Jump to content
 

Servos used as point motors


Dan6470

Recommended Posts

Taking an outside perspective I can appreciate the beneficial side of MERG but I can also advise that there should be clarity on promotion of particular products (a society recommending its own products is a vested interest even if not directly remunerative).

 

 

 

Being combative and hostile won't resolve the issue and I think ScRSG's pragmatic view is good that both side shouldn't escalate the issue.

When someone makes veiled and anonymous allegations of vested interests they deserve to be challenged, robustly.

 

I don't consider my post to be combative or hostile. If you do, then RMWeb is obviously not the place for me, and you can delete my account.

 

Regards,

 

Andrew Crosland

Link to post
Share on other sites

Guest baldrick25

When someone makes veiled and anonymous allegations of vested interests they deserve to be challenged, robustly.

 

I don't consider my post to be combative or hostile. If you do, then RMWeb is obviously not the place for me, and you can delete my account.

 

Regards,

 

Andrew Crosland

AndyY is fully aware of the evidence I have, as can be seen by the tone of his reply.

Link to post
Share on other sites

With regard to Andy Y's comment about the interpretation of vested interests, does this mean that membership of any society, specialist or otherwise in this field, automatically prevents a person from making mention of their membership for fear of promoting something simply because they try to help a 'non-member' out with some information gained due to that specialist knowledge? I would guess (again) that a sizeable proportion of RMWebbers are a member of some society or other, so I really can't see how that approach would help anyone. Blatant plugging of a product for financial gain is obviously undesirable, but there must be scope for discussion, if only to raise peoples general awareness of what options are out there.

 

No, it doesn't mean that the society or its products shouldn't be discussed; it means there should be clarity.

 

 

I don't consider my post to be combative or hostile

 

And I can't see "Stop hiding behind an avatar and allow people a chance to address your issues." being anything else.

 

 

 

Link to post
Share on other sites

Dan

 

I think that since cost is clearly an issue for you a microcontroller based solution is the best idea. A 20-pin PIC like a 16F690 or similar will be able to smoothly operate eight or more servos from one chip and will only cost a couple of pounds. If you cannot afford the MERG solution (I am not promoting MERG here, it is just a freely downloadable project from a website and as far as I am aware the only freely published microcontroller based one to do this task) you will really have to homebrew your own solution. Any analogue 555 based solution is going to be far more complex and much more expensive.

 

What you are trying to do is an excellent starter project and should not be beyond the capabilities of any beginner to microcontroller projects. Here is a basic circuit diagram of all that you need:-

 

post-7495-0-32850400-1297295687_thumb.gif

 

I have only shown one servo but the other seven are connected to the PIC just the same. I have always favoured simply moving the servo from one end to the other and adjusting the linkage mechanicaly to get the throw right, and that method will keep the software simple.

 

The only tool that you will need that you will not already have is a PIC programmer, and they are not expensive.

 

The switches are all SPST, but you can just use one pole from a DPDT switch if you are using that to switch your frog polarity.

 

There is no need to be frightened of microcontrollers. It is a lot easier and cheaper to build a project based on a microcontroller than it is to use analogue electronics. It should be reasonably simple to get 250 steps between one end and the other for smooth movement, but 2000 steps is achievable - far better than any servo can track.

 

Hi Suzie,

 

Thanks for contributing. I feel that I should explain my preoccupation with cost. As a child I had a train set, mainly Tri-ang Hornby with a couple of Wrenn engines, which has been sat in a box in the loft for the last forty years. Also, last year, I received a quantity of Hornby Dublo from the 1960s, it had belonged to a long departed relative. This caused me once again to consider building a model railway. Having looked at the contents of my old train set along with the Hornby Dublo stuff, I came to the conclusion that all of the engines, rolling stock, points and signals, the whole lot had to be sold and the proceeds, almost £1300.00, used to purchase new. One of the main objectives of building the new model railway is to see how much of it can be built from the proceeds of the old. Track will be SMP and points/turnouts what ever you want to call them, will (hopefully) be hand built. The model railway that I am designing will require in excess of 70 point motors, for some reason the number of points keeps growing. From this you will realize that the purchase of off the shelve RTR point motors such as Tortoise would probably have exhausted all of the proceeds. Servos are a good, cheap, alternative. I just have to drive them. Microcontrollers can do this but is there a cheaper, reliable, alternative? The 555 timer may fit the bill, why spend a £1.20 per servo on a microprocessor solution if components costing just a few pence can drive all of the servos? This is what I want to explore.

 

Last year you contributed to a thread http://www.rmweb.co....ap-help-please/ with the following diagram;

 

post-9064-0-54401600-1297422619_thumb.jpg

If the MERG decoders and relays in the above diagram are replaced with two way switches then this would seem to be the ideal solution.

 

 

With regard to;

Any analogue 555 based solution is going to be far more complex and much more expensive.

Can you please explain why you think the 555 timer solution would be more complex and expensive?

 

 

I've come across another solution for driving the servos, it was proposed by Bertiedog almost 4 years ago http://www.rmweb.co....hp?f=88&t=49041, it proposes to strip out the electronics from each servo and to fit a couple of micro switches, that are actuated by the horns of the servo, with a diode connected across each of the micro switches for reversing the servo, see the following diagram;

 

post-9064-0-13245700-1297425465_thumb.jpg

 

With this solution, it would seem that all that is required is a slow pulse train to slowly drive the servo motors, the pulse train is disconnected when the micro switch is actuated by the horns of the servo to register correct position.

This perhaps seems the easiest solution! What do you think?

 

Regards

 

Dan

Link to post
Share on other sites

Dan

 

I think that your two 555 timers and a bank of switches will offer the cheapest solution and should work well. You will have to keep an eye on the 555 circuits to ensure that the pulses stay accurate since they will be prone to temperature drift affecting the resistors in the timing circuit and the capacitors are likely to change in value as they age since you will be using large capacitors to get the 50Hz that these circuits typically work at.

 

The microcontroller solution is still very cheap since you only need one microcontroller for eight servos (around 25p per servo), and the timing will be very accurate. You also have the advantage of being able to control the rate of the servos as well as the endpoints with suitable software in the microcontroller making it very future proof.

 

I cannot see any merit in stripping the electronics from a servo. It is a lot of work for an inferior and expensive solution by the time you have added all the microswitches. Servos are very easy to drive as they come either by analogue means from a 555 or by digital control.

 

It is possible to use one 555 timer per point and get slow rate of movement - this is the 555 solution that would be more expensive.

 

I favour the microcontroller solution since it should be the most reliable.

Link to post
Share on other sites

Guest baldrick25

Use the CMOS version of the 555 and you don't need electrolytics at all, except perhaps across the supply. A polyester 0.01Mfd is the maximum. SG90 servos will work with almost any low frequency pulse train, I've tested them with 30HZ to 100HZ and there absolutely no malfunction with the 100 plus servo's I've bought and used. What is important is the pulse width of between 1 and 2 ms , and if you operate the circuitry off a 78M05 regulator ( which will drive 10 or more servo's simulataneously ) apart from being good electronic practice for the servos, there is no drift that a Racal-Dana 9902 6 digit - 100MHZ timer/counter can detect over time or normal temperature range.

As for being more complex to add slow change circuitry , its a small capacitor and resistor, and you 'sweep' the pulse range between 1ms and 2ms ( or any inbetween points you like) rather than 'switch' it. Apart from that you might need a discharge diode if you intend to sweep the servo every couple of seconds.

Hardly rocket science.

Link to post
Share on other sites

I am sure that a 555 can be made in to a precision timer with a bit of thought, I have read long articles that discuss exactly that, and a 555 can be made to sweep slowly from one pulse width to another by smoothly adjusting the threshold voltage as you say with nothing more than a capacitor and resistor - but using a single 555 for each point is getting away from Dan's core objective of minimum cost, especially with the myriad components required to make an effective and reliable 555 solution. There is no way that I can build a suitable 555 circuit for 25p - even a traditional 555 will cost more than that let alone a 7555 (about 65p I think).

 

The 16F690 is £1.42 from Farnell. VAT, a decoupling capacitor and a socket adds a few more pennies, but since you can control eight servos from one microcontroller the cost is very low per point.

 

It is not about rocket science, but cost and complexity. With a cost of 25p per point, and a very simple circuit (see servo-8 in the above post) I commend the microcontroller solution as meeting Dan's objective.

 

 

Link to post
Share on other sites

Guest baldrick25

... let alone a 7555 (about 65p I think).

 

The 16F690 is £1.42 from Farnell. VAT, a decoupling capacitor and a socket adds a few more pennies,

 

Perhaps you should have checked the link I posted earlier in the thread 556 Link at £2.25 for 10 + 1.25postage.

Does Farnell not demand a minimum order charge of £30 or more or some high charge like £5 for postage.?

The price Of 22.5pence each is not even the cheapest I can find, its just some supplier who is fast and relaible with despatch and stocks many many hobbyist type components with a very reasonable postage charge.

There is also Link2 for 556 at 5 for a £1 - 20p each.

Link to post
Share on other sites

  • 1 month later...

I have only just found this forum theme and that is because I have recently added to my blog some detail of my point operators and have comments about various problems in using RC servos on this duty.

Having read all three pages of this particular theme, I'm sorry to say that I get a very poor impression.

Surely these pages were set up by our very kind administrative team to help railway modellers of all kinds and technical ability to progress in their chosen hobby, and here we have a modeller asking for some help with using servos as point operators. We even have these pages under the heading "Servos used as Point Operators" but to any non-techie of electronics there is very little which makes sense.

I found this theme by searching on "RC Servos" and in the results was "Servos used as Point Operators". Hurray! exactly what I'm looking for. But no! what I find is several pages of techno gubbins that is all but useless for me.

It seems to me that such inputs would be better on MERG's web pages where the majority may well benifit, while the majority reading these pages are unlikely to progress because of what they read.

For me, actually not a member of MERG, Merg have on offer a system which demands no electronic knowledge beyond being able to solder tiny things to pc board and follow an electrical diagram. Having built the board, wire it to your servos, requiring no more knowledge than wiring any other point motor, and start 'playing trains'. Now that's what I wanted but the coverage on these few pages did not give me that.

Link to post
Share on other sites

Hi Dave

 

It’s a shame that you didn’t find the thread interesting. I spent a lot of time researching the use of servos for point motors but unfortunately, I didn’t understand everything that I read. Consequently I started this thread, in the first instance to gain some clarification of how servos operate. Clearly, you will see from my first post that my understanding was wrong and I was kindly put on the right track. I also wanted to learn how to control the servo and of course as you indicated I could have used a microprocessor solution from MERG and that has been discussed in the thread. However I was very interested in alternative solutions and one that I had come across during my research was the 555 timer. I was aware that other members had used the 555 timer to control the servos but again I didn’t understand how, so a good many questions were asked.

 

With regard to

 

 

It seems to me that such inputs would be better on MERG's web pages where the majority may well benifit, while the majority reading these pages are unlikely to progress because of what they read.

 

........ I disagree. I am not a member of MERG and I didn’t think it was inappropriate to ask these questions here on RMweb. Servos and the control of servos have been discuss here before. I simply wanted clarification. Of course you are right there is an element of “techno gubbins†but isn’t that the same with everything that we don’t understand. Currently I’m trying to build my own turnouts, I’m not finding it particularly easy but I will keep at it, previously I didn’t know what the toe or the heel referred to nor did I know about crossing angles and vees but I’m asking questions and learning. Isn’t that the purpose of the forum? I’m quite happy with the responses that I received, sure the thread went off track a little bit but my subject knowledge has certainly been enhanced and I thank everybody for their contributions.

 

 

Dan

Link to post
Share on other sites

Hi :)

 

I dont know if its of any interest but I have used sevos with the MERG controllers to operate the points on Highclere, details can be found HERE and HERE. I had to be a bit choosy about what servos I used but they have been fitted for a while now and so far so good, they do everything I wanted them to and the plan is to use them for the signals too.

 

Missy :)

Link to post
Share on other sites

but to any non-techie of electronics there is very little which makes sense.

With this I do agree - though the OP obviously is no non-techie

 

I have dabbled with servos and they continue to bewilder me - I was lost very early on as the thread started going far too "techie" and seemed to be debating a separate issue - though one the OP was obviously understanding (and it is his thread).

 

We have also a thread by Bertidog on the previous RMWeb3 which I made some progress with .. but once again the project was shelved.

 

I think the real problem is that this subject very quickly overwhelms (or is overwhelmed by those who do understand the electronics and are very comfortable with it) and fails to cover the basics in the simplest of terms. Either that, or it becomes an advert platform for MERG.

Link to post
Share on other sites

With this I do agree - though the OP obviously is no non-techie

A very good point - are we expected to 'dumb-down' (for want of a better term) all answers for fear of upsetting other readers who might not understand the details discussed, even though the person asking the original question clearly would? Any answers I try to give are aimed at the level of the question asked. Part of the problem is that some of these subjects ARE technical in nature, like it or not, so answers are likely to go that way.

 

I think the real problem is that this subject very quickly overwhelms (or is overwhelmed by those who do understand the electronics and are very comfortable with it) and fails to cover the basics in the simplest of terms. Either that, or it becomes an advert platform for MERG.

Another interesting point. Taking this particular topic as an example, but taking no sides - just observing - one solution offered (the 555 timer) could be considered to be more 'electronic' than the other (microcontroller) simply because the 555 solution is purely analogue and therefore requires some calculation based on capacitors and resistors to get the timings and variability right, whereas the microcontroller method is purely numerical in nature.

It is inevitable that MERG gets mentioned in many of these discussions, simply because electronics is the specialist subject that group deals with, and many of its members contribute here one way or another. My own dilemma is that I have the inverse problem to the majority, i.e. little experience of commercial equipment but lots with MERG kit and how to adapt it, hence solutions offered tend to be based around that situation, however others might interpret the intention.

Link to post
Share on other sites

are we expected to 'dumb-down' (for want of a better term) all answers for fear of upsetting other readers who might not understand the details discussed, even though the person asking the original question clearly would?

Absolutely not.

 

But there is a serious risk of such threads being seen as the basic level from those searching from outside the forum or just arriving here from the "active topics" when in fact they are far from that level.

 

In an ideal world such "dumb" readers (myself included on this subject) would simply exit and start their own new thread on the subject.

But then we run the risk of that thread simply being seen as a duplicate and being merged by the Mods or, just as bad, duplication the same "hi-brow" level of response.

 

This can, of course, happen with any topic subject matter and I appreciate it is very difficult for those "in the know" on a subject from "bamboozling" with their response or simply wandering slightly off topic. Pitching a response at the level that the OP of any thread can understand can be a big challenge.

 

 

The MERG solutions are just one solution and should not be seen or promoted as necessarily the best or the simplest. Though we all appreciate being made aware of them, as non-members how else would we even know that they exist? The requirement of membership will put many off as the involvement in such a group implies a certain basic level (if not much greater level) of understanding.

Link to post
Share on other sites

All that most of us want is a solution that works, preferably one that doesn't need to be built, tested and require a membership to buy. As I pointed out on the other topic regarding use for signals (much more required as we have a number of solutions for points) the best unit, so far IMHO was the Embedded Controls system, followed by the only one really available just now which is the Heathcote boards. But these are not a patch on the Emmbedded Controls units. This is not a gripe at MERGS methods and I have seen examples of their boards and very good they are, but the majority of us shy away from building the things! Give us them ready made and I might join!

 

Just to repeat what I said above!

Link to post
Share on other sites

Just to repeat what I said above!

But there are ready made boards out there already, some even with DCC.

 

As far as SERVO4 is concerned all the info is freely available to all (AFAIK) on MERG's website, you just need a means of programming the firmware into the microprocessor, if you don't want to join. This is not exactly difficult once you 've got a PICKIT2 or know somebody who does. Of course join and it's done for you. Of course if you want bounce for your signals, then unless you buy ready-made then the only alternative is to join.

Link to post
Share on other sites

I have only just found this forum theme and that is because I have recently added to my blog some detail of my point operators and have comments about various problems in using RC servos on this duty.

Having read all three pages of this particular theme, I'm sorry to say that I get a very poor impression.

Surely these pages were set up by our very kind administrative team to help railway modellers of all kinds and technical ability to progress in their chosen hobby, and here we have a modeller asking for some help with using servos as point operators. We even have these pages under the heading "Servos used as Point Operators" but to any non-techie of electronics there is very little which makes sense.

I found this theme by searching on "RC Servos" and in the results was "Servos used as Point Operators". Hurray! exactly what I'm looking for. But no! what I find is several pages of techno gubbins that is all but useless for me.

It seems to me that such inputs would be better on MERG's web pages where the majority may well benifit, while the majority reading these pages are unlikely to progress because of what they read.

For me, actually not a member of MERG, Merg have on offer a system which demands no electronic knowledge beyond being able to solder tiny things to pc board and follow an electrical diagram. Having built the board, wire it to your servos, requiring no more knowledge than wiring any other point motor, and start 'playing trains'. Now that's what I wanted but the coverage on these few pages did not give me that.

Perhaps you should start a new thread of your own. There are off the shelf servo based solutions you can use which will probably be better suited to you than the minimum cost DIY based solutions that are being discussed in this thread. Most of the discussion here is about reducing cost with little regard to increased complexity which is probably why it does not suit your aims.

 

 

 

Link to post
Share on other sites

Hi Dave

 

It’s a shame that you didn’t find the thread interesting. I spent a lot of time researching the use of servos for point motors but unfortunately, I didn’t understand everything that I read. Consequently I started this thread, in the first instance to gain some clarification of how servos operate. Clearly, you will see from my first post that my understanding was wrong and I was kindly put on the right track. I also wanted to learn how to control the servo and of course as you indicated I could have used a microprocessor solution from MERG and that has been discussed in the thread. However I was very interested in alternative solutions and one that I had come across during my research was the 555 timer. I was aware that other members had used the 555 timer to control the servos but again I didn’t understand how, so a good many questions were asked.

 

With regard to

 

 

 

 

........ I disagree. I am not a member of MERG and I didn’t think it was inappropriate to ask these questions here on RMweb. Servos and the control of servos have been discuss here before. I simply wanted clarification. Of course you are right there is an element of “techno gubbins†but isn’t that the same with everything that we don’t understand. Currently I’m trying to build my own turnouts, I’m not finding it particularly easy but I will keep at it, previously I didn’t know what the toe or the heel referred to nor did I know about crossing angles and vees but I’m asking questions and learning. Isn’t that the purpose of the forum? I’m quite happy with the responses that I received, sure the thread went off track a little bit but my subject knowledge has certainly been enhanced and I thank everybody for their contributions.

 

 

Dan

 

Hello Dan,

I'm sorry to have poured cold water on your theme. Quite obviously it has done what you wanted & it is after all your theme.

My posting was perhaps made in the frustration of clicking onto the theme and not finding what I was seeking. Perhaps also I had unrealistic expectation because of its title.

In fact I had already found a solution that satisfied me. It's the Merg solution; and it cost me time which I would rather of spent scratch building locos, but the cost per point has proved to be less than Tortoise and very little more than a solenoid solution. I suppose too that the time spent on the project has increased my knowledge of the subject matter.

Good luck with your endeavors Dan'

Dave

Link to post
Share on other sites

Perhaps you should start a new thread of your own. There are off the shelf servo based solutions you can use which will probably be better suited to you than the minimum cost DIY based solutions that are being discussed in this thread. Most of the discussion here is about reducing cost with little regard to increased complexity which is probably why it does not suit your aims.

 

 

 

 

Hi Suzie,

A very succinct summing up. And quite correct!

I think I could claim to have started a thread of my own but on my blog on RMWeb rather than in one of the fora.

When I started my project of trying to apply servos to point operation I was only aware of the MERG solution which I see as assisted DIY and which I have found totally meets my needs. I would however feel happier If I could over-come some of the foibles it has presented. I have found my own solution to these but only time will tell if that solution is good.

I jealously guard my modeling time, very much biased to loco building so that even working on the layout seems to intrude on what I would prefer to be doing. Of course without a layout on which to run the locos

there is little point in building the locos!! In a similar way, I definitely limit my time on the computer and there is no way I am ever likely to do justice to the use of this or any other such fotum.

As example :- Sorry but it's time I got back in the workshop.

Bye and thanks for your interest.

Dave

Link to post
Share on other sites

  • 3 months later...

I've just read through this thread and the one referenced on the older RMWeb site after thinking servos might be good for point control. Seems lots of other people have the same idea!

 

Anyway my first instinct is to go for bertiedog's direct connection to the motor and a couple of switches with diodes solution.

 

But being a programmer I am also interested in how the servos are meant to be used with PWM control and found the following page contains a clear explanation: http://pcbheaven.com/wikipages/How_RC_Servos_Works/

 

Regards,

David Taylor.

Link to post
Share on other sites

Hi David,

 

Thanks for the link, They provide a good explanation of servo operation.

 

With regard to the servo operation by means of the bertiedog solution, I can see a major advantage here and that is that with both the 555 time and the PIC solution there is the possibility of servo twitching. This seems to be totally eliminated with bertiedogs circuit since when the servo gets to the end of its travel it disengages the microswitch which in turn disconnects the power to the servo. The servo will only operate again when the polarity is reversed.

 

I still haven't given up on the other two solutions, I just seem to have stagnated over the last months.

 

Regards

Dan

Link to post
Share on other sites

Just to add that I'm going down the route of servo motor with micro-switch latching. Cheap & cheerful to make, & most important, replicate in quantity, as I have around 30 to install. I use a small base of mdf, and proprietary micro-switches, screws/nuts etc all available easily off the net, diodes are also easy to source though I have 100's already in stock.

The psu (5v-ish) came from an old xmas decoration. (As an aside, these modern phone type chargers which appear redundant from many sources can prove to be useful psu's in railway modelling). Control switches are again easily sourced toggles though I tend to go to wholesale trade sources for cost reasons, rather than the likes of Maplins etc. I can halve the prices that way!

A false start of using 3x micro-switches (2 for limit of travel, 1x for frog polarity switching has slowed progress; a renewed effort is now underway in eliminating the frog microswitch. This makes set-up and reliability of travel limits much easier. Instead I am now using a 4-pole toggle switch for control (though at extra cost). The 2 extra poles can be used for 100% reliable frog switching, normally only 1 frog, but on a crossover (2x points) both poles will be in use.

Stewart

Link to post
Share on other sites

Hi Stewart,

 

I looked at the Tam Valley servo controller yesterday and was very impressed, but I'm still going to try cheaper first, being the direct control of the servo motor limited by microswitches. That solution just really appeals to me - far more so than using an analogue circuit to generate pulses etc... I struggle badly with analogue electronics. Each point unit should come out somewhere in the region of $4-$5 dollars including switches and diodes etc! Makes the commercial ones seem a bit pricey. Of course, I haven't got my serves and switches yet, and I may not get them to work :huh:

 

I agree with your idea re crossing polarity switching. I'm also going to drive that from the control panel switches rather than an extra microswitch drive by the servo - it just seems more obvious with less to go wrong, with the tradeoff of longer wire runs.

 

Regards,

David.

Link to post
Share on other sites

I'd also like to add that I have a good source of layout wire stashed away. Last car I scrapped was a Rover 800 (It had a serious engine failure & I replaced the car with an identical model, removing many parts for spares, and transferring the LPG kit over to the new one, then disposing of the shell to the scrappie). I also cut out much of the wiring loom in as long a length as possible, this has given me plenty of flex in different colours, ideal for modelling. Just shows what a tight old git I am.........

Stewart

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...