control panel, so 30 wires, 2mtrs long, plus the wiring on the layout, 50cm runs from servo boards. I have overcome the EMI issues on this by removing the electro-magnets and replacing them with servo-operated permanent ones along the lines devised by Ian Morgan, (cheaper and can span multiple tracks at once so a better method in the long run), but the knowledge would be handy to have for the future.

 

Izzy

 

 

Hi Izzy,

The ones I used in the photo are 100uf because that's what I had in my bits box at the time, but a smaller value of 22 or 10uf would do just as well.

The length of cable runs on a layout of several meters makes no difference to the size of capacitor to use.

 

Model railways with all the transformers, electromagnets and motors do make a lot of electrical noise, hence the use of suppression capacitors in locos to stop all those zzzzzzzzzzzzzzzzzzzzzzz's on your TV or radio. By shielding or decoupling your electronics you over come this.

If you're not a rivet counting purist you could try Dapol's new uncoupling system which use permanent magnets instead of electromagnets, so reducing noise further. I've started to use these and I like them.

Rgds,

Stuart

[StuartM]

I last posted on this model 2 and half years ago!

At that time, I'd worked on this model in a pretty full on way, the last six months being dedicated to getting the electronics to work.

I got bored, I also thought the track could  better, so I started work on the MK2, which I completed last year.

In the mean time, the MK1 was hung on the wall and I would stare at it often, thinking "I should finish that".

 

I started to have a clear out, too many projects, far too much stuff for my small modelling room, and eventually after weeks of ummimg & arhing, I decided that the trackwork on this model was not as good as the track I can currently produce and that the full layout would never realistically get built, so I decided to scrap the thing and placed it by the bin ready to dump, but at the very last moment I decided to complete the scenic section as a diorama only.

This photo shows the current state of play, with platforms and completed trackwork, which is waiting to be ballasted.

[Budgie]

The way you've photographed that makes me ask "is it really an outside double slip?" I ask because to me it looks like a scissors crossover.

[StuartM]

The way you've photographed that makes me ask "is it really an outside double slip?" I ask because to me it looks like a scissors crossover.

Go back and start at page one and work your way forward.

The double slip is at the other end of this photo

[Budgie]

OK. Sorry.

 

Anyway, I like your work.