Decoders & Capacitors.

[Captain_Francisco.]

  Enlightment please. In my simple understanding of running DCC locos, all that has changed, as far as the DC motor is concerned, is where the DC is sourced from: originally directly from the rails, post decoder fitting, now originating from the decoder: from the motor's point of view then, nothing has changed. Why then is it, for example in the guide to fitting a R8429, or any decoder, in the standard 0-4-0,  is it necessary to "mess" with the motor's, previously quite acceptable power supply, and remove a capacitor? Why can't the wiring from what was previously soldered to the "pick-ups" just be soldered directly onto the decoder "motor" leads?

  Secondly:-  Is this a "constant," or does it vary from "motor to motor," "decoder to decoder," "decoder/motor marraige?" If affirmative, how does one arrive with a reasonable degree of confidence as to what is removed, and what remains? My reason for asking, apart from expanding my understanding of our hobby as it now stands, is that I have some 1960's/1970's vintage, (including a much treasured "Britannia",) Triang/Hornby loco's, and a selestion of "Lima" HO models of the South African Railways Electric & Diesel locomotives, 1980's vintage, that I would like to convert.

  How may one proceed with some degree of confidence? All my decoders for the intended installations, that I have accummulated, are Hornby R 8429's.

  Thanking all in advance.

[morairamike]

The capacitors that are required on a DC loco are fitted to make the loco comply to TV interference regulation.

When you fit a decoder the system is fed from the rail by what is to all intense and purposes a modified AC waveform. 

The capacitor fitted in the power to rail connector can modify this waveform and should be removed from the clip or even better the wires soldered to the rails and the clip put in your spares box. The DCC version doesn't have the capacitor in it in the first place.

Now your decoder is busy crunching the info and trying to run your loco for you but the capacitor across the motor can disrupt the final output from the decoder. So the best policy is to remove all capacitors from the motor. The decoder has the required suppression built into it.

In reality there isn't any difference between a 1960's loco and a 2019 BUT you may find that the 60's loco needs a beefier decoder to handle the larger stall current coated to the 2019 one, i.e. open frame motor or ringfield compared to can motor. 

[walkingthedog]

In short, capacitors not required, decoder has suppression built in. 

[morairamike]

True but he in my opinion required a more detailed reply as to why not required.

[Chrissaf]

If one takes the original manufacturers approach to suppression (all manufacturers, not just Hornby) then it would appear that there is no common 'hard and fast' rule for this. Ignoring 'DCC factory fitted' locos for a moment. Focus on the approach manufacturers take with 'DCC Ready' locos, as these locos are more representative of what manufacturers do for suppression as these locos need to comply with suppression standards for both DC & DCC working.

.

To comply with UK & Euro suppression standards, motors should be suppressed against generating 'radio-wave' noise. A 'spark' any spark is a broadband radio-wave noise generator. Some motors can generate more sparks than others as the carbon brushes rub against the spinning commutator. The more current the motor draws, then potentially the more intense the sparking can be and the greater the range of generated radio-wave noise. And as older model motors tend to draw more current then the need to suppress them is greater.

.

If you are old enough. You may remember that your old 405 line Analogue TV would announce the pending arrival of an unsuppressed motor bike driving past the house before you could hear its exhaust. The TV sound would start to crackle and the picture exhibit interference lines. Later, the laws changed such that the ignition on bike engines had to be suppressed.

.

Anyway, back to the plot.

.

Some 'DCC Ready' manufacturers keep the suppression components across the motor in the factory build and insert a basic bypass 'blanking plate / plug' that just provides pass through for the wheel pickups to the motor. Thus, when the bypass blanking plate / plug is replaced with the DCC decoder. The motor suppression components are still in-situ and somewhat a duplication of the suppression provided inherently by the decoder.

.

Some manufacturers of 'DCC Ready' locos include a slightly modified 'blanking plate / plug' in their factory build. Where the bypass plate / plug has the suppression components mounted on the 'plate / plug'. Thus, in that scenario when the 'plate / plug' is replaced with the decoder, the suppression components directly across the motor are removed to be replaced by the suppression function of the decoder.

.

One would assume that in these cases, the manufacturer has deemed that for the motor specification that they have used in the model, removing the 'DCC Ready' suppression is technically the optimum solution. But as you can see, there is no 'hard and fast' rule that the manufacturers adhere to.

.

This doesn't really help you with your question, because at the end of the day the answer is 'it depends'.

.

My personal choice which I cannot justify with a logical explanation. Is to keep the components (if fitted) if it is a 'DCC Ready' loco. If the loco predates 'DCC Ready', then in the first instance I would also keep the factory fitted suppression in-situ. Monitor the operation under DCC, and then remove them if operational DCC issues are observed. If it ain't broke, don't fix it is the rule I apply. Others here, may have other opinions and rules that they apply instead.

.

Unfortunately, your question is an example of a 'Marmite' question. There are equally conflicting opinions about what should be done with the factory fitted suppression components.

.

PS - As M.Mike correctly stated. Any capacitors on the 'track side' of the decoder must be removed for DCC working. So in a 'hard wired' loco DCC conversion, it is imperative to insert the decoder on the wheel pick-up side of the suppression components (if they are retained). As well as removing any capacitors across the track in the track power connector products.

.

 

[ColinB]

I recently did this with one of my locos I converted to DCC, only because I was having issues. DCC is a high frequency signal and the capacitors on the motor are meant to get rid of high frequency signals, that is what interference is. So worst case the capacitor could effectively in the case of DCC stop it working. In reality, in my case it just made the motor a bit sluggish. I could go into the detail of low pass filters, but I don't think you really want to know. Hornby should do what Bachmann and Dapol have started to do, put the capacitor and inductor on the circuit board that goes in the 8 pin DCC socket (with Bachmann and Dapol they use 21 pin and have more room),  so that when you convert to DCC they get removed with the "plug in" DC adapter. With Hornby putting the connector in the tender, the only issue is it would be a bit far away from the motor and not be so effective. The issue of removing them is if you ever decide to go back to DC, which is why it would be better on the DC "plug in". I don't think removing the motor capacitor and inductor has ever corrected a fault, with my DCC, but that is my personnel opinion, everybody tells you to remove them. The only thing I can possibly see them effecting is the very slow speed control of the motor. 

[Chrissaf]

The high frequency DCC signal (nominal 7,000 cycles per second) is on the track (pick-up) side of the decoder. When the decoder is fitted, the decoder should (if wired and working correctly) block this high frequency DCC signal from being applied to the motor. Thus the capacitor across the motor should not have any direct impact on the DCC data, the same way that a capacitor on the track side of the decoder would have. Like for example a suppression capacitor that is left in-situ within a track power connection product.

[Captain_Francisco.]

Thank you all.  I am getting a far better understanding of the business of powering and controlling small electric motors, and the supression of interference, both from the more practical, and Nationally required, standpoints. "walkingthedog" absolutely "end result" focused; "modern gadgetry" takes care of the necessary, "out with the old, in with the new," and sit back and enjoy. Great. "Morairamike," thanks, I have followed previous advice and any power clips or tracks I possess have been "neutered." I now have a collection of suppressors and am wondering what they may be useful for in the "scenic" area. Chrissaf, lovely comprehesible reply, and I particularly liked your, "If it aint broke don't fix it approach," and allied with ColinB's experience, I think that is the approach I shall adopt; discarding if and when necessary. I am really pleased to find that this most practical approach is not "catastrophically" damaging to any components: their replacement uses a up a fair amount of the 'disposeable for entertainment "Filthy Lucre".' Thanks all, once again. I "sign off" an appreciably happier fellow.

[ColinB]

The only thing I was concerned about with the motor was if they were using a pulse width modultated signal to control the motor. I just assumed they were, as you get better start up torque, and it is easily to do if you have a microprocessor which I assume the DCC decoder has, but if is straight DC to the motor then it doesn't matter. 

[Chrissaf]

The motor control output on a DCC decoder is constructed as a H Bridge motor control circuit. For this to have any speed control over a motor then it has to use some form of PWM as the FETS used as the output H bridge semi-conductors are either fully on or fully off when switched.

.

A typical (example) of a FET H bridge motor control circuit.

/media/tinymce_upload/16186a10bd09f4acb5aff84ddda464d0.png

.

• When Q1 & Q4 are on....Q2 & Q3 are off and the motor turns one way.
• When Q1 & Q4 are off....Q2 & Q3 are on and the motor turns the other way.
• When all Qs are off, the motor doesn't turn.
• The decoder controls the switching on & off and sequence of the four Qs. If the off periods are longer than the on periods then the motor turns more slowly.

[ColinB]

That explains a lot, it also explains why it is so easy to blow them up if you accidentally get the track voltage on the motor outputs. Do you know if they use separate chips, or is it 1 iintegrated circuit. I am just wondering if I might be able to fix some of my dead ones.

[96RAF]

R8249 motor control is by way of a pair of dual-mosfets in H-bridge exactly as per Chris’s diagram. Main control is by way of a PIC chip signalling a pair of Fets that take care of the H-bridge switching.