Here is a mixture of issues that can trip you up and tips on getting more out of your old locos (DC or DCC), primarily for Dublo and Wrenn locos but some of these, particularly on soldering are also relevant to many other makes. Not in any particular order and will be added to as I go along, split into parts within this thread due to thread post limits on image count and text length, contributions by reply are welcome. Part 1 Includes :- 01 - Trip - Undersized Magnets 02 - Trip - Pole Pieces 03 - Tip - Cylindrical carbon brush lengths 04 - Trip - Split brush tube insulation 05 - Trip - Incorrect Orientation of Worm Gear 06 - Trip - Coarse or Fine Armature Worms 07 - Tip - Seized Axles 08 - Trip - Missing Thrust Ball Bearings 09 - Tip - Ringfield Brush Springs 10 - Trip - State of Wire Attached to Collector 11 - Tip - Collector Cover 12 - Tip - Evaluating Max Current Rating for a Decoder 13 - Tip - Run Naked! 14 - Trip - In full view dry joints 15 - Trip - Commutator fitted wrong Part 2 Includes :- 16 - Trip - Over tightened valve gear rivets 17 - Tip - Stall current capability for a decoder may not be enough 18 - Trip - Why does BEMF no longer work? 19 - Tip - Cylindrical Brush Shaping 20 - Trip - Rotating Cylindrical Brushes
Dublo & Wrenn Trips & Tips - Part 1 1 - Trip - Undersized Magnets - the magnet could barely make contact with a pole piece, no wonder the loco was a weak runner even after a re-mag. 2 - Trip - Pole Pieces - (open frame motors, e.g. N2, 2-6-4T) - these are often badly formed, good ones are curved around the armature some later ones had corners, with poor alignment and spacing. As high current toys on train set ovals this did not really matter much, except for excessive current being drawn leading to faster brush wear and the early demise of the armature. When converting to DCC we need these motors to be as efficient as possible, typically 0.25A @ 6V - 10V is good target on an old H&M power unit (No body fitted, on a rolling road) some up to 0.4A tolerable. Minimising the gap between pole piece and armature can make quite a difference as it maximises the magnetic field strength at the armature. Also some are curved where they should be flat, e.g. past the clamping screw and to the magnet, I've come across some awful so called pole pieces. Most are fairly soft and can be easily reformed, others are very tough, the grade of metal used varies enormously. 3 - Tip - Cylindrical carbon brush lengths - this can make quite a difference, too short they become unstable against the commutator and finally drop out if the spring force was still adequate for the motor to run. But also I have found that new brushes (3/8 inch, approx 9mm long) in a brush holder of an open frame motor can result in too high a contact force. Trying out three lengths, almost too short at 5mm, 7mm and new 9mm, 7mm gave optimal performance, being nearly 1/8th amp less current draw compared to a new brush. Even at 5mm it was much lower but not much running time left. They will wear down, but what is the point of creating unnecessary carbon dust, higher current consumption and armature heating. 4 - Trip - Split brush tube insulation - found on an N2, it looked suspicious at the plug end, when extracted a two layer plastic tube was split along its length. How long before that brush tube would have made contact with the live chassis and caused a decoder to burn? 5 - Trip - Incorrect Orientation of Worm Gear - Wrenn A4, excess wear, generally they fit one way around due to the boss with grub screws. For a Dublo open frame motor make sure that the worm gear teeth line up with the centre of the armature lower bearing. Similarly for Ringfield motors. The worm had chewed so much off one edge of the worm gear it would no longer mesh at all, I suspect the worm gear was installed incorrectly from new. 6 - Trip - Coarse or Fine Armature Worms - Very deceptive, I recently ordered a fine then realised it should be a coarse, and when they arrived I was right originally - fine! 7 - Tip - Seized Axles - caused by sticky oil, over time the oil's volatile components evaporate away and leave a sticky residue and renders the chassis seized. I find that thin oil often fixes this but in a serious case the chassis may need soaking in a degreasent to remove the residue first. Warning, ensure that any degreasent used will not attack the wheel bushes or the wheels will fall off! 8 - Trip - Missing Thrust Ball Bearings - not always apparent but can cause significant friction when missing. If you cannot get them as Dublo or Wrenn spares then just buy 3rd party ball bearings, they will not need to be high specification for these motors. 9 - Tip - Ringfield Brush Springs - adjusting the ends against the brush to be parallel to the housing allows them to press evenly on the brushes, when the brushes wear significantly then tweak the springs accordingly. 10 - Trip - State of Wire Attached to Collector - on many locos, this wire is often frayed at the connection to the collector due to poor wire stripping during manufacture or user replacement of a collector wire such that only a few strands of wire remain. Disturbing this hidden connection when fitting a decoder may result in breakage and mysteriously your newly decoder fitted loco does not respond. So it is always worth checking this connection when the loco is stripped down for servicing, e.g. to lubricate the centre axle of a six coupled steam loco chassis. Here are two examples of bad joints, from two City locos, a Hornby Dublo and a Wrenn. The solder blob wraps around the wire but has not wetted the collector surface, at best a resistive connection at worse fails intermittently. In this Wrenn example the black wire has a solder blob attached to it, the wire is not soldered to the collector, only the hook of the wire and blob hold it loosely in place. In the next image (top) the desirable connection is shown between wire and collector, the original connection is shown by the white arrow. A small hole between the solder blob and the insulation was where the collector wire sat. The next is a close up of what is a good joint, both wires visible, and a smooth, shiny fillet of solder merging with the surface of the joined items. Big blobs mean trouble! 11 - Tip - Collector Cover - attempting to solder to the collector wire whilst the collector cover is still in place can easily result in a melted cover, do not risk it, some of these are now expensive spare parts. Finally only nip up the screws, I find many have split covers due to over tightening of the cover screws. 12 - Tip - Evaluating Max Current Rating for a Decoder - Consider, if the mechanism jams solid for any reason, which is cheaper to replace the decoder or the armature? This is why I opt for wheel slip current instead of stall current, this has the additional benefit of using a physically smaller decoder. I am generally using decoders designed for N and a few for Z gauge locos in most of my OO gauge DCC loco conversions. 13 - Tip - Run Naked! - Get greater power delivery from your shrink wrapped decoder by running it naked, just ensure it is restrained from contacting live metal work, not all decoders are fully wrapped by design, e.g. Lenz. Mount with the 2 or 4 three legged components skywards. 14 - Trip - In full view dry joints - Whilst removing suppressor parts from my Wrenn City I find that the wire attached to the chassis solder tag was a loose fit in the tag hole for the wire. All that stopped it from coming adrift was a blob of solder on the end of the wire which was too large to pass through the hole for the wire! 15 - Trip - Commutator fitted wrong - this is a real cracker generally the commutator slots on these open frame motors are midway between the windings. My Wrenn City, again, which struggled to run, had a significant angular error of the slots relative to the windings. To fix I de-soldered the winding wires, bent them clear and then used a wheel extractor to pull the commutator free. However each time I pushed it back on it would twist out of line, but at least now I could see that the commutator was fitted to a plain shaft and not a splined one. So after three attempts I wrapped some masking tape around the jaws of some long nosed locking pliers and twisted the commutator back into its correct angular position. The armature wires were re-soldered under slight tension, rather tricky, after testing at low speeds the wires were tacked into position with adhesive to the wire already wound onto the armature frame. Dublo locos often had a coating of goo to stop windings coming loose, I did not see any such coating on the Wrenn. I could not get the camera lens square on for this next shot, but at least it shows two out of three slots aligned correctly! Minor angular errors can also contribute to a motor running faster one way than the other, but only up to a point. For example, tuning an electric motor in a radio controlled racing car to run slightly faster in the forward direction by adjusting the angular position of the brush gear, too far performance reduces again, reverse was only for backing out of a crash! Continued in Part 2
Dublo & Wrenn Trips & Tips - Part 2 16 – Trip - Over tightened valve gear rivets - The Wrenn City again, the forward pivot on the union or connecting link was stiff enough to cause a noticeable strain on the valve gear (illustrated after replacement below). Attempts to free it failed, the rolling chassis would not roll. This fault must have been there from new, the rivet head was damaged probably due to misalignment with tooling on assembly. If you cannot obtain the correct rivets an alternative is to use a tiny self tapping screw sold to repair spectacles and the like, assorted sizes in a sorted box of 1000, readily available online. These screws have quite flat heads similar to the rivets, are steel but low grade, an appropriate sized screw was fitted into slightly enlarged holes in the links to be joined. Whilst in situ the excess length was cropped off with piano wire cutters, and tidied up, result as shown below. To refine the fitting the thread directly under the screw head could be reduced to avoid trying to cut the brass of the link but remember this will make for a sloppy fit, or even fit a thin brass tube to the screw. So a reasonable solution, some current locos have these type of screw heads visible even now, it is not as though this loco will get significant running so any wear is unlikely to be an issue. Of course after making this modification I found some Wrenn rivets on eBay, typical, but when they arrived they weren't suitable either! My Wrenn City loco was very much a Friday night and Monday morning loco as the assembly faults were numerous, mechanical and electrical. 17 - Tip - Stall current capability for a decoder may not be enough When converting old Dublo and Wrenn locomotives they may have suffered thermal damage to the armature windings due to weakened magnet by age or poor maintenance procedure, e.g. not using a keeper. This can materialise when you least expect it as in my Dublo City of London which fried a RoS-6D in the process, the mystery is that the windings appear to be of relatively high resistance but required 0.5A in forward and 0.4A in reverse, but could adversely overheat a decoder when running. The windings do have ominous dark patches, so possibly a damaged wire is shorting when the armature spins, so although greater than 0.65A is considered the absolute top current for a Dublo armature I have found that 0.25 - 0.3A can be achieved when on a rolling road without the body on DC. On the layout when hauling 6 coaches the DCC RRamp meter can show the current as low as 0.2A. Some decoders including Lenz include short circuit protection, however e.g. the Lenz is rather larger and was rather more expensive than most which really mounts up if you have many locos to convert. Removing the shrink wrapping from the fried RoS-6D did not reveal any damaged components. Probably one of the power transistors feeding the motor was stuck 'on' and shorting the incoming power to the decoder for the motor when it's opposite number was also turned on. It is a drawback of DCC that instead of the track being fed with a maximum of 1 to 1.5A as in a DC system with DCC it can easily be up to 6A. Even with breakers installed to limit current in seperate areas it can still be 2 to 4A, so when a short occurs it can cause far more damage, especially to electronics and delicate loco pickups. 18 - Trip - Why does BEMF no longer work? Whilst converting my Dublo & Wrenn City locos, yes those two again, I initially found that they appeared to hunt on the RoS-6D, an occurrence mainly noticed on the 2-6-4 tank engines so I switched them to the good old next in line decoder the GM DCC93, but for some reason BEMF did not work on either. The loco speed would run away downhill and slow significantly up hill. The RoS-6D is self adjusting, which means it either works or you can do nothing about it if it doesn't, the DCC93 appears to share terminology with the Digitrax decoders, Digitrax publishes info on manually adjusting the BEMF settings via CV 55,56 & 57. But to no avail, so I tried a Digitrax DZ126 and the beefier DH126, still no joy. Then I noticed that loco torque (the amount of strength to rotate the driving wheels) exhibited by them was low, if the loco was stopped by a finger the wheels just stopped rotating. Now if this was due to a weak magnet the motor current would rise steeply, but it did not, it stayed low. Closer examination of the motors whilst on the rolling road showed unusual commutator activity such as a ring of fire effect on one of them. It turns out that I had over oiled the top bearing and the oil had found its way onto the commutator and was acting as a partial insulator. Once the commutator and brushes were cleaned up torque and BEMF was restored. 19 - Tip - Cylindrical Brush Shaping Pre-forming new brushes for an open frame or Ringfield can be achieved reasonably well by wrapping some 400 grit abrasive paper around a piece of tube or rod to match the diameter of the commutator, e.g. 1/4 inch or 3/8 inch for the open frame motors. If one turn of the paper is not enough, wrap it a turn or two extra to obtain the required diameter, the result is a custom made cylindrical file. The less forming done within the motor the less carbon dust to clog the commutator. 20 - Trip - Rotating Cylindrical Brushes Just when you think you have pre-formed the brushes for maximum contact, what happens, well due to eccentricity of some commutators, the looseness of the brush in the holder, and nothing to stop it doing so it proceeds to rotate. It does so because brush droop causes the top edge of the brush to be more in contact with the commutator than the lower edge. The commutator behaves as a pinion and the brush as a crown gear, aggravated by the slots nibbling away at the brush like cutters. At lower speeds the brush can be seen to rotate as any eccentricity of the commutator vibrates them against their springs above which speed they may settle at 90 degrees to the shape that was formed because those edges are the weakest part of the brush surface. So, above a critical speed they stop rotating and if they are at 90 degrees to the original formed shape they get grooved by the commutator which results in a castellated face. Even with a perfectly aligned brush holder tube the brush can still droop as it is a loose fit in the tube, this is made worse by the gap between the tube and the commutator and increases as the brush wears and reduces in length until it finally drops out. This rotation was clearly visible on my Dublo City, viewed with an eyeglass, the brushes rotated under power, direction dependent on armature rotation. I have also seen the result of this happen on a Deltic Ringfield motor bogie which due to the nature of the conversion the brush and its spring was rotated and 'unscrewed' the spring past the retaining pin. This only stopped when the lack of spring force on the brush to the commutator was insufficient to power the armature. The retaining pin is illustrated in the Dublo Co-Co (Deltic) conversion. This problem can occur in the open frame motors and the Ringfield power bogies. The Ringfield motors that utilise spring wires that engage in slots in the brush ends will have this rotation mechanically restricted, e.g. 0-6-0DS and 8F etc. This means that when replacing those brushes it would be advisable to ensure they have a slot cut at the spring end if not already present. Contributions to this thread are very Jim Return to Index
