Heljan Churchward Mogul in 7mm scale

[GWR57xx]

As it will be a while before I can get a replacement nylon gear I thought I might as well try something else.

I took the Loksound decoder off and put the supplied DC blanking plate back on and tried it on DC again. Here's the result:

 

Not very good at very slow speed but ok at medium to high.

 

Next I put the Loksound back in and ran it whatever motor parameters were left after previous experiments:

 

Bit lumpy, and there's still that sticking point at one point in the wheel revolution which I didn't notice as much on DC. I suppose it might be an interaction between the motor sticking slightly which changes the back-EMF sensed by the decoder which then over-compensates. Does that make sense?

 

Then I went back to the original motor parameters as supplied with the sound project, which are the same as the factory defaults described in the ESU documentation for the decoder:

 

To my mind that looks much better at low speed but much noisier and rougher at medium speed.

These all show starting from stationary, gradual increase up to speed step 48 (of 128) then coasting to a halt, forward & reverse.

 

Next I ran the Loksound auto-tune feature:

 

Doesn't look much on the rolling road - looks much more dramatic on a piece of track when the loco shoots off at top speed!

 

And this is what it looks like with the motor parameters as set by the auto-tune:

 

My impression is that the auto-tuned parameters are worse than the defaults at slow speed but better at medium speed. Much room for improvement I think.

 

For me the slow speed running is far more important because the engine will spend all its time "on shed", so no express running for this one.

 

However, I think the damaged tooth may not be too much of an issue now that I removed the "lump" (leaving a gap in the tooth), but as @norman pointed out above the motor gear will only mesh with the ends of the nylon gear, not the middle (until it's worn down a lot!). We'll see...

 

More playing testing to do...

 

[Stringfingerling]

I spent an hour or two tweaking CV's on my Mogul yesterday (Loksound V5L).  Most of the changes were to do with acceleration and deceleration within the practical constraints of my layout.  I needed the loco to be capable of reaching a plausible running speed by the time it exits the tunnel mouth at the entrance to the storage table, and to be able to slow down in time for a stop at the platform.  I think I've achieved those goals now, though it meant resetting the wheel/chuff synchronisation.  I fiddled around quite a bit with CV54 (Back EMF - I think) but ended up doing the procedure that lets it set itself;  set CV54 to zero, press Function 1 and let the loco shoot off and then stop.  It ended up on a value of 43 this time.
I notice your rolling road not supporting the centre drivers:  when I was running my loco in, I was a bit concerned that the centre drivers would be able to oscillate a bit because they have a degree of springing.  I don't have  a rolling road, and ended up with the loco propped up a bit at each end and ran it in at a very gentle speed to avoid the centre drivers jumping up and down too violently.
The only slight fly in the ointment now is, as I 've said in my earlier posts, I think the loco is susceptible to dirty track.  It doesn't stall but it will jerk forwards a bit occasionally on pointwork that causes no problems for other locos.  My layout is in a shed and I think the very damp air in the evenings, before the shed has had a chance to warm up may be part of the problem.   I've been very interested to read of people using graphite sticks to give a very slight film on their track and achieve very reliable running, but I am a bit wary about potential damage to the vulnerable nylon gears on this loco.
 

[7007GreatWestern]

1 hour ago, Stringfingerling said:

I spent an hour or two tweaking CV's on my Mogul yesterday (Loksound V5L).  Most of the changes were to do with acceleration and deceleration within the practical constraints of my layout.  I needed the loco to be capable of reaching a plausible running speed by the time it exits the tunnel mouth at the entrance to the storage table, and to be able to slow down in time for a stop at the platform. 
 

 

I fiddled with CVs 5 & 6 in attempt to give me a very high speed curve for exactly the same reason as you - to try to get the loco from a standing start, up to a reasonable speed and then back to a stand in a very short distance. Most of my locos (OO) use Zimo decoders and I have always been able to achieve this using CVs 5 & 6 alone. However, I'm using ESU Lokpilots in my Heljan 43xx and 61xx and I'm finding my NCE Powercab can't change these values on the ESU decoders! I think I read on another forum this is a 'bug' on some ESU decoders. I did find a 'workaround' however - I simply switched from 128 to 28 speed steps. Now my locos accelerate far more quickly! There is a button on the NCE that switches the appropriate CV in the decoder to achieve this. Because it is an attribute of the decoder, not the controller, the loco 'remembers' that it is to run in 28 step mode when power is removed.

[GWR57xx]

I’ve been messing about with CVs all day as well and have failed to get the result I want, which I am sure is possible.

Trouble is the ESU manual is very superficial in its descriptions of CVs and leaves more questions unanswered than not.

I thought it was CVs 3 & 4 that controlled acceleration and deceleration? As in the values in those CVs set the time from standstill to max speed (CV3) and from max speed to standstill (CV4)?

CV5 is max speed, CV6 is obsolete on the V5 ( which always uses the 28 step speed table) but used to be mid-point for a “3 point speed line” on V4 and earlier.

At least, that’s how I understand it, but I think the manual probably lost quite a lot in translation from the native German.

 

[Stringfingerling]

27 minutes ago, GWR57xx said:

I’ve been messing about with CVs all day as well and have failed to get the result I want, which I am sure is possible.

Trouble is the ESU manual is very superficial in its descriptions of CVs and leaves more questions unanswered than not.

I thought it was CVs 3 & 4 that controlled acceleration and deceleration? As in the values in those CVs set the time from standstill to max speed (CV3) and from max speed to standstill (CV4)?

CV5 is max speed, CV6 is obsolete on the V5 ( which always uses the 28 step speed table) but used to be mid-point for a “3 point speed line” on V4 and earlier.

At least, that’s how I understand it, but I think the manual probably lost quite a lot in translation from the native German.

 

I agree, some of the manual is in weakly translated English, and, yes, I think unless you're a serious programmer you can't do anything with CV 6, but it strikes me that you could use the facility to set each of 28 steps on the speed curve individually , if you can spend the time;

Between speed step 1 and the maximum speed, LokSound 5 decoders will superimpose the speed using a speed curve. The speed curve hast 28 values (Cv 67 to 94). The CVs 67 and 94 are permanently set to the values of 1 respectively 255 in this decoder. The intermediate values can be distributed at will. The speed curve can not be switched off. CV 2 and CV 5 define a scale factor around which the points of the speed curve are scaled. Have a look at the speed curve in Fig. 26. The last entry (CV 94) is 255. This represents full speed. If you wish to reduce the maximum speed you only have to reduce the value of CV 5. The decoder computes (squeezes) the speed curve in such a way that the shape of the speed curve remains despite the lower maximum speed. The same is true for the first entry. The speed curve will be raised and scaled subject to the value in CV 2.

Setting each CV from 67 to 94, you should be able to define any speed curve you like, in spite of not being able to set a mid-point.

I seem to have reached quite an acceptable set of values for everything now; I can drive the engine more or less as I would like to, even though that means using much higher numbers from the 128 speed steps than I am used to for other locos.  If I find the time, I might look at redefining the speed curve by setting the steps from 68 - 93 one at a time.  It would take a long time though because of all the trial and error along the way.

If it's of any interest, I've experimented with a graphite 6B pencil at the places where the loco was jerking, and I think it's much better!  I took note of the warning elsewhere on RMWeb to use only a minimal film of graphite (" if you can see it, you've used too much") but the Mogul is now running consistently smoothly through my pointwork.  I'll post a video sometime in the next few days.
 

[GWR57xx]

1 hour ago, Stringfingerling said:

Setting each CV from 67 to 94, you should be able to define any speed curve you like, in spite of not being able to set a mid-point.

If I find the time, I might look at redefining the speed curve by setting the steps from 68 - 93 one at a time.  It would take a long time though because of all the trial and error along the way.
 

 

Please to hear you’ve sorted your problem with it.

I found trying to change CVs from the command station a bit of a challenge so decided to invest in a Sprog 3 using DecoderPro. Much easier and highly recommended. It also provides a basic throttle control which is all I need to drive locos on the rolling road ( aka programming track).
 

I think you’ll find that the author of your sound project will have put suitable values into the 28 speed curve CVs. Mine did and I’m happy with the smooth curve which starts slow and gives a lot of control over the slow steps, which is more important to me.

[GWR57xx]

While having the loco apart I noticed another example of poor design: the motor is a circular can which is retained in it’s housing by circular clamps. The clamps exert very little pressure on the ridges of the motor so there is virtually nothing to prevent the motor turning in the mount. Using gentle pressure from my finger I can very easily push the motor round by several degrees.

If the wheels and motion hit a rough patch or start to stick, if it is easier for the motor to rotate rather than the drive shaft then it will. This will obviously confuse the decoder which will try to compensate for the change in bemf it sees.

This again would have been easy to prevent at the design stage by simply having a flat on the motor ridge and using a flat clamp to engage with it instead of the silly (and more complicated) half-arched clamps we’ve got.

To prevent the motor turning I’ve applied some double-sided tape to stick it into its mount.

It’s a simple fix, but come on Heljan this really shouldn’t be necessary.

I look at some of the videos of superb slow running of people’s new Dapol OO Moguls out of the box and am starting to think I made a mistake deciding on O gauge .

On the plus side though, my Dapol and Minerva Panniers run beautifully . Straight out of the box, no messing about.

[Hal Nail]

1 hour ago, GWR57xx said:

I look at some of the videos of superb slow running of people’s new Dapol OO Moguls out of the box and am starting to think I made a mistake deciding on O gauge .

 

Surely just looking at your 7mm one affirms that decision, even if it doesn't move at all!

 

joking aside they, are all commenting about the running anyway - its gearing is hopeless apparently.

Edited December 31, 2020 by Hal Nail

[GWR57xx]

34 minutes ago, Hal Nail said:

Surely just looking at your 7mm one affirms that decision, even if it doesn't move at all!

 

Agreed it’s beautiful to look at and the level of detail is superb (but fragile), but if it doesn’t run smoothly it will just become an expensive shelf queen. 
Great design of the exterior (thanks mostly to Churchward & the GWR), poor design of the mechanism.

While we’re handing out awards, good packaging, poor customer support. Trying to order spares is difficult because all that is provided is a blurry exploded parts diagram with no parts descriptions (even on the Gaugemaster website, the official spares distributor I believe, who don’t currently list any spares for the 43xx).

[Hal Nail]

1 minute ago, GWR57xx said:

(even on the Gaugemaster website, the official spares distributor I believe, who don’t currently list any spares for the 43xx).

i got an e mail saying they have suspended sales of spares while they get all parts properly listed so you can order online.

I'd imagine it proved to be ludicrously time consuming for the minimal value involved.

[GWR57xx]

I’ve sent an enquiry to Ultrascale to see if they do a suitable replacement, but they are quoting 5 months plus delivery timescales.

Anyone know of another source of good model gears that would be suitable?

[GWR57xx]

Superb service from Ultrascale, who replied very swiftly to my enquiry.

 

However, in order to supply the information Ultrascale require, I needed to dismantle the gearbox from the engine (which I would have to do anyway to replace the damaged gear).

 

That's the next problem - how to get the gearbox out?

 

The photo shows two screw holes (in red) which hold the gearbox in, but removing them sadly only allows the gearbox to be raised about 5mm before something obstructs it coming any further:

 

Apart from the two screws holding the "firebox flicker" led in place (undone in the photo to gain access to the others), the only other screws visible from the top are the three circled in lime green. 

 

Have any of you managed to remove the gearbox?

I'd like to be sure before I go any further - don't want to end up with a worthless bag of bits that I can't put back together again!

 

Heljan don't seem to have given much thought to maintenance and repair of these engines. If they're going to design a sacrificial weak point (the nylon gear) then it ought to be easy to replace. This could easily have been so much better.

 

[Hal Nail]

52 minutes ago, GWR57xx said:

Heljan don't seem to have given much thought to maintenance and repair of these engines.

I'm getting deja vue from another Heljan thread here. You should put it back together and send it back flagging the broken gear. The fact you have worked on it isnt necessarily a problem (as discussed at length elsewhere).

 

Their approach is to factor in a number if duds and just replace them.

 

Eother way you've nothing to lose ringing Rails first.

[GWR57xx]

"By God, Sir, I've lost my leg!" 

"By God, Sir, so you have!"

 

 

 

 

[7007GreatWestern]

On 31/12/2020 at 11:59, GWR57xx said:

I’ve sent an enquiry to Ultrascale to see if they do a suitable replacement, but they are quoting 5 months plus delivery timescales.

Anyone know of another source of good model gears that would be suitable?

Hello 57xx,

As you may know, we have 'been here before' with Heljan, as recently as two years ago. In fact there is at least one RMWeb thread dedicated to the issue of Heljan's gears and the difficulty in sourcing spares:-

 

It seems a veritable 'cottage industry' of aftermarket Heljan gears sprang up, coordinated by Jim Snowdon (sometimes of this Parish) and manufactured by Ultrascale:-

 

 

Heljan have suffered considerable 'repetitional damage' from this in the past and as recently as March this year they made public statement on their Facebook seeking to allay customers' fears:-

 

https://www.facebook.com/heljanas/posts/Heljan-spares-gaugemasterwere-delighted-to-report-that-our-spare-parts-service-i/3034541133275399/

 

Buoyed by their reassurances I bought a Heljan 43xx this October, confident that Heljan's problems with their nylon gears were behind them and that satisfactory after-sales support was now in place. Now admittedly the nature of the gear failure is different here to the previous incidents and we are not talking about a recurrence of the same problem. I am however watching Heljan's reaction to this with interest a some concern.

 

FWIW, I would also be prepared to participate in a 'group order' to commission Ultrascale to produce replacement brass gears if Heljan and Gaugemaster don't get their act together. However, I imagine it's going to be considerably more difficult to replace the gears on steam outline models because of quartering, notwithstanding the issues 57xx is encountering just getting to the gearbox!

Edited January 1, 2021 by 7007GreatWestern

[GWR57xx]

On 31/12/2020 at 15:32, GWR57xx said:

 

 

... the only other screws visible from the top are the three circled in lime green. 

 

 

I can report that the three screws circled in lime green do not release the gearbox. In fact it was hard to tell what they did, since nothing seemed to move after I undid them.

 

I decided that I wasn't going to go any further with this particular engine so it was returned to Rails for a replacement.

 

The replacement has arrived and went straight onto the rolling road for a test on DC. In reverse it ran super smoothly and was able to run very slowly without hesitating. Sadly when I put it into forward gear the same wasn't true. There was a sticking point somewhere with the crank pin on the centre driver (driver's side) at 10 o'clock. No obvious binding on any of the motion.

 

I didn't want to have a repeat performance of the previous example where it chewed up its drive gear, so off came the body for a look-see.

Various parts on this engine are unbelievably fragile, I hardly touched it and a couple of drain cocks broke off and the thin plastic "cable" running along the bottom of the driver's side running plate broke in a couple of places. I really think that for a couple of pennies extra these should have been wire, not plastic.

 

Anyway, happy to report that on this example the can motor was firmly held in place by a couple of sticky pads so couldn't disengage from the drive gear and couldn't rotate in its housing. The drive gear was also undamaged as far as I could tell, so didn't appear to be the cause of the sticking. It's now back together again for a bit more gentle running in, still sticking when going forwards, but far far better than the first example (so far).

 

Edited February 11 by GWR57xx
restore images

[Hal Nail]

7 hours ago, GWR57xx said:

 

I can report that the three screws circled in lime green do not release the gearbox. In fact it was hard to tell what they did, since nothing seemed to move after I undid them.

 

I decided that I wasn't going to go any further with this particular engine so it was returned to Rails for a replacement.

 

The replacement has arrived and went straight onto the rolling road for a test on DC. In reverse it ran super smoothly and was able to run very slowly without hesitating. Sadly when I put it into forward gear the same wasn't true. There was a sticking point somewhere with the crank pin on the centre driver (driver's side) at 10 o'clock. No obvious binding on any of the motion.

 

I didn't want to have a repeat performance of the previous example where it chewed up its drive gear, so off came the body for a look-see.

Various parts on this engine are unbelievably fragile, I hardly touched it and a couple of drain cocks broke off and the thin plastic "cable" running along the bottom of the driver's side running plate broke in a couple of places. I really think that for a couple of pennies extra these should have been wire, not plastic.

 

Anyway, happy to report that on this example the can motor was firmly held in place by a couple of sticky pads so couldn't disengage from the drive gear and couldn't rotate in its housing. The drive gear was also undamaged as far as I could tell, so didn't appear to be the cause of the sticking. It's now back together again for a bit more gentle running in, still sticking when going forwards, but far far better than the first example (so far).

I havent tried at Rails but other box shifters will test them first if you ask, which removes these issues. Knowing stocks won't last indefinitely, I'd be returning it quickly rather than risking damaging things taking it apart which might limit your options (in the interests of full disclosure, I think this is the first model I haven't immediately dismantled, mostly on the back of your warnings!).

 

Given the way their business model works, expecting a certain level of returns, we are entitled to keep trying till we get one that works well, (and unfortunately the only way things will change is if the level of returns is higher than they expected).

 

Frustrating of course but we dont set the rules.

[neal]

I was lucky enough to acquire a second of these, at the discount rate of Rails. This one was the green unlined late crest. 
 

For this, I thought I would chance my arm with the Rails own brand high power decoder (itself very economically priced).

 

The loco was pristine, with just the rear brake pipe fractured (easy fix).  I was amazed to see the loco run beautifully with the Rails decoder, smoother than more expensive Lenz and DCC Concepts screw terminal chips in Prairies and Hattons Gresleys.

 

I know that Heljan’s QA can be variable, and I may have been particularly lucky with this one, however comparing this model (and Prairies) with the Hattons A3 and A4, the quality and finish is much better.  

I would agree with others that the Hattons commissions are built down to a price (with a margin for manufacturer and commissioning retailer).  Sadly, quality issues now appear to have afflicted the Gresley coaches.

 

I won’t give up on Heljan, and whilst the Collett is my least favourite GW loco, I trust that Heljan can learn from their O Gauge steamers so far and get this one right.

 

Neal

[GWR57xx]

After four hours of running in on DC the replacement 4339 now runs nicely and very slowly so will be staying.

This time I decided to fit the "Chunky Boom Box" speaker in the tender alongside the LokSound V5 L decoder:

 

Wired to the SP2 terminals:

 

 

Sounds good, and from a normal operating/viewing distance of >0.5m it is not obvious that the sound is coming from the tender rather than the engine, which was my concern.

 

Happy again 🙂

 

Edited February 11 by GWR57xx
restore images

[dpgibbons]

A Loksound setup is very pricey, especially with the highly desirable stay-alive. So I'm tempted to try a cheaper hardwired solution, such as the Zimo sound install from YouChoos here.

 

Any thoughts please?

 

 

[Adrian Stevenson]

As long as you are confident with a soldering iron, no problem, go for it.

 

Cheers, Ade.

[dpgibbons]

BR versions now half price - GBP395 at Rails

[John Tomlinson]

3 hours ago, dpgibbons said:

BR versions now half price - GBP395 at Rails

Many thanks indeed for this.

 

I've had my eye on one for a while now, and this latest drop has tipped the balance for a green, lined example. What's not to like at this price!

 

John.

[dpgibbons]

9 hours ago, John Tomlinson said:

What's not to like at this price!

Paying GBP495 in the March sale! 

[brossard]

On 30/03/2021 at 22:55, dpgibbons said:

A Loksound setup is very pricey, especially with the highly desirable stay-alive. So I'm tempted to try a cheaper hardwired solution, such as the Zimo sound install from YouChoos here.

 

Any thoughts please?

 

 

 

I would always go for Zimo.  All but one of mine have Zimo, with one outlier having Loksound (because Digitrains stopped shipping to Canada for a while, all OK now).

 

John