bertiedog

There's a bit about Rossi in the article about Polks and the efforts to get Rossi to change from toy making to scale locos especially for the US hobby market, which influence also brought about the move to HO European models after 1950. Early course scale Rivarossi branded items are very rare in the UK, the early US HO items are far more common. No imports were allowed from Italy till 1951 anyway, same as Germany, due to the war, but often items like these were available to US troops serving in the UK, and got sold on. Early Japanese models also turned up here via US servicemen, at a time when nobody would touch Japanese goods on principle, let alone an import ban. The first Japanese made items in the UK in the 1950's were cheap cameras and even cheaper tin plate toys, which hardly gave an inkling of what was to come from the Far East........ Stephen.

I trust you noticed I did not mention DCC, where the back EMF of coreless is dealt with by the Chip Circuit, (on suitable chips). I was commenting on a PWM circuit delivering 12 volts PWM, a very different beast, which without clamping the back EMF, causes the arcing, and heating, that harms the commutators on coreless types. Now.... a couple of more general points, back to the circuits being discussed, In an ordinary motor the iron core damps the effect as it operates, and coreless means that, no iron core, so adding the bigger and finer coils they are efficient generators as well. The lack of a core also means heat builds up faster in coreless under high load, and PWM DC makes this worse. PWM DC will work at high frequency with coreless if some protection is added to the circuit, and even at low frequency if fully clamped to remove the back EMF. As mentioned before the Pentroller takes advantage of the back EMF, turning it from a problem into a feedback signal to modify the speed against load. DCC can do this as well, and requires the Cv to be matched to the motor in the same way the Pentroller requires it to be used with matched motor types. The general purpose circuits here for DC do not need such a sophisticated approach, but also fall short of the Pentroller or the DCC control on offer from the model trade. Bear in mind that one chip can cost more than a complete home produced DC controller, and cannot match a computer DCC system costing many hundreds of pounds, which many people cannot possibly afford. Personally I am not anti DCC I have a home made NMRA compatible controller that does the lot for the few chipped locos I have. Stephen

Useful collection, "The Throttle" looks interesting, as everything is adjustable. The ones based on the IC regulators are good, but up rate the regulators!, not worth saving a few pence. The Scaler is interesting as it adds onto other controllers, scaling can be applied to most designs easily, switch in or out, it just spreads out the area the controller acts over. The only nuisance of these US based circuits are the relays, they specify "Radio Shack" and these types are not easy to source here, you may have to find substitutes if you want the hand held units, although a cased unit can use direct switching reverse instead. Stephen.

Yes, the timing of the pulses is down to the 555's, the bursts would have ripple, but it does not matter with motor control. In effect the 5000uf isolates the pulse circuit from the motor supply effects.

As I said all PWM can make noise, and it depends on the mechanism, a stiffer, tighter tolerance motor will be all right, but a Lima pancake will rattle like loose teeth. Coreless motors should work, but the coils generate a lot of back emf at a higher than applied voltage and can have arcing troubles on PWM across the brushes. Better motors have potted windings, coated with resin, and these are OK, but cheaper motors do not, and the looser windings vibrate just like a loudspeaker. Can motors are generally quieter on PWM signals, the casing quietens it, whilst open frame are worse, and flat pancake the very worse. With the simple 555 circuit, you can dampen the action with caps on the output transistors, or the feed to the base, this will distort the square wave but reduce the noise, but also lessen the control at crawl speeds. Stephen.

Quick check, it is the power during each cycle for the 555 timers, without C2 they would shut down. The pictures separates the simple PWM parts from the output transistors which varies the fixed max to zero each cycle as defined by the 555 oscillators. Stephen

I would have to check the whole circuit's action , but as the power transistor takes the applied voltage from supply to near zero on each cycle, the lowering of voltage would make the 555's inoperative, but with the 5000uf it effectively un-couples the oscillators from the effects of the PWM output, as they can draw power from C2 on each cycle. The top of each variable width square wave would have a ripple from the rectification, but it has no detrimental effect in practice, a perfect square wave is not needed, just the mean result. Stephen.

The copy of the circuit is from the book, I know I re-drew it several times, and I do remember a modification, but I have not got the notes either my own or from the Bambini book. I am pretty sure it worked though from the book, I used one for ages. Anyway I am not using it again, the IC regulated PWM one from Shortliner's reference is far superior, especially as it is pretty fool proof on shorts etc., the regulator will shut down. Stephen.

The circuit from WW has not been published here so far, so I think that you are referring to the dual 555 PWM circuit, (totally un-related to the WW one). The 5000uf is simply smoothing in there, it is high, but decouples any ripples as far as I can see. The circuit was by Penfold in Bambini books. I have used the circuit several times and it works very well.

And a quick point...... often over looked .....is the type that progressively reduce the Pulses as the DC level increase only do this in the upper range by definition, ......do you always run your locos at top speeds???? If you have a vast layout with long drags, then a plain DC controller is frankly much better, no complexity at all, apart from possibly feedback. Maybe the very best would have all these modes switchable !!!! but the most useful for most layouts is crawler speeds, (without too much noise). Stephen.

The WW version makes the reference to the patent type unnecessary, as the patent one has curiosities in it , and for actual ease of building a proven published one will be better, even an older design, all components are available, or better alternatives made to substitute, like power transistors that used to cost pounds and now cost pence!! For instance virtually any older circuit with the 2N3055 as output can have TWO in parallel to halve the heat, the plastic versions are so cheap now!! Stephen.

It sounds like the WW controller was PWM, with an increasing level of DC as the speed increases, this is the typical way they operate. This type uses the load to assess the current supplied, just like a regulator chip. Only a full PWM with dormant parts of the waveform can check the back EMF to use as feedback control, although I have seen circuits set to function with a core DC level as well. This is the PWM circuit based on 555 timers, not as sophisticated as the one Shortliner introduced, no momentum setting, although it could be added.

It pulls about ten US cars as "standard", (Central Valley/Kadee trucks), but can't pull my test house brick on delrin trucks, too heavy, but with about 300grms added it will pull it. But it is relying on the traction tyres to do this, and they would wear quickly if used with such heavy trains. The brick is equivalent to about 30 boxcars, but without the friction of the bogies. The gears would be all right, but the top worm would wear the first gear as the pivot does not turn the worm gear, relying on slack. It would suit light use, it matches the 44 tonner type in size, and pulls about the same as Bachmann's model with smaller dual motors. The later 44 tonner works better, the early suffered from split gears, from the usual moulding standards being too low.

Well, the Tri-ang oil did not sell at £19....and can't find the darn box anyway, it is a shop tray of them, about 20 or so, came from a shop closing down......have to search harder. Stephen.

I have got a shipping box of these bottles of oil upstairs, Ebay here they come if they sell for this!!

The chassis with sideframes, push on, the pickups are split frame, and all wheel pickup, but to work well it needs weight added. The traction tyre is on the inner pair of the power car. The design is a little odd, the gear remains fixed as the bogie pivots, so the worm contact is slack by design. BE WARNED, there is a very tiny plastic drive shaft between the the motor and gearbox, easily lost if the motor is removed. It is not possible to have a dual drive with the original motor, there is no shaft at other end as you can see in shots. To dual drive the chassis, it would need a second unit, and a dual shaft Mashima to make it work. The gears are Nylon, with a brass worm. They seem accurate and are not forced on to steel axles, so should not split. The shot shows the drive shaft, I don't think it worth dropping the traction tyre, it works fine with weight. The fitted motor is quiet and powerful, a three pole standard design. The Chassis is DCC ready. It is quite quiet and smooth, and powerful when weight is added above the power bogie. The whole design is totally different to the UE inspection chassis, every detail is different. Stephen.

The Stewart Hine Pentroller mentioned is a full back EMF sensing unit with pulses at a higher frequency, mainly developed for small 2mm and finescale 4mm models, and it not really suited to older high current motors. It can work easily with coreless types. Another published controller for home build in the Basin Street design via MERG in PDF form, free to everybody, a pulse and variable output up to pure DC that can handle higher ampages easily. After all of this one should bear in mind once moving, little difference will be detected between each type, or a battery!! It is the start and low speed that makes the difference, and more down to the quality of the mechanism than the controller.

It action is similar to commercial designs like ECM , the feedback is current sensing, not back EMF type, I would not say it is advanced over other circuits these days. A chip regulator could be driven like this, and there are variants that use discrete components, but it depends a lot on exact uses and I would not bother with the complexity of the patent spec version. Decent running is far more to do with the loco, not the controller, it is not a cure for a bad mechanism. PS I do seem to remember the WW version, but no copy of that date to hand, so will need the text etc. I think the Codar may have used the design. Stephen.

I am interested in the Wireless World and will arrange a copy, without seeing it I would assume PWM in base, as no other way gives such control. Feedback can help slow speed, but most aim at constant speed for gradients, Codar used to make one, and I have the circuit stored, but it is not wonderful on crawl. It was an early user of momentum effects. It would be fine for larger layouts. All feedback controllers are lost with multiple motors, banking, US units in tandem or more, the intermittent contact that must occur despite all efforts to keep everything spotless and all wheel contact, will lead to lurches as the tiniest feedback signal is used to correct what has not happened. Now I hear the man at the back saying he uses feedback with multiple motors, and in real life the makers of the units add hysteresis to the circuit, sudden changes do not cause sudden jumps, but some locos may still be sensitive and react, and the choice of the lag is critical, too much damping, and you might as well have no feedback. Personally I do not like feedback where there is a large amount of different locos to run, as they all behave differently, only slight, but it is there......and my feeling is that you are the driver, and should be in control on gradients anyway!!....... To get the best of both ways, do not modify one type or the other, just have both and switch in as needed, A regulator based 0 to 12vdc, with pure DC(handles anything in sound order)., a PWM for shunting, and maybe a feedback type for mainline loop running on a big layout.................or buy DCC...........trouble is the word BUY...............DCC is very pricey indeed, as all three circuits coud be built for about the cost of a single Lenz decoder. Also as before, join MERG it is the source of a lot of help, and masses of experience, although prepare to have your arm twisted towards the dark side, DCC.........there are those in there who still resist it though.

Unfortunately as far as increasing the frequency of the PWM type, you can't do much, as above the lower frequencies of about 100 hz it becomes very audible, and would then need to jump to 20 khz to work without the sound generated. PWM would work at 20khz, but only to adjust the range of speed around a value, control from zero to max impossible because of hysteresis in the magnetic field in the motor. It works with specialist coreless motors quite well for laboratory uses etc, and control equipment, but not models. Stephen.

I have breadboarded the design this afternoon, and it all works fine, although the diagram shows the transistor as normal, it is a Darlington pair type, and any rated above the specified would do, it is the gain they are after here. Also I used the higher rated regulator, why down rate for a few pence. No fuse is needed, it's self regulated, but I would add LEDs to show AC power in and DC out. It works with 5 pole motors very well, it does make a three pole a touch noisy, but very controllable. For coreless near useless, it could be tamed, but why bother there are better plain DC controllers for these. I will transfer to a PC pad board and re-test, perfect for the new Shortline I am building.

Have you the circuit to post here, bearing in mind the authors copyright, or it could be posted to Photobucket etc., and referenced. Stephen.

A Full PWM type circuit, (Constant voltage), very interesting, I use Mr Penfolds version, (in Bambini books on the 555) with 555's generating the pulses, resulting in exactly the same square wave forms. This type works very well indeed for control. especially shunting, but as the other postings on the US design for variable voltage with pulse say, it can over heat the motors, but only under load on long runs on the main lines, light load, and short runs no issue. Crawl control cannot be bettered with PWM, no way any other circuit can be superior, but only at these low speeds. Such PWM controllers can make some motors hum and noisy, again it varies, well set up motors do not mind, but add loose bearings, and dry gears, and you have noise, add tight gears, resilient mounts, and grease lube, no issue. My own mods to the 555 driven type of PWM are to raise the frequency and use capacitors to smooth the output, which are switchable to suit different motors, this also reduces potential motor heating as the RMS value of the output is lowered. Motor heating is not serious on PWM as such, it is after all nothing to do with applied voltage, it is the average current drawn under load, less efficiency, that is the heat source, not the higher voltage pulses, as nothing is applied at all in the gaps!! But yes, under load the magnetic field behaves dynamically, and despite all theory, the motors heats up more on square wave pulses, than plain DC. The motor operates on the average of the pulses, but the mean value is not linear as the load increases and more inefficiency occurs...result heat. The back EMF is also high with pulses, and contribute to the heating, as with coreless motors, which should not be used on PWM, they can burn out in extreme load conditions,. Coreless are often driven by PWM signals in electronic equipment, but are in constant known load limits, and are quite safe then. Therefore PWM CV type controller...perfect for micro layouts and switching layouts...not for large club layouts with legions of locos of different types. Stephen.

The circuit is fairly straightforward, although for non electronics builders a bit short of some basics like the power connections to it! ..you have to assume a lot, .......also the UK 50 hz mains frequency difference from 60hz in the States may explain the slower responses to the commands, meaning lower some of the trimming values ,as you found, to compensate. All parts are available easily and not costly, although the reversing relay adds a bit. Substitute transistors,(properly chosen), would work, nothing is too critical. It would be a very good general purpose circuit, and removes most risk from pulses noise and potential heating effects. Stephen.

There are lots of suitable designs on the net, but as you say how do you know what they really do? I know, as an electrical engineer, but even then the circuits may not work as expected, so it needs recommendation, and the best course is to join the MERG group, who are a club, shop, group of enthusiasts etc, and have the experience to say what works and what does not. Your requirement is for control of older "type" motors, none the worst for that, straight DC, with feedback may be the best, with variable pulse available in a switchable form. It would be the same for an all Mashima fleet, they just have lower draw on current. Feedback works where a fleet are basically the same motors, otherwise it has got to be in some way adjustable. Double motored locos cause problems with feedback and don't work with the set-up. The best control for low speeds are variable mark space CV (constant voltage) controllers, but if the motors are 3 pole and 5 pole on the same layout there may be differences, and such controllers are able to make the motors hum due to the pulses. The next best are feedback controllers, which can be CV mark space types or pure DC, or a mixture, a circuit that senses the load and back EMF, trying to maintain a constant current drain, again when pulses are used the motor may hum at low speeds. But after all the extras are removed a pure DC supply is able to run any type of motor without hum, and straight forward variable voltage regulator chips, which have built in overload protection are the best all rounder. Commercial designs try to deal with as many variables as possible, so building your own can fine tune to what you have much better. It is also very inexpensive,... £10 would build a comprehensive spec., controller, with the exception of the transformer for power, the most expensive item. hope this helps, Stephen.