C30T RTR HO scale - Wombat Models surprise release

 
  NSWRcars Assistant Commissioner

the eBayer still has several 1020s and I have one ordered.
NSWRcars
Having received the correct Mashima 1020, I pushed on with re-motoring my Ixion C32.

As the flywheel was no longer needed, shaft at that end of the 1020 motor (opposite to worm gear) was first cut off. Wires to the OEM motor were unsoldered and the motor removed. Now to retrieve the worm gear from OEM motor for re-use. I had no trouble getting the flywheel off the shaft, but the worm would not budge. I decided to sacrifice the motor and then gripped the armature while pulling the worm. You guessed it, the armature came off the shaft, with the worm still stuck solid. I then gripped the (now bent) shaft and finally got the worm off without damaging it. Now to bore out the worm to fit a 1.5mm shaft. I have no lathe, and only 1.25 and 1.5mm drill bits, so a two-step bore. I chose to drill from the worm end only, and after completing the bore noticed that the drill had run slightly off-centre, yuck! Anyway, the worm was a reasonable fit to the 1020 shaft and I secured it with Loctite 243. Hopefully it is not too eccentric...

Next yuck was the chassis motor cradle which has to be ground out to accept the Mashima 1020. Unfortunately lugs that secure the motor retaining cage were ground too thin and broke away, leaving the cage retained at one end only. I examined this and scratched my head for about an hour, but concluded that a repair of the lugs was too difficult, so I have bonded the unsecured end of the motor to the chassis using silicon sealant, which is removable. The cage retains the motor at the worm end only, which ought to be adequate to counter driveline forces. Not entirely happy with this solution – perhaps a Mashima 1015 would have fitted with less grinding? Another minor issue to deal with is longer wires needed to reach the 1020.

I attempted to fit lead around the flywheel area, but clearances were really too tight, and I abandoned that idea. There is no way that a Mashima 1024 would fit this area. I already have the Ixion up from 157 grams to a respectable 178 grams, with lead in the sides of the firebox, over the gearbox, in the sandboxes, keeper plate, and under the dome and chimney.

Overall more work to re-motor than I anticipated, but hopefully the running improvement will be worth the effort.

Sponsored advertisement

  Oscar Train Controller

I've got a few Mashima motors stashed now for the future and an Ixion 32 but wanted to find gear replacements for the 1.5mm shafts before I did any other mods. Then summer happened, fire and 'rona, and kind of lost interest, until now. Despite my DC flag-waving loyalties I wanted to do some comparisons with DCC to see how these 3 pole motors handled it, plus have always been curious about DCC.  There's no videos or definitive answers yet but I need some DCC advice.  

Q. My problem is my Wombats wont coast to a stop. Rather they decelerate with the throttle and come to an abrupt halt before the throttle is turned to zero. The Ixion accelerates and decelerates as intended.  Has anyone had this issue or got any idea what's going on?

Here's my setup:
Digitrax Zephyr Extra.
3258 Digitrax DN132PS chip, OEM motor
3004 Digitrax DN132PS chip, Mashima 1020 motor
3063 DCC Concepts Zen Black, OEM motor

Some steps I took to remedy:
After initial playing round and not being able to fix this issue I started from scratch, resetting all decoders independently on a programming track using CV8=8, and reset the Zephyr. Once reset and reprogrammed all locos were left off any track to avoid conflicts.
All had CV 29 changed to 38.
4 digit loco addresses added
CV3=25, CV4=25
The Zephyr's braking was set to 2 but even at 15 didn't change anything. The Wombats would coast for a while, decelerate a bit then abrupt halt.

It doesn't make sense especially given that although the same Digitrax decoder is in the Ixion 32 as the remotored 30T, the later behaves the same as the unmodded 30T with Zen decoder.  I read instructions on the locos and as much as I could on the decoders and I can't work out what CV if any needs changing. I changed light settings on the tenders but no change to behaviour. I have no idea what else to do but it's clearly to do with the Wombats.
  Oscar Train Controller

Whilst trying to figure this out I thought perhaps something on the loco's circuit board is messing with the signal to decelerate slowly so I went about hardwiring the decoder to the 8 pin tender/loco plug. I haven't quite worked out all the lighting connections, the whole two of them, but I got the loco going, though without joy on the momentum front. I'll provide a schematic eventually when I work it all out.  What do the resistors do and what are the bits bridging them? Surface mount caps or resistors?  Removing them out of the equation didn't solve the issue.





So I'll plonk a video in here showing the difference between an Ixion C32 with the same decoder and settings coasting when throttle turned to zero compared to the Wombat 30t which stops abruptly.  This is 3004 with Mashima but it does the same on 3063 with OEM motor and regardless of tender circuit board in use or not. Makes me wonder if now I have to dig into the boiler and remove the lighting circuit board. There's very little experiences shared online publicly regarding DCC'ing this thing and by far most videos show the loco always in motion. Surely I'm not alone. I don't really want to open up the Ixion just yet but given their similarities I can only think that the cablight is the only difference electrically, kind of.



https://www.youtube.com/watch?v=--T9tO4e7tA
  Oscar Train Controller

The thought occurred to me after the last post that maybe it's not something to do with the Wombats but more to do with the Ixion. Most of my diesels are 21 pin socketed so I dug through some older Traino's to find an 8 pin setup. Enter 4201, new out of the box after at least 10yrs. Such a breeze putting a chip in this and man it was heavy at 515g. Anyway, programmed the same as the others, this time with a Zen.  The thing slowly accelerated away but unfortunately, came to an abrupt stop as throttle reduced to zero despite CV4 being given the same 25 value. So I don't know. All of a sudden the focus is on the Ixion. Why does it work properly and is this just a CV setting and user error. Any ideas from the DCCers?
  Oscar Train Controller

An update with a hallelujah moment finally.  I swapped decoders between the Ixion 32 and Traino 42.  The 42 now behaved faultlessly whilst the 32 now shutdown early whilst throttling down. It's not the Wombats, the Ixion nor the 42. It's one good decoder. But I swear I set all the same a few times yet two Zens and one Digitrax weren't working as they should. When in doubt reset the system again, though this time I took off the quick connects to the tracks and reset the Zephyr by pressing Prog, Switch, 39, CV-WR, then Exit. Screen blanked then lit up and then success. They all worked! It's only taken a couple of days.

And there was one casualty, I killed my first decoder.  That tender-to-loco plug that we've talked about before, it was a little loose in 3063.  So during the excitement I pushed it home with a little flat screwdriver not even thinking about track power being on thanks to being more familiar with DC, and I must have bridged a +ve and -ve contact for less than a second as a puff of smoked escaped from a Zen. Arrgh, lesson learned, but geez.

The point of all this was to test OEM motors with DCC v good DC and I just haven't got much to share. I will say one thing though, I have been searching randomnly since late last year for sources of Mashima alternatives and there is nothing that's 5 pole skew wound in this size at all.
  NSWRcars Assistant Commissioner

The point of all this was to test OEM motors with DCC v good DC and I just haven't got much to share. I will say one thing though, I have been searching randomnly since late last year for sources of Mashima alternatives and there is nothing that's 5 pole skew wound in this size at all.
Oscar
Thanks for the updates. Surely there must be demand for a decent motor of this size? The OEM ones are unsatisfactory.
I haven't touched my Ixion C32 since October last year when I posted the above. It is re-motored, but there is some finishing off required, and I haven't got around to testing it. There is more interesting work to do on the new layout!
  Aaron The Ghost of George Stephenson

Location: University of Adelaide SA
I've got a few Mashima motors stashed now for the future and an Ixion 32 but wanted to find gear replacements for the 1.5mm shafts before I did any other mods. Then summer happened, fire and 'rona, and kind of lost interest, until now. Despite my DC flag-waving loyalties I wanted to do some comparisons with DCC to see how these 3 pole motors handled it, plus have always been curious about DCC.  There's no videos or definitive answers yet but I need some DCC advice.  

Q. My problem is my Wombats wont coast to a stop. Rather they decelerate with the throttle and come to an abrupt halt before the throttle is turned to zero. The Ixion accelerates and decelerates as intended.  Has anyone had this issue or got any idea what's going on?

Here's my setup:
Digitrax Zephyr Extra.
3258 Digitrax DN132PS chip, OEM motor
3004 Digitrax DN132PS chip, Mashima 1020 motor
3063 DCC Concepts Zen Black, OEM motor

Some steps I took to remedy:
After initial playing round and not being able to fix this issue I started from scratch, resetting all decoders independently on a programming track using CV8=8, and reset the Zephyr. Once reset and reprogrammed all locos were left off any track to avoid conflicts.
All had CV 29 changed to 38.
4 digit loco addresses added
CV3=25, CV4=25
The Zephyr's braking was set to 2 but even at 15 didn't change anything. The Wombats would coast for a while, decelerate a bit then abrupt halt.

It doesn't make sense especially given that although the same Digitrax decoder is in the Ixion 32 as the remotored 30T, the later behaves the same as the unmodded 30T with Zen decoder.  I read instructions on the locos and as much as I could on the decoders and I can't work out what CV if any needs changing. I changed light settings on the tenders but no change to behaviour. I have no idea what else to do but it's clearly to do with the Wombats.
Oscar
I have not been reading this thread, but Allan has just alerted me to it.

What are CVs 2,5 and 6?

It will be a bit of work if you don’t have JMRI but try setting CVs 67 to 94 to these values in order: 1,1,1,1,2,2,3,4,5,6,7,9,11,14,17,21,26,32,40,49,60,74,91,112,137,169,204,254.

Then set CV29 to 54.

That should set your decoder up to follow what I use as a fairly standard curve with little ‘low end’ acceleration, giving you some hopefully decent deceleration performance as you come down in speed too.

If you can, see what happens if you take a decoder you are happy with the throttle response of, and put that in the model you are not happy with, if the model still performs badly, don’t discount the idea of a bind, or maybe a slow speed pickup issue or similar killing the decoder’s attempts to provide a decent slow speed stop.
  Oscar Train Controller

Hi Aaron, thanks for the input and thanks Allan. I thought about starting a new thread given the DCC based issues but as it's all tied into experiments with these locos that I hadn't finished I thought I'd add it here for reference. So, glad you fellas saw it.

From memory CV2 was at 1 or 2 and CVs 5 and 6 were all over the place in the end but I think around 100 to 120 and 50 to 80 respectively.  I changed them so many times when I couldn't work out the momentum issue but now the system is behaving after the reset I will enter the figures you quoted.  From what I understand looking at the Digitrax calculator for CV29, I see now 54 will give 28 speed steps capability, (honestly I didn't fully realise I was stuck in 3 step) and doing some further comparison say to the DCC Concepts example which is more lineal at the start, your numbers should provide an excellent opportunity for the motors to demonstrate take off and slow speed performance.  I look forward to trying this out and reporting back but will have to wait till later in the week when I've got time off.
  Oscar Train Controller

Whilst becoming consumed with trying to find an answer to better motor performance during startup and slow speed, I’ve had to ask myself if I’m taking this a bit too far. Then I view a bit of footage of a real 30 or 32 and plonk one of mine on a DC track trying to emulate their movement and immediately convince myself I’m not crazy and get back to trying to find solutions.

To that end I’ve gone a few routes both DC and DCC.  I’ll cover DCC developments first then DC in a following post.  Even though I was a staunch DC advocate I’m thinking of continuing down the DCC route partly due to my layout will be easier to wire rather than the relay logic I wanted to employ on a DC layout.  That’s a whole other discussion but knowing that there was a couple of videos showing smooth starts in DCC I wanted to test a few decoders to see if I could achieve the same.

DCC
Disclaimer: please note when reading the following and viewing the video I’ve approximately one month experience on DCC.

So I’ve got a Digitrax Zephyr DCS51 as command station, added a DT500 throttle plus UR90 to give a bit of walkaround ability.  After a day playing with CVs as Aaron suggested I promptly dug out an old laptop, bought a Digitrax PR4 for connectivity and loaded JMRI.  DecoderPro is amazing and I couldn’t be happier.

Six decoders were tested: NB, all 8 pin plug. I heard second lot of Wombats were 21pin??, I only have Wombats from first run.
-          Digitrax DN136PS
-          Digitrax DH126P
-          DCC Concepts Zen Black Mini
-          ESU LokPilot 5
-          NCE D13WP
-          TCS DP2X-UK

I could not copy the smoothness of a Mashima on DC but after hours spent tuning them and experimenting there were two clear standouts.

ESU LokPilot 5 and TCS DP2X-UK were the best performers and were also the most expensive at $55 and $63 respectively.  Both managed to make the OEM motor crawl and transition smoothly from forward to reverse and vice versa.  Both exhibited some jitters kicking the motor over but very minor. I was very impressed with how these two managed to make movement prototypical.  I liked how the ESU would be less influenced by initial throttle setting and would slowly accelerate following a high CV3 setting rather than max throttle setting. To make the TCS crawl it was more dependent on starting with low throttle inputs and gradually increasing.  Both took a while to setup and required resetting when I lost track of which CVs were changed.

The ESU’s eureka moment occurred after reading further into the downloadable instructions for the LokSound 5 as there is none for the LokPilot 5 that I could find. (AFAIK, same LokPilot and LokSound versions have the same bits that influence motor control).  Setting CV54 to 0 and using Program on the Main, activate Function 1 and the Wombat takes off at full speed for a second or so then stops.  Magically, all those CVs for Parameter K and I and EMF settings that I previously messed up were set correctly. I just had to play with acceleration and deceleration and that was it.  I think this chip is the pick of the crop.

The TCS would be my preferred choice if it wasn’t for a couple of issues. One was that reliance of slow throttle increments to transition from really slow start and really slow acceleration. I suppose it’s akin to accelerating hard in a car to slowly accelerating, so no real big deal.  The other and more important issue is the auto-adjusting BEMF which although responsible for outstanding slow speed operation, it caused the loco to slow or sometimes stop when nearing full speed then start up again. No connectivity issues or dirty track. Even with the full speed reduced to speed step 170 or so it may briefly get there then slow. If I wind the throttle from 99% to something less, it’ll jump to full speed again. I could be missing something here. I played with EMF cutoff and dither and I just couldn’t rid this behaviour though I did somehow limit it.  It was the easiest to install being non-wired direct fitment and the “UK” suffix refers to the pin orientation on the board of which there are three orientations.  This one fits the Wombat perfectly.

Digitrax DN136 and DH126.  The suffixes refer to “P”lug and “S”hort referring to a harness with plug and length of harness. They’re meant for N and HO scale respectively with the former capable of running small HO locos.  They’re both series 6 decoders and have 3 and 2 functions and came in at $35 and $38.  In a nutshell, I like these.  The DN was jittery though, not so much with the DH.  Neither could match the ESU or TCS for fine slow control but the DH version was acceptable.  Apart from that my only dislike is the acceleration setting only goes to 31.  All the others allow more freedom there and I made use of it. I’m liking the set and forget kind of approach to accelerating. It’s why my Gaugemasters have simulation/inertia.

DCC Concepts Zen Black MINI. A four function decoder at $37 was good value. I wanted this one to win to be honest.  My only complaint is the initial response from stationary to moving sees the loco jump.  As best I tried I couldn’t stop that from happening.  The other decoders above all showed some pause between changing directions whereas the Zen changed directions near immediately and jumped to the opposite direction. The acceleration and deceleration was excellent except for the initial start and stops were a little abrupt.  The Zen Black is not defined in DecoderPro yet so I may be missing a setting to improve starts and stops. The Zen has the best included instructions of all of them though you may still want to visit the website for more.  I find DCC Concepts has excellent in depth documentation on their site for a lot of products. Along with the TCS I’m going to have to peruse their websites a bit more.

NCE D13WP. $43. I could not get this decoder to work any better than on DC. It was kind of worse which surprised me.  With my little experience in DCC I have only heard good things about NCE and the D13 series of decoders.  From what I could tell I could only adjust torque compensation through kick strength and kick frequency, plus the usual speed steps and acceleration etc. I couldn’t get it to run a Wombat smoothly at all. However, I threw it into a Trainorama 42class with the same settings and found it performed completely normal. So whilst not suitable for the 3 pole motors it works well in modern 5 pole diesels.  Slightly on the downside with these is there’s no insulation/heatshrink around the decoder which all the others have.

Another disclaimer but regarding the vid. Despite an Ixion 32 appearing up first with the ESU, all decoders were tested on a Wombat 30. The ESU, TCS and DH126 were tested on the 32 as well and appeared to behave the same as the Wombat.



https://youtu.be/Fi0s80tu9OU
  FirstStopCentral Chief Train Controller

Interesting observation that I think is adding to the jump when starting.

It appears that the 30 motor is driving on the rear driving wheel and that due to 'slop' in the siderods, the rear wheel begins to rotate but effectively slips because all the other wheels are holding the loco where it is. So when the 'slop' is taken up and all the driving wheels begin rotating together, the rear wheel (and motor) have already begun speeding up from zero.

The 'slip' due to the 'slop' (sounds like a sun tan lotion commercial) is quite noticeable when reversing direction.

Paul
  Aaron The Ghost of George Stephenson

Location: University of Adelaide SA
I didn’t see reference to you having an ESU decoder, otherwise I would have mentioned the CV54 set up.

Also, I didn’t figure on you being a DC user.

ESU, TCS and to a slightly lesser extent, Lenz and Zimo almost certainly make the best decoders in the market I would usually buy them in that order, and I wouldn’t buy another brand.
  lkernan Deputy Commissioner

Location: Melbourne
ESU, TCS and to a slightly lesser extent, Lenz and Zimo almost certainly make the best decoders in the market I would usually buy them in that order, and I wouldn’t buy another brand.
Aaron

I'm pretty much in the same boat, although a lot of shops around here are pushing the cheap blue ones.  
I've thrown out enough of those to never go there again.
  NSWRcars Assistant Commissioner

Interesting observation that I think is adding to the jump when starting.

It appears that the 30 motor is driving on the rear driving wheel and that due to 'slop' in the siderods, the rear wheel begins to rotate but effectively slips because all the other wheels are holding the loco where it is. So when the 'slop' is taken up and all the driving wheels begin rotating together, the rear wheel (and motor) have already begun speeding up from zero.

The 'slip' due to the 'slop' (sounds like a sun tan lotion commercial) is quite noticeable when reversing direction.

Paul
FirstStopCentral
Not a good look. And this highlights the poor weight balance of these locos, centre of gravity is too far forward, less weight on the rear (driving) axle.
  Oscar Train Controller

Aaron, yeah the ESU, TCS and NCE decoders were bought just after your last post and I'd only just got hold of them a few days before posting the update. I planned at buying a Lenz Silver mainly due to youtuber "Everard Junction" praising them over cheap alternatives years ago. And I know they're popular in Europe but they're a bit thin on the ground in Aus, ie can't find any. But I do like the ESU.  Regarding DC, I don't know what it is about it. There's a fascination about analogue with mechanical levers and contacts plus relays tapping and solenoids clicking compared to screens and CVs and programming. However the automation aspect also interests and I could quite easily sit back and just watch as a 'puter runs the show.  I'll aim for some hybridisation between the two.

About that slop - It'd be interesting to see a difference if that slack was eliminated if all wheels were cogged like the Trainorama 32, instead of relying on the side rods.  However it would make other characteristics of the drivetrain difficult to include.  For example, I discovered the other day there's compensation built in to the middle wheel.  It's something I was surprised but happy to see.  The middle drive wheel can float a little above and below rail level and is supported by light coiled springs inserted into the cast chassis above the bushes of the centre drive wheel.  All wheels have some side play but only the middle has this suspension. This goes for the 30 and 32 but perhaps there's a little too much allowance there in the rods. I dunno, but it's a cool feature. The geartrain has negligible backlash. There's a photo of the springs in an upcoming post about motor mods where I disassemble a 30 a bit more.

Regarding the balance issue.  I must admit I haven't really paid close attention except to the weight till now. The CoG of the 30 appears to be between the middle and forward driving wheels and on the 32 it is on the leading edge of the middle wheel. It'd be great to see it over the middle more or better still, equidistant between front and rear drive wheels given that both locos rear wheels are set back more, the 32 especially. Given the light suspension of the middle wheel, when you have a closer look it does little to nothing for traction and all the work's done by the front and rear drive wheels IMHO. I'm almost positive I'm right here. Man this is another rabbit hole I'll have to dive into.....
  LaidlayM Chief Commissioner

Location: Research
I didn’t see reference to you having an ESU decoder, otherwise I would have mentioned the CV54 set up.

Also, I didn’t figure on you being a DC user.

ESU, TCS and to a slightly lesser extent, Lenz and Zimo almost certainly make the best decoders in the market I would usually buy them in that order, and I wouldn’t buy another brand.
Aaron
TCS as the cheapies, Lenz or ESU for even better quality.  I had one NCE decoder and gave it away.  The only other decoders I have are QSI and they are troublesome.

Mark in Research
  LaidlayM Chief Commissioner

Location: Research
I didn’t see reference to you having an ESU decoder, otherwise I would have mentioned the CV54 set up.

Also, I didn’t figure on you being a DC user.

ESU, TCS and to a slightly lesser extent, Lenz and Zimo almost certainly make the best decoders in the market I would usually buy them in that order, and I wouldn’t buy another brand.
TCS as the cheapies, Lenz or ESU for even better quality.  I had one NCE decoder and gave it away.  The only other decoders I have are QSI and they are troublesome.

Mark in Research
  Oscar Train Controller

Mark I was going to offer my NCE up for grabs too, until I tried it on the diesel so I'll keep it now, but wont try more.

More on motor mods and replacements

In this post I successfully try out a different motor, attempt a skewed rewind and search for worms.


1.5mm bored Worms

Mashimas are thin on the ground now.  Hollywood Foundry’s out, the ebayer I got mine from are out but there’s still 1020s and 1024s. If you have got a Mashima you’ve still got the issue of accurately boring or reaming out the OEM worm to 1.5mm otherwise they are ideal replacements performance wise.  I thought if only I could find a suitable worm I wouldn’t have to try rebore anything. I went on a search and found the world of gear cutting is more complex than I thought even on small scales. Basically, if I can use that term, the defining measurements of the Wombat’s worm are as follows:



Right Hand
Module = 0.4mm
Outside Diameter = 5.3mm
Pitch Diameter = 4.5
Helix Angle = 5.08deg
Bore = 1.0mm
Length = 9.0mm



The only commercially off the shelf worm I could find to nearly match the above and fit a Mashima shaft was from Northwest Short Line who sells a steel worm with the following measurements:



Right Hand
Module = 0.4mm
Outside Diameter = 6.0mm
Pitch Diameter = 5.2
Helix Angle = 4.40deg
Bore = 1.5mm
Length = 9.5mm



Unfortunately it’s not going to work.  Despite the larger diameter and different helix angle the worm meshed ok with the worm gear when not fitted but sat proud.  The housing is too small and the entrance blocked by a bulkhead that limits motor movement, similar to the one I cut away at the other end to fit a longer motor.  I could cut that away and cut the gear housing in half but that’s all a bit extreme. Furthermore, the larger diameter and the mesh not being perfect means the motor has to sit higher which means the cage can’t be used and the body might not fit.

So if you’re using a Mashima and don’t want to or can’t rebore accurately, search for the top measurements but with 1.5mm bore and please let us know if you find any, I couldn’t.

This is just a pic showing a 30 nearly stripped right back so I could determine above measurements plus close up of worm and spur gears. You may notice the suspension springs I mentioned previously in the middle drive wheels hornblocks/bearing housing.  Note the gearing is 50:1







Skewed Rewind

Sumida Crossing is a great read that I’ve referred back to a number of times over the past year or so and in particular the DC motor page linked shows examples of N scale motors, skewed 3 pole and 5 pole motors plus straight winds. Haven't found specs available on any to see if any of those motors could be a drop in replacement but I got the idea to maybe try and skew one of our OEM motors.

They are surprisingly easy to skew.  I removed the windings on one older motor, trying to count the number of winds when after about 150 the rotor dropped and it was like a cotton spool unwinding.  The remaining two attempts of counting winds were unsuccessful through breakages and loss of concentration.  I used a thin screwdriver’s shaft and twisted between the slot’s gap and got instant skew.  



I’ve no idea about which way to skew or wind and what sort of wire, thin or thick. A bit of misinformation was available claiming thick was better but then it’s all about resistance and long thin winds which means more resistance and less current.  Some 0.2mm enamelled copper I had which I thought would be thin enough is grossly oversize.  The OEM winding is I think 0.06mm to 0.10mm. Big variance but dodgy callipers and no micrometer led to a bit of guess work.  Regardless I used the 0.2mm as it was all I had and got about 40-50 turns, I forget.  




My first attempt caused a bit of humming and finally some rotation but then it quit.  After repeated attempts I realised the Gaugemaster was cutting out. I waited some time then ramped it up and before full speed was reached I heard a little crackle and then it stalled. On disassembly I noted the brushes/contacts on one pole had vaporised.  Windings were ok though.  I tried another cap end with new brushes but this time connected to analogue volt and amp meters. To my shock during a quick turn of the controller the ammeter shot to near two amps. No wonder the fine brushes cooked at near full speed. So disassemble and wait for some thinner wire to arrive

Note the missing/shorter brushes closer to camera





The new enamelled wire arrived in 0.112 and 0.085mm from “physicalsciences” on ebay in Canberra. I removed the wire from that rewind and skewed as far as I could.



I got about 170-180 winds with about 150 winds on the last pole/slot owing to space constraint.  It’s a shame cos I was getting good at it with good tension and tight wind but the first winding in particular was loose. It was obvious it was going to be out of balance but I had to try it. It started turning at a lower indication on the gaugemaster, between 10 and 20%. Amps rose to approx 300mA before dropping as the motor turned. It vibrated ridiculously as I sped up and as I hit max power a couple of times the amps maxed out at 300-400mA and voltage near 15v before failure.  Just a dead stop.  On disassembly I noted the brushes were fine but some loose winding clearly was thrown outwards due to centripetal force then broke, lost continuity and jammed between the magnets.  I was pretty chuffed regardless. It took 3 hours to rewind and 10 seconds to fail but meh, fun stuff.




But rather than rewind again I had a different spare OEM motor (I bought a few from Casula Hobbies to experiment on) with bent shafts after being a bit rough when removing worm and flywheel.  I inserted the shaft of the thin screwdriver between the outer edges of the slots then twisted and got 3 pole skew with factory wind. I reassembled and tested and got movement. The bent shafts prevented me from testing this further other than observing it on the bench.  



So I grabbed another spare which is the one above, skewed the laminations and tested in the stripped loco.  Results were pretty good. Windings weren’t damaged and in the video below you’ll see a comparison of OEM first with forward and reverse on manual control then with simulation followed by the same test for the OEM skewed motor.  NB there is a slight bind in the mechanism of the stripped back test loco which I didn’t investigate when I had the loco apart but it’s there in some older vids too so it's not the motor causing those little stalls.  The motor to final drive ratio is 50:1 so for at least three quarters of wheel rotation you’re seeing the motor rotate evenly a number of times displaying reduced cogging.  The end result sees the skew modded factory wound motor starting off sooner and smoother than the unaltered motor and generally operating slower though at mid to full speed both locos were travelling at the same speed.  The skewed motor stopped before the OEM motor on simulation.  Not as good as the Mashimas but it’s a possible mod to see better performance on DC.  I had the idea of adding 53grams of lead folded on top (equals the weight of the body shell), to make things fair.  No real change in results.



https://youtu.be/l6LX_dcVMVM





A No Mod Needed Drop in Replacement Motor, that does works better.


There’s three potential candidates and all are 3 pole straight wound.

               - N-Drive 1015 from the UK available on ebay gets good reviews on RMWeb,
               - High Level Kits have got a 1015 and 1020 size
               - NorthWest Short Line have the 1215 which is 10mm on the flat.  So same size as the 1015.

All have 1.0mm shafts.



High Level Kits and NWSL have been good to converse with via email.  All three candidates say in one way or another that they’re worthy Mashima replacements. Owing to an earlier communication with NWSL and months ago promising here I’d buy their motor next pay day, plus wanting to buy those worms mentioned above, I decided to buy a 1215.




NWSL were really good, items arrived in 12 days and about $18USD shipping. I have to say I was a bit sceptical when I looked at the 1215.  It’s so similar to the OEM replacements.  I put on a flywheel and turned the 1215 by hand and compared with other OEM motors.  Amazingly, very little to no cogging as was suggested to me. I had to open this sucker and see what was inside.  Again, so similar. Here compared with one of the OEM rotors I skewed.




I think the magnets may be different, I’ve a dim thought of noticing that but never took a photo. The windings may be different, I'm not sure.  Regardless, it really is better. There’s a short 1min video comparing older vids of the Mashima 1015, OEM, this NWSL motor and remembering how awsome the Mashima 1024 motors are on DC and using a Gaugemaster controller.  I’m impressed with the NWSL alternative.  When all is said and done I may just convert all three of my locos to this motor. It’s better than OEM on DC and I’m sure will be better on DCC too.



https://youtu.be/XAgD8dwZdL0
  NSWRcars Assistant Commissioner

NWSL were really good, items arrived in 12 days and about $18USD shipping. I have to say I was a bit sceptical when I looked at the 1215.  It’s so similar to the OEM replacements.  I put on a flywheel and turned the 1215 by hand and compared with other OEM motors.  Amazingly, very little to no cogging as was suggested to me. I had to open this sucker and see what was inside.  Again, so similar. Here compared with one of the OEM rotors I skewed.

I think the magnets may be different, I’ve a dim thought of noticing that but never took a photo. The windings may be different, I'm not sure.  Regardless, it really is better. There’s a short 1min video comparing older vids of the Mashima 1015, OEM, this NWSL motor and remembering how awsome the Mashima 1024 motors are on DC and using a Gaugemaster controller.  I’m impressed with the NWSL alternative.  When all is said and done I may just convert all three of my locos to this motor. It’s better than OEM on DC and I’m sure will be better on DCC too.
Oscar
Just to clarify, does the the NWSL 1215 have the same diameter shaft as the OEM? Did you use the OEM worm?
  NSWRcars Assistant Commissioner

The centre driving axle suspension in conjunction with weight imbalance and narrow 88 wheel treads reduces the stability of these locos through pointwork. It is possible to improve the balance (without shifting the Centre of Gravity) by fitting a stiffer spring to the front pony truck, but you lose valuable tractive weight, transferred to the pony truck. I've made this mod to my Ixion C32 and it improves stability and tracking of the loco considerably, and virtually eliminates that ugly dropping of the front drivers into Peco point frogs. I got most of the lost traction back by packing lead weight into every nook and cranny. The Wombat C30T has less scope for adding weight towards the rear.
It's unfortunate, but there are so many "features" of these locos that make them cranky, that I often wonder if they are worth persevering with. I've spent hours on my Ixion C32 getting it up to scratch, time that could have been much better spent. And I still haven't completed the re-motoring job I started last year... lost interest. Maybe next week?
  Oscar Train Controller

Just to clarify, does the the NWSL 1215 have the same diameter shaft as the OEM? Did you use the OEM worm?
NSWRcars
Yes, absolutely.  Shaft is 1.0mm, the stock flywheel and worm will be both interference fit requiring pressing on just like the OEM motor.  Other major measurements of the motor match OEM and mashima 1015. ie; 10mm on the flat, 12mm on the round and 15mm long.  All three are the same but the NWSL has the 1.0mm shaft like OEM.  Similarly removal requires gear puller or tapping out like I do with a 1.0mm bright brad nail and light hammer plus the use of a vice and something like cardstock to protect the worm and flywheel from the vice.   If I was just starting to think about a change now, and didn't have other options I'd be happy to go with NWSL. Also, since the mashima 1020 was better than the 1015 I'm going to get a NWSL 1220 to match the 1020 you and I have.

Interesting about the front bogie spring.  I'd love to cast lead into every space available but I dunno.  I deleted a bit about peco points and the 88 wheels. I really like the finer wheels but I'm going to have to fill in frog flangeways to minimise the drop.  I actually edited videos to avoid showing the drop and how obvious it can look and spoil the visual of a loco traversing crossings and point work.
  a6et Minister for Railways

Just to clarify, does the the NWSL 1215 have the same diameter shaft as the OEM? Did you use the OEM worm?
Yes, absolutely.  Shaft is 1.0mm, the stock flywheel and worm will be both interference fit requiring pressing on just like the OEM motor.  Other major measurements of the motor match OEM and mashima 1015. ie; 10mm on the flat, 12mm on the round and 15mm long.  All three are the same but the NWSL has the 1.0mm shaft like OEM.  Similarly removal requires gear puller or tapping out like I do with a 1.0mm bright brad nail and light hammer plus the use of a vice and something like cardstock to protect the worm and flywheel from the vice.   If I was just starting to think about a change now, and didn't have other options I'd be happy to go with NWSL. Also, since the mashima 1020 was better than the 1015 I'm going to get a NWSL 1220 to match the 1020 you and I have.

Interesting about the front bogie spring.  I'd love to cast lead into every space available but I dunno.  I deleted a bit about peco points and the 88 wheels. I really like the finer wheels but I'm going to have to fill in frog flangeways to minimise the drop.  I actually edited videos to avoid showing the drop and how obvious it can look and spoil the visual of a loco traversing crossings and point work.
Oscar
An issue I find with both the Wombat 30T & the Ixion 32, (unless design/production changes took place since I saw the preproduction models) is that rather than metal boilers they only had an internal metal sleeve in the boiler, basically from smokebox end to front of firebox, the primary reason being to have the bit extra weight over the drivers. Putting a heavier spring in the leading bogie will create more problems as they tend to lift the front of the model and the leading driving wheel is often not making contact with the track.

The original Austrains 36cl were notorious in that area but remove bogie spring and put a couple of small strips of lead flashing on top bogie made a huge difference except the bogie was a bit more floppy but much better. The next run of them had a different frame, that was also used on the 35cl which precluded doing that same work.

While my model still sits in a box, with only a cursory look, I would suggest checking if the dome, and the sandboxes are plastic or metal, if plastic, it may help if both items could be replaced with metal casting replacements, while only small in extra weight it may help.

What does not help these two models is the 88 wheel sets, as they are less likely to have constant contact on track around curves and possible  grades, as indicated with the peco points.  Most model importers no longer have the 110 wheels but Trainorama now use a 100 wheel, they are mid point sizes and look quite ok and noticeably so from 110 wheels.

I believe the key is to try and balance the actual weight from the motor end over the 3rd wheels and the front driver, maybe a small amount extra over the leading driver.  With the body off, try measuring the distances from top of motor and flywheel, and how much available space between it and the inside boiler bottom, try some lead flashing in there to add weight. To glue to the inside use some roof and gutter clear silicone it is strong and will hold the flashing in place but is also removeable.
  Oscar Train Controller

I've been under the impression they were metal boilers, albeit thin. I'm sure that's the case, I will check again next chance I get.

Brass dome and funnels have been available from Ray Pilgrim's Shapeways shop for the Ixion 32 and Wombat 30T but apart from googling them specifically I'm not sure if they're still available when I look through the Signals Branch listings.  Only 3d printed plastic replacements are appearing in the shop.  I'm sure he had brass sandboxes for the 32 as well. I said months and months ago the funnel in particular didn't bother me but it does. I'd like to have the bolt detail and non segmented look.  Anyway, at least the plastic ones are still available.  

100 or 99 like SEM rollingstock replacements should be a standard these days I reckon. It's a nice balance between a finer tread and wide enough to ride through crossings.
  NSWRcars Assistant Commissioner

Brass dome and funnels have been available from Ray Pilgrim's Shapeways shop for the Ixion 32 and Wombat 30T but apart from googling them specifically I'm not sure if they're still available when I look through the Signals Branch listings.  Only 3d printed plastic replacements are appearing in the shop.  I'm sure he had brass sandboxes for the 32 as well. I said months and months ago the funnel in particular didn't bother me but it does. I'd like to have the bolt detail and non segmented look.  Anyway, at least the plastic ones are still available.
Oscar
You can buy a sheet of 1.5mm Lead flashing from your favourite warehouse, it weighs 0.015g/sqmm.
Cut it up and pack it into your loco. This is what I managed with the Ixion C32, which starts out at 157g:

C32 Firebox sides – 22x11x2 = 7g

C32 Gearbox wrapper – 14x28 = 5g
C32 Motor – 9x6 = 1g
C32 Sandbox x 2 - 9x3x2 = 2g
C32 Keeper plate 6x18 + 6x9 = 2g
C32 dome = 3g
C32 chimney = 2g
C32 Ixion modified 157 + 22 = 179g

Note the chimney and dome on my C32 are the Ray Pilgrim plastic ones. They will accept a lead slug. However, the main problem with these locos is weight is already too far forward. You should only add weight forward of the front driving wheel if you fit a stiffer spring to the pony truck. I did this to my C32 using a Century S-1093 spring, and I can assure you that it greatly improves the balance and tracking of the loco. The brass sandboxes that others have mentioned are actually a later, larger type, and some may like to fit these, however I'm told it is not a simple job and requires cutting the footplate. I was happy to pack a little lead into the OEM sandboxes.


Now, I don't have a Wombat C30T, but just from looking at photos that Oscar and others have posted, I reckon this is what could be done. Estimates only. Note that the C30T has a shorter firebox than the C32 and likely an even worse weight imbalance. For this reason I think you would have to consider adding lead flashing under the cab roof, and I don't think adding anything to the front of the loco (such as chimney or smokebox) would be a good idea, unless you are prepared to experiment with pony truck springs.

C30T Firebox sides – 18x11x2 = 6g
C30T Cab roof    23W x 21L = 7g
C30T Gearbox wrapper – 14x28 = 5g
C30T Motor – 9x6 = 1g
C30T Keeper plate 6x18 + 6x9 = 2g
C30T Wombat modified 150 + 21 = 171g

Go to it, fellows!
  a6et Minister for Railways

I've been under the impression they were metal boilers, albeit thin. I'm sure that's the case, I will check again next chance I get.

Brass dome and funnels have been available from Ray Pilgrim's Shapeways shop for the Ixion 32 and Wombat 30T but apart from googling them specifically I'm not sure if they're still available when I look through the Signals Branch listings.  Only 3d printed plastic replacements are appearing in the shop.  I'm sure he had brass sandboxes for the 32 as well. I said months and months ago the funnel in particular didn't bother me but it does. I'd like to have the bolt detail and non segmented look.  Anyway, at least the plastic ones are still available.  

100 or 99 like SEM rollingstock replacements should be a standard these days I reckon. It's a nice balance between a finer tread and wide enough to ride through crossings.
Oscar
Oscar, not sure but kerroby hobbies http://kerrobymodels.com/ho-scale.html  make parts and mostly in white metal, they have domes and funnels, sometimes the parts are a bit sus with poor molding.

The options are Ray Pilgrims domes and funnels, but not sure if they are solid moldings or hollow, if the dome is solid & I would say the Wombat one isn't (guessing) then perhaps Ray's one would be heavier maybe not by much and therefore better, and they are better as far as the correctness of them.  The option though in using them is perhaps to drill a 1/4" hole up to halfway and add lead into the cavity.

Getting the balance right with wheels and other items is crucial.

Once I get my 30T out and ready, I have to take the tail rods off the cylinders as they were not on superheated loco's would also liked to have seen the slide bar covers produced, at least one can get the Jacks from Steven Ottoways shop.
  a6et Minister for Railways

Brass dome and funnels have been available from Ray Pilgrim's Shapeways shop for the Ixion 32 and Wombat 30T but apart from googling them specifically I'm not sure if they're still available when I look through the Signals Branch listings.  Only 3d printed plastic replacements are appearing in the shop.  I'm sure he had brass sandboxes for the 32 as well. I said months and months ago the funnel in particular didn't bother me but it does. I'd like to have the bolt detail and non segmented look.  Anyway, at least the plastic ones are still available.
You can buy a sheet of 1.5mm Lead flashing from your favourite warehouse, it weighs 0.015g/sqmm.
Cut it up and pack it into your loco. This is what I managed with the Ixion C32, which starts out at 157g:

C32 Firebox sides – 22x11x2 = 7g

C32 Gearbox wrapper – 14x28 = 5g
C32 Motor – 9x6 = 1g
C32 Sandbox x 2 - 9x3x2 = 2g
C32 Keeper plate 6x18 + 6x9 = 2g
C32 dome = 3g
C32 chimney = 2g
C32 Ixion modified 157 + 22 = 179g

Note the chimney and dome on my C32 are the Ray Pilgrim plastic ones. They will accept a lead slug. However, the main problem with these locos is weight is already too far forward. You should only add weight forward of the front driving wheel if you fit a stiffer spring to the pony truck. I did this to my C32 using a Century S-1093 spring, and I can assure you that it greatly improves the balance and tracking of the loco. The brass sandboxes that others have mentioned are actually a later, larger type, and some may like to fit these, however I'm told it is not a simple job and requires cutting the footplate. I was happy to pack a little lead into the OEM sandboxes.


Now, I don't have a Wombat C30T, but just from looking at photos that Oscar and others have posted, I reckon this is what could be done. Estimates only. Note that the C30T has a shorter firebox than the C32 and likely an even worse weight imbalance. For this reason I think you would have to consider adding lead flashing under the cab roof, and I don't think adding anything to the front of the loco (such as chimney or smokebox) would be a good idea, unless you are prepared to experiment with pony truck springs.

C30T Firebox sides – 18x11x2 = 6g
C30T Cab roof    23W x 21L = 7g
C30T Gearbox wrapper – 14x28 = 5g
C30T Motor – 9x6 = 1g
C30T Keeper plate 6x18 + 6x9 = 2g
C30T Wombat modified 150 + 21 = 171g

Go to it, fellows!
NSWRcars
The weight to inside cab roof is a good one I forgot, as I put it in my 35 & 36cl models, there can be some also placed on them in front of the seats. On the 30T a small piece on both seats and just in front of the firebox, so not visible would also help.

Sponsored advertisement

Display from: