DCC Loco Conversions - 24 - Dublo Duchess

One of my oldest to convert, she was created from two wrecks in 2005 including conversion to 2 rail and fitted with a NEO magnet. A bit battered (as me) but I like her!

Bear in mind that if the loco, like this one has unplated alloy wheels they will require frequent wheel cleaning, whether in use or in storage.

Installation of a decoder is straightforward, modifying the chassis to insulate the original brush holder takes most of the time.

Very Important - the chassis was modified as in articles 22 and 23, failure to do so can destroy a decoder.

https://platform1mrc.com/p1mrc/inde...rsions-22-dublo-brush-holder-extraction.6017/

https://platform1mrc.com/p1mrc/inde...sions-23-dublo-brush-holder-replacement.6025/

This particular loco had the hole for the new insulated brush holder drilled out by hand and thin walled black heat shrink sleeving used as an insulator, future locos will be reamed out and be insulated with a thick walled heat shrink sleeving.


Rails RoS-218 Decoder

The larger package version of the Rails 'Connect' family is used as there is the space for it and it comes supplied with a wire harness which reduces installation effort compared to the RoS-6D. At the time of writing they cost the same.

All surplus wires are cut off in a saw tooth fashion as described in Ref 24.1, "General Techniques Part 2"

Update 04/06/22

Shown mounted on offcuts of decoder foam packing, however this has proved unreliable whilst working on 2-6-4 tank loco chassis as it does not have sufficient strength. I have replaced the foam on this loco with offcuts of laser board from scenic building kits.


This mounts the decoder so it bridges the screws which retain the cylinders in place and moves it into the cavernous loco body.

Apart from extending the grey wire the remaining wires were long enough as supplied.




In Operation

The current consumption of this open frame motor when on the rolling road without the body on DC is slightly higher than the Ringfield locos converted so far at just over 0.3A.

However according to the DCC meter it is taking 0.4A at which the decoder feels noticeably warm, it is rated 750mA Continuous (1.1 Amps Peak Power ) so it should be okay.

Anyway I have decided to run it open so it runs cooler, most Lenz decoders I have seen are all open to view their excellent build quality, wrapping in heat shrink makes it safer to install but does reduce the load capacity as there is no ventilation.

As luck would have it the two transistors that get hot are on the top surface as mounted on the loco chassis and by trimming the heat shrink carefully it still grips the edges of the circuit board to retain it in place.



Currently (May 2022) the data for the high power version RoS-8HP states continuous as 1.5mA (yes milliamp), 3 Amps Peak Power, I have queried this with Rails of Sheffield who have replied that it is an error and the continuous rating is actually 1.5A.

So the RoS-8HP is a suitable alternative for these locos and is currently priced the same as the other family members, however it is slightly larger physically and so cannot necessarily be used as a drop in upgrade to the RoS-218.

Cogging

Testing on the testrack and rolling road the loco exhibited coggng at low speed settings, the armature visibly jumping between poles.

Replaced it with a remagnetised original magnet and it was noticeably worse, unusable!

This is a first, less cogging with a NEO compared to an Alnico.

Refitted the NEO, back to cogging as before, next adjusted the starting voltage CV under basic speed settings from the default of 0, normally BEMF eliminates the need for a non-zero value as the voltage to the motor is raised until it moves on step 1 (of 28).

Setting the CV for a start value of 4 eliminated the cogging below step 2, which is fine for an express loco on my railway as I do not have a passenger station for them to call at.

They may be rarely stopped at a signal near the marshalling yard entry or exit but otherwise they start and stop in fiddle yards.









Summary

Locos that are 50 to 60 years old, worn and crudely engineered compared to the miniature skew wound motors and fine mechanisms of today cannot be expected to perform as well, but I am satisfied the way the old girl runs now.

On to the next loco type, Dublo & Wrenn 2-6-4 tank engines.

Discussion always

Jim

Return to Index

References

24.1 - DCC Loco Conversions - 11 - General Techniques - 2 of 2

https://platform1mrc.com/p1mrc/inde...onversions-11-general-techniques-2-of-2.5917/

 

Update 04/06/22 - applied above

Shown mounted on offcuts of decoder foam packing, however this has proved unreliable whilst working on 2-6-4 tank loco chassis as it does not have sufficient strength. I have replaced the foam on this loco with offcuts of laser board from scenic building kits.

Jim

 

Hi Echidna

Thanks, this was certainly an unusual case.

In practice even if the the power supply was un-interrupted each armature winding in these simple DC motors is energised in sequence via the commutator as it rotates.

On top of this the decoder output to the motor is pulse width modulated, that is it supplies a constant voltage but applied in varying durations. For example at full speed the motor may receive full voltage all the time, at half speed the full voltage is applied for half the time, momentum of the motor keeps it running in the off periods.

In overly simplistic simple figures e.g. on for 1 second and off for 1 second, or it may be turned on and off every tenth of a second so it averages being applied for 50% of the time.

These days the switching frequency is in the high kHz region, above human ear response so we do not hear an annoying buzz from the motor.



The NEO magnet was fitted in place of its original Alnico magnet long before I was able to re-magnetise Alnico magnets, so I cannot be sure how it ran with an Alnico magnet.
I have not reversed the changes to find out because ultimately it is how it runs on DCC that matters now.

Generally cogging occurs in these old Dublo locomotives when a NEO magnet replaces the original Alnico magnet. It is possible for the field strength of a magnet to be too strong for the design of the armature, causing it to jump from one pole to another at low speeds, especially as these locos are of a very basic 3-pole design with quite crude pole pieces.



However a further complication is the use of BEMF for automatic speed control, this is necessary for setting and leaving (background) trains running on the continuous mainline which has gradients or for slow speed control when shunting in yards.

Worn parts and coarse gear trains can cause BEMF performance to be unstable which can manifest itself in many ways, from juddering as backlash is taken up and released, to cyclic speed variations, this can vary between decoder brands/designs depending on the firmware (the built in internal programming) of the decoders.


Ultimately mixing worn old technology with new technology control systems designed for refined motors and fine gear trains is just an additional challenge to using DCC, my love of running my old relics with my contemporary locos drives me to solutions even if the reasons seem to defy general experience.

So there may be an armature fault, but the set up for this loco now allows it to run satisfactorily so I will call it a day and just run it

Jim

 

Thanks Paul.
Plus a skew wound coreless motor, alternatively just buy a current model

What drew me to the Rails Connect family of decoders is their 'advanced brown out protection', IMO it works well with these old Dublo locos on Peco streamline, but I doubt they would be good enough for running on original 3 rail track where a full blown stay alive system would be required, which are somewhat more expensive and bulky to install.

A few old locos I had with incandescent lamps even had corroded LES fittings, so I just upgrade them to LEDs, warm white works well on older prototype models, cool white for 'modern day' models whose prototypes would have halogen lamps, or now high intensity LED lights.

Jim