Feedback Controllers

[37114]

Hi Guys,

 

I am contemplating getting a Feedback controller having seen a friends one in operation. I currently have a Gaugemaster 100 series, which is in full working order, while my track is peco code 100 streamline with insulfrog points. As my layout is a small shunting layout slow speed control is pretty crucial, but at present it is not to how I would like it. My locos are all Bachmann diesels except some Lima DMU's and one powered by a black beetle.

 

Last night I tried my friends gaugemaster HH feedback controller in the layout and the difference was unbelievable, but having looked at the gaugemaster website, they caveat that some DCC ready and portescap motors are not suitable for Feedback controller. The loco we tried did seem much louder in operation and worked fine but I am worried about longer term damage if I was to buy one. To put it into perspective my layout at the moment only gets used at shows so circa 8 days a year and each loco probably only works for a total 2 hours over a whole weekend.

 

Has anybody any experience/thoughts they could share please?

 

Thanks/

[Pete the Elaner]

Please take caution with feedback.

I have heard from several sources that these are not really suited to modern mechanisms & have been known to burn out newer Bachmann motors.

That is only hearsay but when I tried a Bachmann loco with a feedback controller for myself, I did not think it was any smoother than with a 'smooth' DC 'troller.

[bike2steam]

Meant for constant speed rather than smooth, feedback is ideal for layouts run a low speeds, the problems arise at fast speeds where the back emf can create more heat within the motor. I used one (HH) on 'Bembridge' for 15 years at some 60 + shows with no problems what so ever. Have a word with Rob Kinsey of the Wolverhampton MRC, I regard him as the most knowledgable person on model railway electrics/electronics, as far as he is concerned there are many unwarranted scare stories regarding feedback, used wisely - no problem.

[Theakerr]

I love mine. Had it many years with no problems. I do take care on which loco's I use it one based on the popular story that it is bad for certain motors, but on my open frame Motors, older Bachman and Mainline it is excellent. With my mainline J72 and my Bachmann J39 I can set them to a crawl so that you have to look very hard to see if they are moving.

[Jim104]

I love them as well, I have an extensive OO scale DC layout and have four HH controllers. Have never had any overheating problems with Hornby/Bachmann etc BUT many Bachmann locos do not like the HH controllers, the tend to surge rather than run smoothly. However, to fix this I reduced the input voltage from 16V AC down to 10 V AC. Yes the overall top speed is reduced on all models, but no trains run that fast anyway.

In reducing the input voltage all locos run smoothly and consistently on the layout. I have even tried a portescap loco (against advice) and that has run without burning out as well.

Big thumbs ups for Feedback HH controllers from me.

[Donw]

With coreless motors such as Portescap Feedback can give some poor results. The motor tries to follow the controller which is reacting to the motor. In 0 gauge I found that with quality mechanisms an emitter follower type gives a smoother result and sensible low speeds. Feedback worked well with commercial locos where the quality of the mechanism was not as good.

The concern about it damaging Portescap is due to the quoted brush life as being so many cycles. Variable pulse controllers based on 50HZ acts like on off cycles. I don't know of any proven cases but that was the concern.

If by DCC ready you mean it has provision for fitting a chip that should be unaffected. If however you mean a DCC fitted loco with a chip that will recognise DC and respond accordingly. Such a chip could fail to reconise a pulsed output as DC with unknown results.

If you can find a controller where the feedback can be switched on or off that might suit. I beleive the pentroller is no longer available that used feedback but the output was stepped DC with only a brief emf sampling period. It was designed to work well with coreless motors and others.

Regards Don

[AMJ]

Until I went DCC in the shed area and also the goods yard I used a HH model controller so essentially it was start stop action. Very good for many locos and just like DCC need to keep track clean.

[37114]

Meant for constant speed rather than smooth, feedback is ideal for layouts run a low speeds, the problems arise at fast speeds where the back emf can create more heat within the motor. I used one (HH) on 'Bembridge' for 15 years at some 60 + shows with no problems what so ever. Have a word with Rob Kinsey of the Wolverhampton MRC, I regard him as the most knowledgable person on model railway electrics/electronics, as far as he is concerned there are many unwarranted scare stories regarding feedback, used wisely - no problem.

 

Thanks for the responses guys, I have a HH on order. When I tried my friends HH the Bachmann locos ran fine without any surging (a problem I have without feedback through points/curves), and being an 8ft shunting plank they won't do high speed or operate for any great length of time.

 

Re the DCC bit, I was surprised when I read it could cause problems with DCC Ready locos and presume Gaugemaster mean "DCC ready" refering to locos with DCC chips (which none of mine have) as opposed to locos fitted with the blanking plugs

 

As I already have a non feedback controller which the HH plugs in the back of, I will wire in a DPDT switch so can change between Feedback and non feedback outputs if required.

 

Thanks once again.

[bike2steam]

Until I went DCC in the shed area and also the goods yard I used a HH model controller so essentially it was start stop action. Very good for many locos and just like DCC need to keep track clean.

 

The HH on my old Bembridge layout was wired in series with an HF1 track cleaner, the only time the track was cleaned was before an exhibition if the layout hadn't been used for more than 2 months.

[PWSlack]

........contemplating getting a Feedback controller...........some DCC ready and portescap motors are not suitable for Feedback controller.......

 

It depends upon the motives of the individual making the claim.

 

If, for example, one were peddling the latest, all-singing-all-dancing high-value smooth-DC controller aimed at preying on the concerns of those that have been taken in by the claim that feedback controllers reduce the brush life of coreless motors then one might expect to play-down the feedback type and receive increased sales as a result.

 

Some individuals will make the claim that "coreless motors don't like feedback controllers", which isn't very useful as a claim as no motor can express an opinion, and go on to spend many tens of pounds on a DCC system that does what it says, which is to enable 'n' trains to operate simultaneously, though not necessarily what the purchaser actually wants, which is to operate them realistically.

 

Most DCC controllers have some form of feedback algorithm somewhere in the mix.

 

The fact is that the feedback controller is set up to deal with a certain amount of rotational inertia, the "flywheel" property, of certain types of motor. Because of this property, the feedback type will have no difficulty dealing with keeping the speed constant for any of the more traditional "iron-age" motors, and indeed, some of the larger coreless types. However, as the size of the motor goes down, the inertia goes down with the square of the radius, so beyond a limit, one runs into behaviour problems, as has been pointed out above, with some of the smaller coreless motors. The Portescap 16C1 and the 1616 are usually inside this limit, and the 1219 is beyond it when operated from a heritage ECM Compspeed type CA. The motor and controller when paired are simply too lively for each other, and the loco bounces off down the track like a demented headless chicken. Some of the smaller Maxoms that one might find in larger scale models also exhibit this phenomenon.

 

There are solutions, and the simplest one is given in the EM Gauge Society Manual Sheet on the topic - one merely wires some resistance across the motor terminals to act as an electronic brake, a "slug" in control system parlance. The easiest resistance is a grain-of-wheat bulb, which can be arranged to appear behind the firebox door to simulate the fire which, of course, glows brighter as the locomotive works harder.

 

As to the claim of reduced brush life in coreless motors, well, there has been no failure observed here in >30 years of operation of ECM Compspeed CAs at various locations with both Portescap 16C1 and 1219s in 4mm and Maxoms in 7mm scale locos, totalling some >45 motive units. And there's no issue with electronic track dirt ionisers. That's good enough for most, perhaps?

[Jim104]

Interesting comments PWSlack, I have noticed some newer Hornby and most Bachmann OO gauge locos tend to surge on every roatation of their wheels at slower speeds, the Ivatt tanks being the worst offenders.

Most Hornby locos creep along smoothly when using the HH type controller but some of them do not. I have a Hornby standard 4 tender loco 75XXX which surges all the time. It is most odd how they all differ to the same type of cotroller. I wonder if the grain of wheat bulb would improve these poor runners or is there an alternative 'fix'

[Donw]

There are two issues here the use of feedback and the way the power is delivered.

As mentioned above feedback needs to be matched if it is not then surging is quite likely. DCC chips seem to use higher sampling frequencies to reduce the effect.

The other concern is the use of power pulses. If the power is given in 12v pulses then a slow speeds the back emf is low so the 12v pulse will produce a higher current pulse and increase the heating effect (IsqR) so depending on the motor and its installation overheating could be an issue. For that reason it is best if you have a non- feedback controller as well to suit any loco thats gives problems with the feedback one.

Don

[bike2steam]

I have noticed some newer Hornby and most Bachmann OO gauge locos tend to surge on every roatation of their wheels at slower speeds, the Ivatt tanks being the worst offenders.

 

Not helped by imprecise wheel quartering, more noticable on Bachmann split chassis locos such as Ivatt tanks, and older standard 4's.

[Rocking Shunter]

I have recently bought two Gaugemaster DF controllers with an HH add on and they have certainly made a vast improvemant to the running of my layout. They give excellent performance over points, crossings and dirty track with all my locos (Hornby, Bachmann, Lima, Mainline) with no drop outs or stalling and give a good impression of scale speed.

On the down side, the locos do tend to be a bit noisier and there is a faint hum from the controllers but I believe this is a small price to pay for the superior quality of control. I can't see me going back to a conventional contoller in a hurry.

 

Carry on shunting:)

[paul 27]

I have the same problem with a Gaugemaster DF, some Hornby motors are noisey on slow running, tried putting

resistors in them but have made no difference, anyway of making them run without the noise before i think of replacing the motors, there running on 12v strait from the controller they do run very slowly without problems apart from the noise.

[Donw]

If you have a freiend with a non-feedback controller or a local club it would be good to try them on that. If they are quiet on a steady DC supply then it's the controller/motor combination if still noisy it suggests noisy mechanisms. At a pinch you could try 4.5v from a battery.

Don

[Suzie]

The main problem with 'DC' feedback controllers is that they do not have any way of tuning the PWM frequency to match the motor, so you end up with a compromise frequency which makes some motors noisy at the expense of not overheating, and some motors hot at the expense of quiet running. All feedback controllers have to be PWM to provide a cutout time to measure the BEMF voltage from the motor, and the effectiveness of the BEMF feature is often compromised by all the electronics now fitted to most locos, especially modern ones designed with lighting and such and fitted with a decoder socket let alone the interference suppression components.

 

The problem of high current pulses from PWM controllers with an incorrectly set PWM frequency causing damage to small coreless motors (and even larger pancake ones in some cases) is a real one, especially when the motors are running at low speed/throttle setting, so it is worth heeding the manufacturers warning.

 

Fitting a correctly tuned feedback controller inside each loco gets round the problem for all motors, and cheap ones can be bought for around a tenner, with a remote controller available for a few tens of pounds more.

[paul 27]

What is tuned feedback controller in the loco? as i have said tried resistors to no avail, could the tv suppresor be causing the noise i know these are

removed on DCC models with feedback decoders.

[Jim104]

Fitting a correctly tuned feedback controller inside each loco gets round the problem for all motors, and cheap ones can be bought for around a tenner, with a remote controller available for a few tens of pounds more.

 

Suzie, Can you enlighten us please about tuned feedback controller inside each loco ?

thanks

[stewartingram]

There are quite a few misconceptions in the above posts, I have used a form of pwm contorller since the early 70s with no problems.

 

Using straight DC control, no DCC electronics involved, and if the loco has provision for a DCC chip it is NOT fitted, the following is the simple case.

A voltage is applied to the track, depending which rail is +/- will determine the direction of rotation of the motoer and hence which way the loco will travel.

The applied voltage will determine the speed, more volts =  more speed.

The voltage applied to the track has an electrical circuit through the motor windings, via metal wheels/pickups/wires/motor brushes etc. Thus an electrical circuit is completed and current will flow. In an ideal world there is CONTINUOUS flow of current which in reality does not happen due to dirty track, dead frogs, poor tracklaying etc. This interruption MUST be kept to an absolute minimum. Cleaning of track, wheels, pickups etc is a MUST, extra pickups can and should be added where possible. Lets however assume this is done.

The flow of electrical current through the windings of a motor produces magnetic fields, which intereact with fixed magnets within the motor. From our school days, we learnt 'like poles repel, unlike poles attract'. This is what causes the armature of the motor to rotate, and hence the loco to move. All very simple so far.

 

Now if a coil is moved in a magnetic field, an electrical current is induced in the coil - it acts as a dynamo or generator. The motor armature is rotating by means of the current fed to it from the controller; but this coil movement also acts as the dynamo, producing electrical current, known as the back emf.

 

The controller I use actually measures this back emf, to decide when to send current to the motor, thus if the motor is stalling it will adapt its current to the motor to prevent that stall. (Usual disclaimer about dirty pickups as noted earlier, we are ignoring those).

This is what a feedback contoller actually does; it 'reads' the motor and reacts accordingly.

 

A PWM controller is different, but the 2 functions are often/usually combined.

The typical model motor is designed to operate on (nominally) 12v DC. So for the sake of argument, 12v = max rpm (ie max loco speed, more of that in a moment).

So half of 12v, ie 6v, will give half rpm.

A quarter of 12v, ie 3v, will give 1/4 rpm. Simple? .....well no actually, because another factor comes into play. The STARTING voltage of a motor is higher than the MINIMUM running voltage of the motor. If we start our motor off on 12v, and reduce the voltage until the motor stalls, it might be say 2v. But to get it started again we have to apply 4v. This does not bode well for slow running!

A way round this is to always apply full 12v, but to switch it on and off, in pulses. So 12v on/off/on/off in equal amounts, will give an AVERAGE of 6v, so 1/2 speed. The 'width' of the pulses, otherwise known as the frequency, determines the average voltage/speed of rpm. Hence the term 'pulse width modulation' or PWM for short. Note that in its basic pwm form there is NO reading of the motor and hence no feedback, but as I said these can be combined.

 

My controller is a very old design from Wireless World magazine, which does just that. It appeared around the time the 1st Black & Decker mains powered drills had speed control built in, which used similar technology. I have found it extremely reliable, and will 'cog' over a crude motor such as an old Triang XO4 so slowly that you can see each pole come round and pulse over reliably.

I have found no noise problem WITH THE MOTOR, but the noise is generated in the gear mechanism. An older mechanism is usually geared to a lower ratio, and thus higher loco speed, than more modern types. Usually too, lower gearing means more space between the teeth of the gears, and it is the rattle of these that produces the noise. However, to put it into context, my controller will move an old - very old - R52 type Jinty 0-6-0 so slowly that you can watch the coupling hook climb over the next wagon coupling.

TBH I haven't even bothered to check any of the latest controllers on the market, as I'm satisfied with what I've got. I am aware of its pitfalls, if you want to call them that, nothing is perfect. I think they can be summed up as follows:

Cannot purchase, has to be handbuilt, not to everyone's taste

Causes noise with mechanisms as stated above

Possible problem with the hi-fi coreless motors (hence the design of the Pentroller), but as I don't use those.....

 

 

I have made a fair number of these,with only 1 known problem A friend ran a 009 loco with a headlamp - Minitrains Baldwin if I remember right - and during an exhibition, the chimney meted from the heat of the bulb. Probably because f the continuous 12v (pulse) appled causing heat in the bulb?

 

I have usually also used a RELCO track cleaner with the controller. Together with attention to detail in current collection, such as use of live frogs, maximum number of wheel pickups etc, this has resulted in reliable slow running. I would go as far as to say, I would use that Triang R52 performance wise on a shunting layout, though appearance wise it would not be up to scratch!

 

Though an old design, it was designed by proffessionals in the electronics sector. It was also not designed with commercial restraints such as cost as it was never expected to be manufactured for sale, so I expect to be a good design, Nit that it could never be improved of course! I suspect that most of the controllers on sale will have limitations of design however, being made in a commercial world.

 

Stewart

 

(Just a few thoughts and experiences of a PWM feedback controller and its use over nearly 40 years)

[D1059]

I beleive the pentroller is no longer available that used feedback but the output was stepped DC with only a brief emf sampling period. It was designed to work well with coreless motors and others.

Regards Don

 

We have some old Pentrollers on our club layout and all my N gauge locos run very well using them. This is a largish tailchaser layout and no overheating is evident with Pentrollers either.

 

I purchased a Pictroller a few years ago when designing my new shunting layout, but have only got to use it in the last couple of weeks as construction has finally got to the wiring stage - it works very very well, much better than my straight DC Gaugemaster. Control is ultra smooth and the handheld version is well designed for one handed operation.

 

Essentially the Pictroller is a modernised Pentroller with the advantage of not having to manually switch between iron core and coreless settings. Definitely recommended for those of the non DCC persuasion.

 

 

STEVE

[shortliner]

I can honestly recommend the Varipulse, either ready-built or home-made - I have yet to find a better controller for DC http://www3.sympatico.ca/kstapleton3/Index.html

[cravensdmufan]

Hi,

 

I'm using two old (20+ years! - excellent build quality) Gaugemaster HH controllers powered by a Gaugemaster M1 (2x16v AC output).

 

All locos, modern Bachmann and Hornby diesels as well as older Hornby and Lima diesels, run well with this set up with the exception of two Bachmann Class 101 DMU's that are my most recent purchases.  They surge very badly.  I'm wondering if anyone else has had problems using feedback controllers with these particular models?

 

I'm curious about Jim104's solution of reducing AC input to 10v.  Can anyone advise me what make or type of transformer would do that? And where to buy one?

 

Thanks very much in advance.  

[CourthsVeil]

Hi,

Conrad have one: no. 514292 - 89. Caution: experience with electrics needed! Dangerous device in inexpert hands!!

 

Another solution, simpler: insert two rows of three diodes (1N5400) each antiparallel between transformer and controller. Reduce the voltage by 2.1volt.

 

Hope this helps

  Armin

[cravensdmufan]

Very many thanks for that information Armin.

 

I certainly don't trust myself with the mains electric so I'll try the diodes! 

 

I'll post again soon to report results.