Hacked Servo point control - wiring dia added

[Izzy]

Servos are quite popular for point/signal control these days and are ideal for 2mm use given the limited space often available for point motors. Having struggled over a few years to get them to work at a level I was happy with I now convert them - hacked seems to be the term used - to work on simple DC by stripping out the electronics. This has helped recover some of the outlay initially spent on servo control boards and leads etc by enabling the re-use of the servos without them.

 

These are the latest version hacked servos I have made for point control on my new Tendring circular layout.

 

 

 

 

 

 

 

 

 

 

I got the basic idea from this thread:

 

http://www.ngaugeforum.co.uk/SMFN/index.php?topic=32488.0

 

* update *  As the above link now just leads to the N gauge forum and you have to join & sign in to be able to see the details and wiring dia I have decided to add my own version below.

 

 

 

 

 

 

I though it a neat and simple design and adapted it for my particular needs, above baseboard use, using 60thou plasticard glued to the SG90 servos to locate the switches and secure the units in place. The sub-miniature switches - 1x SPDT (control panel) + 1x DPDT (on servo) - are from Expo tools as they are easy to obtain and the small size means less power is needed to move them. Piano wire actuators are fitted in the arms and used to absorb excess travel/provide adjustment and means the point blades are sprung to sit firmly against the stock rail. These can be adjusted for travel distance and spring strength by varying the wire gauge and length to suit, as the actual travel arc is fixed by the switch movement.

 

Simple to make an advantage is they work off a plain DC supply of between 4-5v and so can be powered by AA's if you want, ( 3 x alkaline AA @4.5v or 4 x rechargeable AA @ 4.8v). This gives the option of a layout powered just by batteries when using the little 9v battery controllers now available should you wish although I currently feed them 4.8v through a voltage regulator.

 

The action is quite fast but not harsh like solenoids and often I don't realise they have worked they are so quick/quiet, but lower voltages don't move the switches reliably enough. I'm content with the fast but reliable action as even at lower voltages they are not that noticeably slower but strangely do sound harsh and noisy.

 

Izzy

 

 

Edited April 3, 2022 by Izzy
add wiring dia

[Brian]

Neat hack  .  Any idea what current they draw as they operate?  Presumably the toggle switches action holds the servo and operating wire in position rather than as per normal the servo being held by the pulses from the control electronics.

Watched the video of them moving, certainly not slow and virtually the same speed as a solenoid, but at around half the price! 

Edited October 7, 2017 by Brian

[Crosland]

Presumably the toggle switches action holds the servo and operating wire in position rather than as per normal the servo being held by the pulses from the control electronics.

 

Using conventional analogue servos it's better for our applications to turn off the control pulses at the end of the movement. This causes the servo to power down and the friction in the gear chain holds the position.

 

If you leave the control pulses running and the mechanical adjustment is not spot on, the the servo will be continually pushing against the stops and will be noisy and very power hungry.

 

Digital servos cannot be turned off this way as they remember the last setting and continue driving the motor.

[66C]

Using conventional analogue servos it's better for our applications to turn off the control pulses at the end of the movement. This causes the servo to power down and the friction in the gear chain holds the position.

 

If you leave the control pulses running and the mechanical adjustment is not spot on, the the servo will be continually pushing against the stops and will be noisy and very power hungry.

 

Digital servos cannot be turned off this way as they remember the last setting and continue driving the motor.

 

The toggle switch acts as a changeover switch to reverse the direction and as a limit switch to turn off the power as it is moved over.  The servo motor only draws current while moving.  The mechanical resistance in the gearing holds the servo position at each end of its travel.  Bear in mind that in this modified form the electronics are stripped out and the servo is simply a well-geared-down DC motor.

 

I tried one out in a test rig that has been well played with by various fellow modellers and has survived OK so I think that reliability should be fine.

 

 

The photo shows my test rig operating a 00 Peco turnout mounted on a piece of 6mm ply.  Power for the test rig is from 4 AA batteries but the planned implementation operating hand-built 0 gauge turnouts will use a regulated 5 volt supply.

 

Regards.

[Izzy]

Neat hack  .  Any idea what current they draw as they operate?  Presumably the toggle switches action holds the servo and operating wire in position rather than as per normal the servo being held by the pulses from the control electronics.

Watched the video of them moving, certainly not slow and virtually the same speed as a solenoid, but at around half the price! 

 

Sorry Brian, I'm afraid I have no idea what current they draw in this format/voltage not having measured it, but past tests have shown about 0.5amp @6v and 0.9amp at stall @6v (this with SG90's).

 

You are correct about the toggle switch determining the arc range of movement, which is less than that of a normal servo. It does hold the servo in postion by default but this is secondary to it's main objective to reverse the direction of travel in combination with the other toggle switch. It is overcoming the toggle switch resistance that determines the level of voltage needed to move them which then sets the speed of movement and why the smallest switches available from Expo are used.

 

You can operate hacked servos without the toggle switch as stall motor type devices, but in this case you must keep the voltage down below about 3v (they will work on down to 1.2v depending on the actual servos used), otherwise they just rip the gears/stops to shreds given the high reduction ratios they have. IIRC the 9gm SG90's are around the 320-1 mark with four stages, while I have some JP 7.5gm EnErGy digital ones that use a smaller coreless motor and are around the 615-1 ratio with five stages. (they have a rolls royce type performance - smoother/sweeter/quieter - in comparison with the cement mixer SG90's)

 

This was my original design of hacked servo, powered via sprung centre-off SPDT toggle switches - Expo again, and steel piano wire used as the acutator, the servo moving through it's normal travel arc of around 87degrees and the wire absorbing the excess travel over that needed, much like the Tortoise/Cobalt stall motors, just smaller - much smaller, and cheaper. However this then needed a separate micro-switch for crossing/frog polarity changing. The design posted here overcomes that issue and gives a neat all-in-one package at the expense of needing a bit more power and slightly faster operation. So far they have proved totally reliable in operation.

 

Izzy

 

 

edited to correct gearbox ratios - see further post for more details

Edited October 8, 2017 by Izzy

[sp1]

How are the switches wired?

[Izzy]

How are the switches wired?

 

Please use the link in the first post to see the original post on N gauge forum by the clever chap who thought up the idea. Nice and clear and better than I could do!

 

 

Having searched out the details of the servos I have used - I knew I had them somewhere -  I thought perhaps they might prove useful/interesting for some.

 

                  Details:            JP EnErGy 7.5gm digital                     Tower Pro SG90 9gm

 

                    Motor:                 6x10mm  coreless                         10x12mm flat can 3-pole

 

                Gearing:                  5 stage    615 - 1                                4 stage   320-1

 

 

Current Draw - first is running/second is stall. A selection of power settings obtainable using both alkaline and re-chargable AA's or set via an adjustable Voltage regulator.

 

                   1.2v:                     60ma/170ma                                        60ma/320ma

 

                   1.5v:                     60ma/220ma                                        70ma/360ma

 

                   2.0v:                     60ma/270ma                                        80ma/500ma

 

                   3.0v:                     70ma/375ma                                        100ma/630ma

 

                   4.5v:                     80ma/560ma                                        120ma/970ma

 

 

You will notice that although running current draw is quite okay the stall figures get worrying when higher voltages are used. You wouldn't really want to use these as stall motor designs when fed with continuous current at any voltage, but especially higher ones. That is where this design implementation scores so well. The action of the servo throwing the DPDT switch ( only one side is used as a SPDT the other switching the crossing polarity), not only reverses the current ready for the next movement but disconnects it too, with the other SPDT switch showing the current direction set, (which of course can be mounted on a control panel).

 

 

Izzy

Edited October 8, 2017 by Izzy

[DavidLong]

Please use the link in the first post to see the original post on N gauge forum by the clever chap who thought up the idea. Nice and clear and better than I could do!

 

 

 

Bob,

 

I don't quite understand the purpose of the dashed black line on the diagram.

 

By the way, do I see that you have soldered the operating wire to a couple of pieces of brass rod (tube?) which are pushed through the operating arm?

 

David

 

Edit: I have an answer on the dashed line but one further question. Is the operating switch on-off-on or on-on and is it latched or sprung?

Edited October 9, 2017 by DavidLong

[Junctionmad]

Nice idea , adding a simple pulse width modulated power supply that could power several such stripped servos and would slow the action recreating the idea of why servos were introduced originally.

 

Ie it would slow the motor , while retaining power.

Edited October 9, 2017 by Junctionmad

[Free At Last]

Bob,

 

I don't quite understand the purpose of the dashed black line on the diagram.

 

David

It shows that the switch is operated by the motor.

[Izzy]

On 09/10/2017 at 10:47, DavidLong said:

Bob,

 

I don't quite understand the purpose of the dashed black line on the diagram.

 

By the way, do I see that you have soldered the operating wire to a couple of pieces of brass rod (tube?) which are pushed through the operating arm?

 

David

 

Edit: I have an answer on the dashed line but one further question. Is the operating switch on-off-on or on-on and is it latched or sprung?

 

Hello David,

 

Yes, small bore albion alloy tube with piano wire to provide a cantilever action since no fixed fulcrum point was involved to provide movement. The particular arrangement involved was to drive rodding above baseboard to the point tie-bars. Both toggle switches are of the SPDT changeover type (on-on) - the DPDT being treated/used as two separate SPDT switches moving in unison.

 

 

On 09/10/2017 at 11:20, Junctionmad said:

Nice idea , adding a simple pulse width modulated power supply that could power several such stripped servos and would slow the action recreating the idea of why servos were introduced originally.

 

Ie it would slow the motor , while retaining power.

 

That sounded such a good idea that I have just tried it with the units retrieved from the layout coupled to an old AMR I have spare, but I am afraid it doesn't work. Slowing the speed down looses the inertia in the multi-stage gearing which carries the switch to the limit of it's travel after the power has ceased to the motor. Hardly any movement thus occurs and mostly the switch on the servo ends up stuck in 'no mans' land between the two poles so no movement becomes possible. A real pity.

 

Izzy

Edited August 16, 2022 by Izzy

[AndyID]

Another "gutted" servo method in this thread.

 

I've found they work quite well at 1.5 volts. One benefit of removing the electronics from the servo is they are completely immune to interference. Also, as Izzy mentioned, they only consume power while they are moving.

Edited October 13, 2017 by AndyID

[DavidLong]

 

Simple to make an advantage is they work off a plain DC supply of between 4-5v and so can be powered by AA's if you want, ( 3 x alkaline AA @4.5v or 4 x rechargeable AA @ 4.8v). This gives the option of a layout powered just by batteries when using the little 9v battery controllers now available should you wish although I currently feed them 4.8v through a voltage regulator.

 

 

 

Bob,

 

If I wanted to control two servos from one switch for a crossover, would your quoted voltages still be sufficient for the servos to move the switches?

 

David

[Gordon H]

It's the current that matters in these circumstances, not the voltage.

[Izzy]

Bob,

 

If I wanted to control two servos from one switch for a crossover, would your quoted voltages still be sufficient for the servos to move the switches?

 

David

 

Hi David,

 

Yes, most certainly. The important bit would be that the power supply used has enough amperage to feed the current draw of two servos at the same time. As the rough ballpark figure for servos is to allow 0.5amp per servo that would be 4-5v @ 1amp.

 

Izzy

[Izzy]

As there are multiple ways of powering hacked servos according to particular voltage needs I thought I would just mention these one amp variable voltage regulator power supply converters I often use to power them.

 

https://www.maplin.co.uk/p/velleman-1a-power-supply-solder-kit-ve58n

 

Passing through Maplin's yesterday I see the kits are currently down to £5.49 and quite easy to assemble with fairly decent instructions. Input can be 5-24v ac / 5-35v dc and I often feed them off my 15v/3.5amp DCC system. Output via the trimmer - you need a MM to set it reasonably well - can be anywhere between 1.2v and around 22v depending of course on the imput current. They do need a heatsink on the voltage regulator for more than very low current/voltage draw and especially if it's continuous but a small offcut of thickish aluminum is all you need. I didn't need one at all in the installation shown in a previous post - you can see the VR in a couple of the shots.

 

Izzy

[Izzy]

 

 

Having recently made and fitted some more of these for a new 2FS layout I thought perhaps an update and a bit more info might not go amiss. These latest ones use Tower Pro MG90’s, which are a metal geared version of the SG90.

 

It is simple and easy to undo the base and unsolder and remove the electronics.

 

 

With metal gears and no end stops rotation through 360 degrees is possible i.e. continuous rotation if required.

 

 

A bit of PCB is glued onto the base to allow wires to connect to the motor tags.

 

Rather than plasticard they have been built using 1/16” ply. The servos are glued to the ply with cryno. The base has slots to allow for some final adjustment in conjunction with the wire acutator.

 

 

 

 

The switches are micro size from Expo tools so the toggles will fit into the arms.

 

 

 

 

You can see the amount of movement the switch allows. How much throw is used and absorbed depends on the arms fitted and the actuator connecting to the point by whatever means chosen.

 

 

These use 29swg piano wire. I use it because not only is it very springy, but it retains it strength and does not become ‘set’.

 

Power is from 4xAA rechargeable batteries @ around 4.8v – 5.2 v depending on the charge state.

 

 

At it’s simplest a DPDT on the servo and a SPDT for control is all that is required. A second DPDT can be used on the control for additional switching if desired the same as on the actual unit. On that one is used to change direction movement with the other side – they are treated as two SPDT’s – for frog polarity.

 

 

 

The yellow/brown is the battery power. Red/black is from track. Green is to frog.

 

This is the hand held control box. I made it the same basic shape as my Prodigy DCC handset.

 

 

I like these because not only are they reliable but cost is very reasonable. £2 for each servo – via ebay – and around £1 each switch (packs of 5). So £4 plus some wire, and with simple basic DC current nothing much to go wrong.

 

[Izzy]

Recently the question was posed as to whether these hacked servos are worth the time and effort involved. Well of course everybody has specific needs so view things differently but I thought that these comparision shots might help explain why I have chosen to make and use them.

 

There are three main reasons. Size, cost, and reliability.

 

Taking cost first, they work out at about, currently, £4 each to produce inc switches.

 

Reliability. I have found setting up and using commercial point motors fraught with problems of one kind or another. Work one minute, not the next. Or not having the power to actually move a point tie-bar, which has quite bemused me, thinking that's what they were meant to do. Althought you do need to make and set up these yourself, once done they never fail to work.

 

Size. In some respects this is probably the most important one for me, mostly working in 2mm/2FS where room to fit things is less than in larger scales. This is particularly so if you want to fit them directly under the tie-bar. But I have used them in both 4mm & 7mm just the same, they have more than enough power. The servos I use are all the small 7.5g/9g micro servos. Currently I would reccomend the metal geared Towerpro MG90 ones. With them there is no danger of the gears being accidently stripped/broken during construction and testing. I have bought mine off e-bay at the same basic prices as the plastic geared SG90's, just under £2 each in batches of 10.

 

While they are around the same basic size, take up about the same space, as Peco or similar solenoids, they are much smaller than even the smallest stall motor type, which many prefer for the less harsh action.

 

Here's a few shots comparing my latest installs, using 7.5g ones, against an original type Colbalt I bought to test out. I never got any more.

 

 

 

 

 

 

* I have now added the basic wiring dia to the original post as the links provided are subject to change and the information might not be accessable.*

 

 

Edited April 1, 2022 by Izzy