Megapoint Controllers

[newbryford]

I'd wager that many modellers without some electrical engineering knowledge are caught out by the wire Cross Section Area issue. One only has to look how it's sold to modellers - mainly by reference to its current carrying capacity and not it's CSA; which in fact determines to a large extent its voltage drop due to higher resistance per unit length.

 

If I remember correctly it's V_d/A/m - volt drop per amp per metre.

 

Cheers,

Mick

[leopardml2341]

It's just Ohms law

 

Voltage drop = load (A) x Length(m) x cable resistance (Ohms/km)

 

Regards

 

Edit; for DC of course.

Edited February 6, 2017 by leopardml2341

[newbryford]

It's just Ohms law

 

Voltage drop = load (A) x Length(m) x cable resistance (Ohms/km)

 

Regards

 

Our cable specs quoted volts/amp/m. We knew the amps and the distance. I can't remember what the allowed volt drop was before we upped the cable size - it's about 30 years  since I did any calculations.

Edited February 6, 2017 by newbryford

[AndyID]

We did some experimenting today on the new large group layout we are building. The (MERG) servo control boards are mounted in the control panel, so are a max. of 20ft. away from the furthest servo. Using the wiring we were used to, which was admittedly quite thin, the servos kept twitching , quite violently. So, to cut a long story short, we doubled the thickness of the wiring of the 5v supply to the servo and they stopped twitching! This suggests that there was a voltage drop on the original wiring which was cured by this simple solution. Two servos were on this line. Would adding further servos to this cause any further problems? Effectively we have created a "bus" for the 5v servo supply. Anyone care to add to this? The control boards themselves are OK, and connecting the servo directly on to the board, it works perfectly with no twitch, which seems to have confirmed the cable length idea.

 

Other than the wire gauge, is the new connection identical to the previous connection (routing, bundling, etc.)? If the new connection follows a different path that might have more to do with the different behavior than the wire gauge.

[Vecchio]

According to BS7671 the allowed voltage drop for private installations is 5% for lighting circuits and 8% for all other cicuits - but I do not believe that the servo twitching has anything to do with the voltage drop...

[ScRSG]

I'd wager that many modellers without some electrical engineering knowledge are caught out by the wire Cross Section Area issue. One only has to look how it's sold to modellers - mainly by reference to its current carrying capacity and not it's CSA; which in fact determines to a large extent its voltage drop due to higher resistance per unit length.

Absolutely correct, and as a non-electrical modeller who just tries things to make them work, this is why I posed the question. Your reply seems to confirm, by experimentation, that we may have found the solution. I kinda like these challenges, however! 

Edited February 6, 2017 by ScRSG

[leopardml2341]

Absolutely correct, and as a non-electrical modeller who just tries things to make them work, this is why I posed the question. Your reply seems to confirm, by experimentation, that we may have found the solution. I kinda like these challenges, however!

 

See also my post #89 and AndyID's note above.

[leopardml2341]

......but I do not believe that the servo twitching has anything to do with the voltage drop...

Unless that 'voltage drop' is actually a response to conducted or (less likely) coupled EMI, into the servo driver, in which case the twitch could be a response to the reinstatement of the nominal position set point?

[ScRSG]

Other than the wire gauge, is the new connection identical to the previous connection (routing, bundling, etc.)? If the new connection follows a different path that might have more to do with the different behavior than the wire gauge.

Yes, all we did, at this stage , was to utilise a second set of identical wires and "tee'd" them together, effectively doubling the thickness of the wire. Just to clarify, the twitch which we had was the servo arm "twitching" about 20 degrees! not just a small amount. Interestingly, it only did this at one setting of the control switch, when this was changed to the other state there was no servo movement after it had set itself.

[leopardml2341]

Yes, all we did, at this stage , was to utilise a second set of identical wires and "tee'd" them together, effectively doubling the thickness of the wire. Just to clarify, the twitch which we had was the servo arm "twitching" about 20 degrees! not just a small amount. Interestingly, it only did this at one setting of the control switch, when this was changed to the other state there was no servo movement after it had set itself.

That ain't 'twitching'!

 

Have you been able to reinstate the problem by removing the second wire?

 

If not, I'd suggest that the problem is elsewhere and 'fixing' it by doubling up wire runs is purely coincidental and it may recurr at any time.

 

Rule number one of fault finding, I'm afraid.

Edited February 7, 2017 by leopardml2341

[ScRSG]

That ain't 'twitching'!

 

Have you been able to reinstate the problem by removing the second wire?

 

If not, I'd suggest that the problem is elsewhere and 'fixing' it by doubling up wire runs is purely coincidental and it may recurr at any time.

 

Rule number one of fault finding, I'm afraid.

Yes, didn't think that was what everyone else regarded as twitching! And, yes, we did recreate the problem again by going back to the single wire.

 

As further experiment and to try to prove where the problem lay, we connected the servo directly to the MERG board and it operated perfectly, but over the distance the errant movement was still there until a) we attached two very thick wires directly from the board to the servo in its operating position and the movement was no longer there. This tended to tell us that the original wiring, we believed, was the problem. So, b) using the original wiring we doubled this up as noted and the problem went away with this solution. From these tests we were reasonably confident that the cabling (wire thickness etc) is at the root of our problem.

 

The MERG boards are in the control panels for the fiddle yard, but we are also using the Megapoints boards in a similar fashion for the scenic part of the layout, so lessons learned here will be carried forward to that area when track is laid.

[Suzie]

Really best to try and minimise the length of wire run to the servos and putting the servo controllers near to the servos. Think about using digital servos too to reduce the likelihood of wild movements, and if you do have a long run using a shared power supply think about decoupling the supply near the servo with a capacitor. Cheap servos use the power supply as a reference voltage to work out what position they should be in, so any variation in the supply voltage is likely to result in movement, with consequent extra draw on the supply and further supply variation. Providing a local (to the servos) 5V power supply might be a viable solution.

 

Don't be fooled by the thin wires attached to the servos in to thinking they are low power devices, they draw nearly an Amp when moving. Not much drop in 30cm of 7/02, but when you have a 5m run there and back you are going to see nearly a volt dropped with one servo moving.

[ScRSG]

, and if you do have a long run using a shared power supply think about decoupling the supply near the servo with a capacitor.

Can you please explain this further to this non electrical illiterate, what values and where do I put them in the circuit. The shared power supply is used to reduce the amount of wiring from baseboard to baseboard, i.e. we have created a "bus" for the 5v supply. Can you recommend a power supply specification? 

[Suzie]

If providing a local supply, 1A for one servo, 2A for two servos, and probably OK with 2.5A for 3 or more servos since they should not all be moving at once after the initial power up.

 

If not using local supplies I would look at putting in 6.3V 100uF capacitor between black (-) and red (+) near to the servos in the first instance, but you will need to experiment to find the best value.

 

If running a 5V power bus, think about using 24/02 or 32/02 just like a DCC track bus - but for a different reason - it is the DC voltage drop you are trying to avoid rather than AC impedance.

 

There is good reason why most servo controllers are just for a small group of around four servos - they can be local to the servos with no need for fat power wires, and little scope for interference being coupled to the signal wire, and there is no need for a high current 5V power supply.

[Izzy]

I regret to say that I have given up on trying to use servos as they stand for model railway purposes. Initially I used SG90's with MERG boards after recommendations to do so. I could not stop errant behaviour of various kinds, twitching, chatter etc. Some was at startup, but also at other times. It would often involve just one board or servo but at others could be found with all those installed irrespective of wire run distance or wire size.

 

Increasing the basic power supply to a dedicated one able to supply sufficient at startup did solve the 'where am I' twitching, ( I have read that each individual servo needs about 0.5amp to 'initialize' on startup and used an old DVD SMS power brick of 12v/3amp capacity). The current draw with multiple servos isn't generally mentioned, and if you only have a few, as most radio controlled cars/boats/planes do ( which also mainly have large current draw dedicated battery packs) then not an issue. Otherwise it can be.

 

Mention has been made of using digital servos. My experience here - and again I might just be generally unlucky - is that these are even more susceptable to outside EMF interference. I fitted one layout with these - JP 7.5g EnErG ones - which were quiet and quick until my DCC system was switched on. Then they went haywire, constantly twitching and chattering. The command station didn't even have to power up the layout for this to happen.........

 

It is possible that all these issues can be overcome given sufficient time, effort, and knowledge, but for me it's just something that isn't 'there' in terms of ease of use and reliablity under all/any conditions. So now it's hacked servos - as here:  http://www.rmweb.co.uk/community/index.php?/topic/112916-hacked-servo-point-control/

 

Izzy

 

 

 

 

[njee20]

If providing a local supply, 1A for one servo, 2A for two servos, and probably OK with 2.5A for 3 or more servos since they should not all be moving at once after the initial power up..

The power up was the issue I had. Once I got to a certain point (no pun intended), they'd all rhythmically twitch in unison when starting up. Running 60 or so servos. Beefier power supply sorted it!

[AndyID]

 

 

It is possible that all these issues can be overcome given sufficient time, effort, and knowledge, but for me it's just something that isn't 'there' in terms of ease of use and reliablity under all/any conditions. So now it's hacked servos - as here:  http://www.rmweb.co.uk/community/index.php?/topic/112916-hacked-servo-point-control/

 

Izzy

 

Some more hacking ideas here: http://www.rmweb.co.uk/community/index.php?/topic/96929-no-stall-servo-point-motor-servo-hack/

[Revolution Mike B]

I don't get any issues with our cheap servos and the Megapoints system other than some infrequent chatter, and even then, in an exhibition hall, you can't hear it any way.

 

If you use 0.6 - 0.8 wire for the control arm then it stops the servos from being overdriven too. I've never had any issues with the servos twitching either and they're all using servo cables plugged together over quite a distance.

[ScRSG]

Well, another club night last night and a further attempt to add a third point motor (out of nine at that location) to the layout. So, as a test, connected the servo directly on to the MERG board and tested, worked perfectly, smooth and quiet. Mounted the motor on the layout and plugged it in, CHAOS, the servo jumped about and then worked its way to the end of its travel (it was supposed to centre itself). Must admit to being quite dispirited by this! So, bigger power supplies on order, capacitors on order, consideration of a different wire to supply the 5v line etc. etc. Last resort will be to move the MERG boards out of the control panels to nearer the point motor locations, redo the switching and change the 5v line to 12v for powering the boards.

 

I hope the Megapoints boards installed in the main control panel work better than this, I can only hope!

[Vecchio]

 

 

It is possible that all these issues can be overcome given sufficient time, effort, and knowledge, but for me it's just something that isn't 'there' in terms of ease of use and reliablity under all/any conditions. So now it's hacked servos - as here:  http://www.rmweb.co.uk/community/index.php?/topic/112916-hacked-servo-point-control/

 

Izzy

 

I wouldn't use hacked servos - I do not see the point. In this case a tortoise or similar with some auxiliary contacts is better. The only advantage of a servo drive is the exact positioning and the variable speed. And in my opinion you will not get that with a hacked servo.

Vecchio

[Izzy]

I wouldn't use hacked servos - I do not see the point. In this case a tortoise or similar with some auxiliary contacts is better. The only advantage of a servo drive is the exact positioning and the variable speed. And in my opinion you will not get that with a hacked servo.

Vecchio

 

True, I would agree. But they are much cheaper and smaller, the latter being the main consideration of why I tried them in the first place. But you can vary hacked servo speed by altering the supply voltage and using sprung acutators means they are no different to Tortoise/Cobalts.

 

I envy those who have no problems and for who it just works because the premise is very appealing, and commiserate with those for whom it doesn't. It gets frustrating and then pointless in the end when the costs involved start mounting up and you feel like your banging your head against the proverbial brick wall.

 

Izzy

[Jol Wilkinson]

Well, another club night last night and a further attempt to add a third point motor (out of nine at that location) to the layout. So, as a test, connected the servo directly on to the MERG board and tested, worked perfectly, smooth and quiet. Mounted the motor on the layout and plugged it in, CHAOS, the servo jumped about and then worked its way to the end of its travel (it was supposed to centre itself). Must admit to being quite dispirited by this! So, bigger power supplies on order, capacitors on order, consideration of a different wire to supply the 5v line etc. etc. Last resort will be to move the MERG boards out of the control panels to nearer the point motor locations, redo the switching and change the 5v line to 12v for powering the boards.

 

I hope the Megapoints boards installed in the main control panel work better than this, I can only hope!

I have experienced problems with servos and MERG controllers so am looking at alternatives. My use is solely for signals where erratic operation is both visible and can damage the signal mechanism.

 

So far it seems DCC systems for loco control suffer less than DC. My kit built locos are all DC, and with the Pentrollers I have all perform well and I have no intention of changing. The problems I have experienced are random movement on powering up and twitching form loco generated interference (this can be from pickups as well as motors). Long servo leads are a definite problem and I have yet to try a more powerful, "stiffer" power supply.

 

The great benefit of the MERG unit is the excellent setting box, something not usually replicated elsewhere (I can't understand Tam Valley system I have to test). The "unreliability"  I have read about with most servo controllers means I want to avoid having to fiddle about underneath the baseboards to get the signals adjusted correctly. With a remote setting box it is much easier. 

 

There is also a lot of confusion over servos. Some servo controller manufacturers say the SG90 is okay, others recommend digital servos, others say that good quality more expensive analogue servos are the best value option.

 

So far the best solution for reliability and ease of use seems to be the GF unit as used on Liverpool Lime Street, but at £170 for the two units and setting box I need, I am a bit reluctant to invest until I have tried the Heathcote unit I have to hand.

[Suzie]

There are pros and cons to making the choice between digital and analogue servos.

 

1. Digital servos always move at full power, even just for a tiny readjustment of position. This will require a substantial power supply if there are a lot of simultaneous servo movement. If the controller has some jitter in the pulse length it is sending there will be a constant high power draw and the servos will noticably 'tick' while they are supposed to be stationary. Analogue servos do not move at full power for small movements so will use less power and not tick if the control signal is a bit jittery.
2. Digital servos sample and filter the applied control pulse so will not move if there are interference pulses - they will just be ignored unless they repeat a lot. If your wiring is prone to picking up interference this will make a lot of difference, and may be key if you have a delicate mechanism that cannot take wild eratic movement. They also should not move until they get a good initial pulse which in most cases of good controller design should prevent the initial start up twitch you usually see with an analogue servo. Sometimes the filtering can filter out desired very small movements, in which case an analogue servo might be a better option, but it would have to be a very good analogue servo!
3. Analogue servos use simple analogue techniques involving charging up a capacitor with a resistor to to create a DC voltage to compare with the voltage on the feedback potentiometer. Variations in supply voltage, pulse voltage, pulse shape, and glitches in the pulse, will all be reflected in how the servo electronics calculates the position. There are lots of variables that can lead to the servo not being where the controller intended it to be. Digital servos use proper signal processing to calculate the desired position based more closely on the pulse sent from the controller and pretty much eliminate all the problems seen with analogue servos not following the pulse correctly.
4. Analogue servos do not apply full power for small movements so if you have a sticky mechanism they can be easily stalled and will just sit there drawing a lot of power and not moving to the desired position. If you stall a digital servo it will be at full power and you will have something seriously wrong.

Generally speaking you will only want an analogue servo if the controller is a bit lacking in how it produces the pulses, the power supply is inadequate from a current point of view (voltage must be very stable indeed), or you need very precise positioning of less than a half of a degree - and your wiring runs are very well separated from other wiring, and very short!

Edited February 9, 2017 by Suzie

[leopardml2341]

This might be useful info ScRSG:

 

[Vecchio]

I only partially agree. Analogue servos were good enough for the last 40 years or longer for the control of model aeroplanes - and you can believe me it is more disturbing on a fast flying model than on a point or signal motor if a servo has a problem. They come in various sizes which means also power is not a problem. but even a SG90 should be fine to switch 0 scale or G scale points if there isn't something really wrong with the mechanism of the point or signal.

 

Digital servos jitter as much as analog servos do, normally the frequency is much higher which can end in an annoying hum with 50 Hz or even higher. As Suzie said the power consumption is higher, which could also be problematic when you switch the system on having some 30 or 50 points in action.

 

I am using only analogue servos on my layout - there is the odd tick from time to time but never ever any wild jitter. Of course I have a separate power supply of 10A for the servo system.