DCC Loco Conversions - 42 - Dublo & Wrenn 1000HP Bo-Bo

Here I attempted to convert two versions of the Bo-Bo loco, I have a third but these two essentially cover that too, there are multiple two rail builds of this loco, with variations in bogie castings, power pickup and brush springs.

I have divided this article into three sections, spread over two parts due to its length.

Part 1

42.1 Hornby Dublo Loco #1​

The relatively straightforward conversion of the first Dublo loco as a successful conversion.​

42.2 Wrenn Loco​

The trouble and failure of converting the Wrenn loco.​

42.3 Hornby Dublo Loco #2​

The trouble and failure of converting the second Dublo loco.​

Part 2

42.4 Summary of the work.​

So you can choose to read the good, the bad or both!

Part 1

Very Important - Both arms of the brush spring must be insulated to ensure track power does not connect to the decoder motor drive output, failure to do so can destroy a decoder.

I do this with PTFE sleeving as it is not affected by soldering to the brushes, if you use PVC ensure it is not melted, either solder before fitting PVC sleeves or insert some protection.


42.1 Hornby Dublo Loco #1 - Successful Conversion

Most Dublo and Wrenn DCC conversions I have performed recently benefited from the Rails of Sheffield (RoS) Connect series of decoders, in particular the RoS-6 decoder, but alas no success with these locos.

All original wiring was removed as it was in the usual poor state of construction plus aging, additional wiring with 10/0.1 PVC insulated wire.

The red wire from the unpowered bogie attaches to the solder tag on the motor bogie and then another wire from the tag to the decoder.



The black wire from the unpowered bogie attaches directly to the decoder via a solder tag anchored to the bogie via an 8BA screw.

The Dublo unpowered bogie required a suitable hole which was drilled and tapped into the bogie to avoid using the original tag at the top of the pillar which attaches the bogie to the body mounting plate. Sound electrical connectivity is an issue with these locos so eliminating at least another potential poor connection is worth the effort.



Orange and grey decoder wires are soldered directly to the brushes as usual with a further piece of unshrunk heat shrink sleeving acting as a strain relief, which incidentally makes soldering to the brushes easier as it keeps the tinned end of the wire in place against the brush arm while flowing solder into the joint. Refer to images 1 and 3.

RoS-6D

Wires are attached to the RoS-6 decoder pins by soldered lap joints with un-shrunk heat shrink tubing providing some extra support and insulation. Surplus pins 5 and 6 cropped off.

Although okay on the test track the loco was jerky in operation to the extent of being unusable on the layout.

Lenz 10443-01 Gold

I have had this spare for many years so I thought I would try it.

Surplus wires were de-soldered from their pads and the remaining ones reinforced with Kapton tape and subsequently with woven nylon lacing tape which also provides extra security to retain the decoder to the mounting plate. I did try cable ties but the clearance between the body and the mounting plate was insufficient for the ties I had.

This decoder is not shrink wrapped so if it comes adrift from the double sided pads it will likely fry on a live metalwork. A second band of lacing tape is applied to the other end of the decoder board.





Up to step 7 of 28 it ran smooth, however at higher speeds, i.e. above step 9 the loco started faltering and once again became unusable.

Next I attached the stay-alive module for this decoder which I also bought many years ago, namely the 10490 (POWER1) module.

Retained by a small double sided sticky pad and a cable tie as space permitted here, refer to image 2.

It now runs well with top speed limited to a scale 70mph as per the prototype, although running on my layout will likely be lower speed freight working.

Looking all around.









42.2 Wrenn loco - failed to convert this one

Multiple decoders tried before giving up

RoS-6D

Loco totally unusable with the RoS-6D.

Lenz 10410 Gold mini

I then tried a Lenz Gold mini I had spare, but without stay alive, performance was poor and jerky as was the 10443-01 Gold without stay alive.

The Lenz stay alive unit at this time is in short supply and available at an inflated price on eBay so not tried with this decoder.

DCC Concepts DP4SA

Next I tried an old DCC Concepts DP4SA, this was an early bulky 8pin plug in decoder with stay alive which I bought a batch of when Hattons were selling them off cheap. Performance was again unusably jerky.

Lais 870021 + 870009 SA

Next I decided to try the Lais 870021 decoder paired with the 870009 stay alive unit, readily available and somewhat cheaper than the Lenz parts.

On the test track with just the Lais decoder the Wrenn chassis was a little jerky but with the stay alive unit attached it ran smoother.

On the layout not so good, ran well cab first, but bonnet first it would just stop and restart at frequent intervals from speed step 14 of 28 upwards (again top speed set to be a scale 70mph).

So was this over current, pickup failure or something else? (See summary).

Tried different set ups :-

1) BEMF off - no difference
2) Brush carbon high or low (some vary enormously) such that they are not diametrically opposite each other - could be made worse but not better
3) Bypassing the pickups with clip on leads - no difference
4) NEO or Al-Nico magnets, worse with Al-Nico
5) Reverting to DC to the motor, runs fine, not an armature failing as on another loco.

So, okay good enough at 35mph forward for unfitted goods train, however later it would not run properly cab first at all!

The unpowered bogie differs from the Dublo bogie as it is a powered bogie chassis without the motor components. I packed the motor void with layers of lead to improve power pickup - very little improvement, but does suggest power pickup is part of the problem.

At this point I have given up with this loco.


42.3 Hornby Dublo Loco #2 - failed to convert this one

Lais 870021 + 871007 SA

Being rather determined in my approaches to these tricky situations I then tried the Lais decoder with (a differently packaged, to fit easier) stay alive in a 2nd Dublo loco, just in case it was an issue with the Wrenn version only.

As with the Wrenn loco it ran in bursts forwards (bonnet first) and most of the time in reverse (cab first) up to speed step 14.

Tried different setups :-

1) BEMF off - no difference
2) Reversing loco by CV - no difference
3) By passing pickups - no difference
4) Reversing motor by reversing magnet or swapping orange and grey wires - no difference.
5) both brushes below armature shaft centre line, adjusted - no difference.

Basically it is related only to the direction the loco runs, and behaves the same whether on the track or rolling road.

It was very noticeable that the commutator on this loco is somewhat eccentric so could it be loss of connection due to brush bounce at certain rotational speeds?

The information from Digitrains include a note that these decoders may be used as static decoders by replacing the motor with a 560 ohm resistor, so as that and the motor in parallel could easily be driven by the decoder I tried them in parallel - no difference!

Digitrax DZ126

Whilst waiting for feedback on availability of the Lenz POWER1 SA unit I thought I would try another reliable decoder the Digitrax DZ126, nothing to lose as they say.

Seemed okay to start with on the test track and was improved by adding weight to the unpowered bogie to bring the total loco weight up to 453gm.

Unlike the Wrenn loco the fixed bearing was loose in the bogie casting, so I fixed this with aluminium foil.

Just for a change on refitting the body there was scraping sounds above step 14 of 28 which sounded like wheels against body, but they are nowhere near each other, in the end I just reversed the body and all was quiet. Then reversed loco forward direction by CV. These locos are unbelievable.

Moving to the layout from the test track, it hesitates and in places totally fails to continue, so it looks like a stay alive is required on the layout for this loco too.

Ho-hum, try the DP4SA here.

DCC Concepts DP4SA

Ran smooth at all speeds on the layout, but, and a big but, over a certain speed the loco became uncontrollable, which means that once it has accelerated to the required speed it cannot be contacted again to reduce the speed let alone stop.

This is a problem because a run-a-way Dublo loco on a layout with many contemporary items of stock around is potentially very destructive, even the relatively light Bo-Bo could still demolish a contemporary wagon or loco in a fast collision.

The only way it could be stopped was by applying an emergency stop which cuts the track power, when released the loco is back in contact with the data from the DCC command station and stops.

This does however give reason why these locos will stop frequently when driven by the Lais decoders, they have a programmable option which will halt a loco if it has not received data for a certain amount of time, the default I believe is 0.25 seconds.

Stay alive keeps the motor running, but if power pickup is just bad enough then command contact is lost as loco track speed rises.

This does not explain why the problem could happen on the rolling road with the pickups bypassed or may be the same loss of communication can happen with brush bounce.

 

Part 2

42.4 Summary

It can be a relatively simple physical conversion but ensuring power pickup is sound is difficult, also there is little space to mount a decoder especially if the unpowered bogie requires ballast.

It seems that the two rail version of this loco was a poor performer by design, the bogies will easily rise up on two wheels either when driven or a load is attached to the couplings due to the single point attachment to the mounting plate (joining bar), reducing traction but far worse for DCC the number of wheels for reliable power pickup.

The Co-Co and Co-Bo have similar mechanical traits due to the single point attachment of their bogies but their weight helps to keep the bogies horizontal, particularly the 6 wheel bogie. This loco model is probably the only Dublo/Wrenn loco I have converted that requires a Stay Alive supported decoder, assuming it has sufficient time to keep the stay alive capacitor(s) charged for intermittent use.

Use of stay alive on a decoder which does not check for communication integrity will lead to run-a-way locos, does the Lenz Gold used on the first Dublo check for communications, that I do not know. Until I obtain another Lenz decoder with stay alive I cannot confirm this, at present the cost of trying this out is prohibitive.

The second Dublo loco leads to the question, can data communication be corrupted by brush bounce on an eccentric commutator which causes the decoder to malfunction because there is not a motor or a suitable resistive load attached to it during the bounce?

Was this worth the effort, I think not, after many frustrating hours I have one that works well and at least two others which may now be left to gather dust before I turn to dust trying to convert them.

Stop Press: An alternative stay alive unit compatible with Lenz Gold is the Train-o-Matic SPP (Smart Power Pack) 02020201 which I may try when it becomes available, it is much cheaper than a Lenz SA unit at £20.


Discussion always

Jim

Return to Index

 

Thanks Echidna.

I have not had any issues with derailing, I would suspect that assuming the back to back is correct then one or both of the screws connecting the bogies to the mounting plate may be binding, also there may be a washer missing at one end, at the motor end brass, fibre at the other end, has that one broken up or frayed? There is a delicate balance to be made there between flexibility for curves and sound electrical pickup.

I think the primary problem with the 2 rail HD cl20 is that the unpowered bogie is unstable, it can pivot like a pendulum from the screw attachment to the body mounting plate.

This appears to be exacerbated by the low body weight, so when driven by the powered bogie the leading wheels having some drag, partly due to pickups and any extra from axle friction which causes the bogie to tip slightly pivoting on the the leading axle and thus reducing the weight on the trailing axle and its ability to pickup power. The action can happen both ways.

Also coupling rollingstock to the unpowered bogie can make it worse.

As the number of wheels with power pickup are limited, and the need for almost perfect wheel plating, difficult on these, as they are quite old now and many are badly worn this lack of wheels in contact with the track makes DCC conversion difficult. I believe the 3R version worked ok because the wheels picked up power from both rails for, lets say the outer return rails and the plunger/slipper from the centre feed rail, or simply it had more power pickup points.

Over the course of converting many 'vintage' locos I have found that some decoders will just not work with certain loco mechanisms, the trouble is that there are no circuit diagrams of these decoders available and certainly no information on how they are internally programmed whether using microcontrollers with embedded software or arrays of fixed logic performing their proprietary motor management algorithms. They are black boxes with relatively limited tweaks available for loco tuning.

Jim

 

Thanks for sharing Dave, appreciated.

That is really useful to know

Where did you use your 3D printed parts in this conversion?

Jim

Return to Index (DCC Loco Conversions Index)

 

By the way, if your magnet is weak be careful that the current the armature draws does not singe your decoder.

Was the 3D printing done with a resin or filament machine, I am sure other members would also be interested in more details if you have the time.

Jim

 

It sure does, although it took me some time to get into using the free SketchUp Make 2017.

Your parts are like most I print they are relatively straightforward but are difficult and time consuming to fabricate and even tedious when many are required, but once you have the CAD model you can then make more if you want, also use as a basis of further designs, you can even have a change material dependent on your printer.

My printer is now becoming a tool like any other in my toolbox, at last, I pick it up (figuratively speaking) and use it.

How well do your parts turn out, I find that with mine, particularly when printing small items like the bush they are a bit hit and miss when it comes to usability, how they look in the CAD tool can be quite different to how they look when printed with a 0.4mm nozzle.

Jim

 

Thanks, we ought to move this discussion onto 3D printing, I certainly found making spare copies helped, Jim