42056 - Porsche 911 GT3 RS - MODs and Improvements

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid
Spellcheck

[Jeroen Ottens]

12 minutes ago, Paknaloid said:

Given the above post, I MUST have mounted the gears incorrectly, as I am following the jb70 version and when testing the gears after completing the box 1'steps, I either have some movement of the gears and drivetrain as they should - maybe 5-10cm - and then everything grinds to a halt; or, the drivetrain continues to move i.e. The wheels would be able to turn, but the gearing and engine catch and stop moving - so it would move across the table, but the gear and engine functions are somehow bypassed.  I spent an hour trying to trouble shoot and couldn't spot the issue,  I have a feeling I'll need to tear the chassis apart and start again, but by any chance is there something very obvious I've missed??

heres a couple of photos...

 

 

It looks like both driving rings are engaged at the same time. That will lock up the drivetrain. Are the orange 2L with axleholder parts (the ones that connect to the 9L steeringlinks) 90 degree rotated with respect to each other?

[Kumbbl]

17 minutes ago, Paknaloid said:

Given the above post, I MUST have mounted the gears incorrectly, as I am following the jb70 version and when testing the gears after completing the box 1'steps, I either have some movement of the gears and drivetrain as they should - maybe 5-10cm - and then everything grinds to a halt; or, the drivetrain continues to move i.e. The wheels would be able to turn, but the gearing and engine catch and stop moving - so it would move across the table, but the gear and engine functions are somehow bypassed.  I spent an hour trying to trouble shoot and couldn't spot the issue,  I have a feeling I'll need to tear the chassis apart and start again, but by any chance is there something very obvious I've missed??

heres a couple of photos...

AFAICS at the first pic there are enganged both driving rings at the same time which results in a moving range of about 5 cm (as you have reported): this is possible because each red clutch gear allows a 1/4 turn of a driving ring before it is catched - and because two different gear ratios are enganged at the same time the whole gear train stalls after both red-clutch gears are catched... the further moving on of the car is possible because then the pin-clutch of Didumos chips in so the car can be moved ... in a sequential gearbox always just one driving ring is allowed to engange!

Solution: check the orange liftarms 2x1 at both side of the indexer module - they must look like as in step 34 (page 28) of the BI of jb70... Caution: the must have an offset of 90 degree to each other, so when the left one (divers side) looks topwards then the right one must look frontwards.

BTW: I strongly recommend to omit the pin-clutch of jb70/Didumos: This means replace at step 38 (page 32) the two blue pins with one axle 4L --> the Didumos-errata-based jb70-Porsche does not need this clutch because the gear-box works now 100% reliable and it never stalls... in contrary: the clutch has the disadvantage that i chips in when not necessary cause of too low friction of the pins!!

Edited June 6, 2017 by Kumbbl

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid

[Kumbbl]

Just now, Paknaloid said:

Thank you both - this must be where I went wrong.  I'll try to fix tonight, hopefully without needing to dismantle too much!  Extremely helpful, thanks again.

no, cool down ;-), you have nothing to dismantle - very easy to fix at each stage before the body is mounted!

Please have a look at the BTW at the end of my previous post - i recommend you to fix now because later one it will be difficult! ;-)

Edited June 6, 2017 by Kumbbl

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid

[Didumos69]

9 hours ago, Appie said:

A question: do people notice more friction in second gear than the other 3 after fixing the sequence to being correct? The funny thing is that when I put gear 2 and 3 like the instructions (so 1>3>2>4) the friction is gone (just like Lego said why they did it).

In gear 1-3-4 I can push the chassis with no rigidity reinforced mods perfectly smooth over the table, but second gear comes to a grinding halt within 5cm.

Considering the same amount of gears are turning, I don't understand how this happens. Additional bracing of parts doesn't seem to help nor moving the gears of gear 2 and 3 in front of the the rear axle. Removing the red 16T of 4th gear made no difference (figured that different speed might be messing it up like Didumos pointed out) , but perhaps I need to fully adjust the axles like @Didumos69 did with his mod.

I noticed this too as did @I3Brick in this and following posts. EDIT: Also @charlesw confirmed this effect in this post. Let me recall my explanation:

On 31-10-2016 at 8:55 PM, Didumos69 said:

After applying Boratko's fix I also experienced the biggest issues in 2nd gear and also others reported on the same issue. There's a known issue with the red clutch gears that are not actually involved in one of the two clutches - the ones lined up between the normal 16t gears just in front of the actual gearbox. There is always one of them involved in transferring drive, which gives problems when its axle is forced into a very different rotation speed. Reasoning from the gear ratios in the gearbox I arrived at the following conclusions with respect to the RPM ratios between the mentioned red clutch gears and their axles for the situation with Boratko's fix:

• 1st gear, right clutch gear involved, gear / axle RPM ratio is 1:1
• 2nd gear, left clutch gear involved, gear / axle RPM ratio is 1:3
• 3rd gear, right clutch gear involved, gear / axle RPM ratio is 25:9
• 4th gear, left clutch gear involved, gear / axle RPM ratio is 1:1

So in 2nd and 3rd gear the axle of the involved red clutch gear rotates almost 300% faster or slower than the gear itself, Taking into account that the clutch gear will be pushed against its axle as it transfers drive, it's not strange that this causes severe friction. Especially in 2nd gear when the engine needs to rotate faster than in 3rd gear. Now let's see what the ratios are when we don't apply Boratko's fix:

• 1st gear, right clutch gear involved, gear / axle RPM ratio is 1:1
• 3rd gear, left clutch gear involved, gear / axle RPM ratio is 5:9
• 2nd gear, right clutch gear involved, gear / axle RPM ratio is 5:3
• 4th gear, left clutch gear involved, gear / axle RPM ratio is 1:1

These ratios are much better compared to the situation with Boratko's sequence fix. Especially 2nd gear, which has to work harder than 3rd gear, has a much better ratio. I think this explains the difference you're experiencing.

Actually, this would make the explanation given by TLG for having the 1-3-2-4 sequence partly correct. They said: "Too many gears are engaged at the same time and smooth running with all those tolerances is just not possible." Indeed, the 1-3-2-4 sequence allows for a more smooth running of the whole drive train. However the first part of the statement remains to be untrue, because in each gear all axles rotate and in each gear the same number of gears is involved in transferring drive from input to output.

Btw, as you probably already know, a way to avoid red gears from transferring drive on axles rotating at different speed has been described here and is also part of the unofficial errata.

Edited June 6, 2017 by Didumos69

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid

[Kumbbl]

9 minutes ago, Paknaloid said:

Quick question - as I'm building the jb70 version, do I still need to implement Blakbird's step-5 fix?  I can see that the gap which has a spacer in the stock version and two bushes in Blakbird's fix has nothing at all at this point on the axle in the jb70 instructions.  Is the issue fixed elsewhere, or do I need to add the two bushes per Blakbird's fix?

this fix is included in jb70s package - you don't need the bushes - the best solution is to put nothing there, neither a pin connector as in the stock BI nor the bushes blakbird applied... excatly this is shown in the BI of jb70... but the only important thing is: do not use a pin connector there ;-)

In general: The jb70-Porsche is errorfree (with one exception concerning the mounting of the front-calipers) and contains all of Didumos errata - the only thing i recommend to omit the pin-clutch of Didumos (s.a.) - apart fro that - just build it exactly as described... if something does not work it is probably your fault...the BI are proven to be correct...

Edited June 6, 2017 by Kumbbl

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid

[Kumbbl]

1 minute ago, Paknaloid said:

Assuming my error, where on earth does this 2L pin go?!

funny, but this is one of just 4 very small "errors" i will describe - i'm working on a special posting for that - will post it probably today afternoon.... so: IMHO this 2L blue pin is not needed...

Edited June 6, 2017 by Kumbbl

[Paknaloid]

.

Edited January 4, 2018 by Paknaloid

[Appie]

5 hours ago, Didumos69 said:

I noticed this too as did @I3Brick in this and following posts. EDIT: Also @charlesw confirmed this effect in this post. Let me recall my explanation:

Thanks and my apologies, should have re-read the topic, I remember that reply from months ago now that I see it again. As a result I added your frictionless style axles to the gearbox. Except I only used it on the side of 4th and 2nd gear. This had 2 reasons: I could mount an 2x4 liftarm here easily providing space for a 4L axle with stop and 3L axle with stop. I personally don't like a 2L axle "floating" halfway in a liftarm. I know the chances of it moving are 0.0001%, but I just don't like it. The other reason was that on the other side of the gearbox I already mounted my new version of a 1 speed reverse (this time same speed as first gear ) and didn't feel like tearing it down again. Luckily for me the real culprit in this story was the red clutch gear in front of the gearbox that goes to the selector axle of 4th and 2nd gear. Once I changed this, second gear became as silky smooth as the other 3 gears. So thanks Didumos for figuring this out.

Another thing that felt off to me was how the rear part of the gearbox was basically mounted on a pair of these mounted only in the middle to the 5x7 frame below it. I could visibly see that stuff bend when it was in second gear (like Didumos said, it's pushing it down). So I added his brother to the dead space of 5L axle which I then mounted to the 5x7 frame below it with a couple of black pins. I did this before changing the axles and could notice a slight decrease in friction, small, but noticeable.

 

@Kumbbl

I agree, alot of these errors have no place in product like this, luckily there are people out here that will fix Lego's utter crap (the pin joiner in step 5 is a serious triple red card Lego, it does not fit!) I do hope however that people who aren't on forums like this, will email Lego to death with the clear flaws in this product so they know better for next time. And I know they get emails for the dumbest crap, but in this case they are warranted. The product was rushed (not finding a proper solution like Didumos for 1>3>2>4 clearly shows this imo) and very flawed in general (see errata). I have been wanting to quit this build many times and I just barely finished box 1. I never take breaks from building a set, yet this has been on my table for 3 weeks now. There's only so much face palming a guy can take in a week.

In other news, I decided to drop the rear wheel steering mod. I got pretty close to installing the Claas method (only method I knew which will auto-align the wheels on gearshifts), but could already see slack getting the best of me when the steering angle at the rear is so small and my space at the front to compensate for that slack being limited. And then I figured: if I am going to bother to make a mini version of the Claas method, why not actually give it a whirl in a mini Claas (yes, I am aware somebody already made one, I am not original )

I am still trying to get my indicator mod to work. I want it below the speed-o-meter in the dash, which I managed, but I need to brace the steering better. After that, I am done with this model though.

Edited June 6, 2017 by Appie

[Kumbbl]

1 hour ago, Appie said:

@Kumbbl

I agree, alot of these errors have no place in product like this, luckily there are people out here that will fix Lego's utter crap (the pin joiner in step 5 is a serious triple red card Lego, it does not fit!) I do hope however that people who aren't on forums like this, will email Lego to death with the clear flaws in this product so they know better for next time. And I know they get emails for the dumbest crap, but in this case they are warranted. The product was rushed (not finding a proper solution like Didumos for 1>3>2>4 clearly shows this imo) and very flawed in general (see errata).

problems can occur, but what is the real shame for TLG, that it would have been not necessary to go back to the drawing board and begin again from scratch... just take into account how fast guys like Blakbird (found during its build the step-5 error) and Didumos (who fixed within some weeks ALL drivetrain-problems and not with some clumsy and flimsy hacks but with very clean and robust and failsafe solutions) have fixed together the problems so a flawless model is the result. Do not tell me that guys like uwe wabra who e.g. has designed the wondeful and tricky gearbox of the 8258 crane truck - long time reference for best gear-trains - can not do the same if they get the time for it - it is IMHO completely impossible that these flaws didn't occur at TLG-internal tests - they knew the problems for sure but the risked to push the product to the market because the assumed that most of the customers would be blended by the gorgeous look of the car and want only put it onto their shelf - so the assumed the complaints would be manageable and would not result in lost future costomers....

so it was a weighting between risking unsatisfied customers and some costs when fixing there problems before going to market... they risked some unsatisfied customers but IMHO they were right that this this would not result in real problems - most of the people who bought the Porsche will buy for sure the next car in this "UCS"-technic series... ;-)

[Paknaloid]

.

 

Edited January 4, 2018 by Paknaloid
added picture and smalled them down

[Kumbbl]

2 hours ago, Appie said:

@Kumbbl I decided to drop the rear wheel steering mod. I got pretty close to installing the Claas method (only method I knew which will auto-align the wheels on gearshifts), but could already see slack getting the best of me when the steering angle at the rear is so small and my space at the front to compensate for that slack being limited. And then I figured: if I am going to bother to make a mini version of the Claas method, why not actually give it a whirl in a mini Claas (yes, I am aware somebody already made one, I am not original )

IMHO the class method is not only the only one which allows auto-align (which is IMO a must feature for AWS with diferent modes) but it is also the only one which does not suffer from backlash which is immanent for gear-based solutions - and driving-ring-clutch.gear-based solutions (for changing mode) is a complete no-go cause of the fact that a clutch-gear based design alwas has at least(!) a full quarter turn backlash so the rear steering is often complete useless...

But my hat off Appie, if you were able to integrate this linkage-monster into the gearbox-stuffed chassi of the Porsche without disrupting the ground clearance and the quite clean and rigid design of the chassis...  IMHO real working and realistic RWS is not really possible with current lego parts (the class-method works only great because the steering angle is equal on both axles and has not to be minimal at the rear axle) - at least not for a race car cause of its minimal steering angle at the rear axle - regardless of the mode.

The RWS of the good old 8880 is a nifty feature but it only works because it has only one mode - which is completely unrealistic .... therefore i built in a mode selector some years ago which is gear based but not driving ring based (cause of the too high backlash) but based on gear-switching like in old 8860 - for just two "gear-modes" the behavior is quite acceptable but of course sometime somehow rough ;-) but it works. But it has of course no autoaligning so it allows only mode-switch when all 4 wheels are straight forward... so overall it was a try but i have to admit it is not really a ready-for-production implementation ;-)

what would be intersting: i heard rumours that TLG had a RWS solution in their beta-models which was not approved by Porsche - would be intersting to see this refused design of TLG....

But also your design would be intersting for me, Appie - could you provide some pics?

Edited June 6, 2017 by Kumbbl

[Appie]

37 minutes ago, Kumbbl said:

just take into account how fast guys like Blakbird (found during its build the step-5 error) and Didumos (who fixed within some weeks ALL drivetrain-problems and not with some clumsy and flimsy hacks but with very clean and robust and failsafe solutions) have fixed together the problems so a flawless model is the result. Do not tell me that guys like uwe wabra who e.g. has designed the wondeful and tricky gearbox of the 8258 crane truck - long time reference for best gear-trains - can not do the same if they get the time for it - it is IMHO completely impossible that these flaws didn't occur at TLG-internal tests

Exactly why I think the product was rushed. Didumos and others fixed this thing so fast it wasn't even funny. Only adds more to the insult that they think we'd be OK with this kind of model. While most customers probably won't care, because it looks nice and they will never remove the body, because that "in a normal gearbox I wouldn't be here" white clutch gear solves all problems there might be, but personally, I can not take another of these sets. I might buy one to build awesome stuff like Jeroen Ottens' DB11, but never again for the actual model if it is as flawed as this one.

 

27 minutes ago, Kumbbl said:

IMHO the class method is not only the only one which allows auto-align (which is IMO a must feature for AWS with diferent modes) but it is also the only one which does not suffer from backlash which is immanent for gear-based solutions

I am all ears if you have other solutions besides the Claas which auto-align. The problem I ran into with the Porsche was that at one point I had to involve gears, so I could route axles to the rear steering, so more slack. Perhaps with a custom chassis I could have removed that problem, but I didn't think it was worth the effort (and tbh, I want to spend my time on other Lego stuff).

I don't have any pictures of this solution, but it was basically a combination of the rear hog route, my 4>2 gear selector from my first attempt at this 2 mode RWS and the Claas method. I used the same as my original attempt, so in front of the midconsole and behind the front wheels was my playground to squeeze it in. The reason I completely dropped it was because I felt I came up one stud short at the top (collided with the normal steering to front axle) and it looked like I couldn't get enough angle on my small Claas mechanism to compensate for slack. Also in the back of my mind was that I had to brace this stuff to hell and back like Lipko's telehandler with Claas mechanism to compensate for the tires acting like glue on a surface. It felt like a little too much for the limited space I had available, it could very well be doable, but I had a hard time visualizing it.

I also heard TLG had a prototype with rws, but that Porsche said no. I have a hard time believing that to be true though, considering the crazy amount of flaws in the current version (if you have time to add another feature, you have time to make sure the stuff you made first actually works). And even if they did have rws, I wonder if it had an actual switching mechanism to switch between 2 modes of rws or if it was just like the 8880. Porsche could have said no, because they didn't find the angles realistic (which they won't be with Lego) or was like: nope, don't want 8880 steering. I don't think such a version was presented to Porsche though, the 8880 version maybe, but with a real switch? Nope.

Edited June 6, 2017 by Appie

[Didumos69]

@Kumbbl, about the door locks. Isn't there a way to get rid of the flat plates? They come very close to the paddles and steering wheel IMO. @DayWalker's version doesn't seem perfect either because the black beams and orange panels are only connected with one pin.

About the linear clutch you're suggesting to drop. There is still a very small chance of double engaged gears when shifting and pushing the model at the same time. To me it's a good practice to protect the drive train in such cases. Also, when the linear clutch tends to slip, I treat that as a sign that something is causing friction that shouldn't. Untill now I always managed to get things working without any slipping.

Final question. I am planning to take apart my Porsche and build yours. Are you planning to bundle your mods as an add-on to @jb70"s instructions?

[Kumbbl]

Just now, Didumos69 said:

@Kumbbl, @DayWalker

 

Final question. I am planning to take apart my Porsche and build yours. Are you planning to bundle your mods as an add-on to @jb70"s instructions?

funny - currently i'M working on a write up of exactly this as an own topic - i'm in the final state... ;-)

[Didumos69]

9 minutes ago, Kumbbl said:

funny - currently i'M working on a write up of exactly this as an own topic - i'm in the final state... ;-)

Great ! Take your time.

[Kumbbl]

10 minutes ago, Didumos69 said:

@Kumbbl, about the door locks. Isn't there a way to get rid of the flat plates? They come very close to the paddles and steering wheel IMO. @DayWalker's version doesn't seem perfect either because the black beams and orange panels are only connected with one pin.

i have noticed this also but they do not collide - you can get rid of the front 1x6 tile by using a 1x8 tile instead of the rear 1x6 tile. But i have refused this because i have also noticed that a males hand can only hardy switch the left paddle if drivers door is closed - regardless which paddle-shifter you use - so i always switch when divers door is open... well therefore i will not change the plate coverage for my Porsche - but of course your finding and hint is correct!

[Paknaloid]

Didumos - as a neutral third party, which of the two door locking mods do you prefer? I'm undecided which one to add but am a way off being at that point anyway , interested to hear if you favour one over the other

[Kumbbl]

54 minutes ago, Didumos69 said:

@Kumbbl,

 

Final question. I am planning to take apart my Porsche and build yours. Are you planning to bundle your mods as an add-on to @jb70"s instructions?

well, here we go - see this new topic:

Its not a full blown PDF BI which contains jb70s BI and all integrated MODs but it is a detailed write up which MODs the Porsche contains and how to integrate all this into the base-package of @jb70 - with some links to my bricksafe-folder which containes many detail photos... I have never used LDD or other tools like this so i'm not able to generate such beautiful BI as jb70 but an experienced builder shoule be easily able to rebuild this Porsche. But maybe @jb70 is following us and our praisings and maybe he would be willing to enhance its BI and to integrate some of the MODs described there - i would assume it would not too much effort for an experienced BI-maker....

11 minutes ago, Paknaloid said:

Didumos - as a neutral third party, which of the two door locking mods do you prefer? I'm undecided which one to add but am a way off being at that point anyway , interested to hear if you favour one over the other

DayWalker, i'm also very neutral and i appreciate all your MODs but in case of the door locking i prefer that one of el Squatter because IMHO it is more clean and robust - based on springs is always prefereable compared to rubber bands - and your is also fixed by only one pin... and working inside door handles does IMHO not add much value - but do not misunderstand me: this should not downrate your its just my personal taste. keep your very good MODing work!

[DayWalker]

2 hours ago, Didumos69 said:

@DayWalker's version doesn't seem perfect either because the black beams and orange panels are only connected with one pin.

I dont understand. The black 11L beam (and the 4x2 L beam)  is fixed by two connectors, on the long and short orange pannels respectively. Just look at the LXF ^^

1 hour ago, Kumbbl said:

DayWalker, i'm also very neutral and i appreciate all your MODs but in case of the door locking i prefer that one of el Squatter because IMHO it is more clean and robust - based on springs is always prefereable compared to rubber bands - and your is also fixed by only one pin... and working inside door handles does IMHO not add much value - but do not misunderstand me: this should not downrate your its just my personal taste. keep your very good MODing work!

No problem ^ ^

What I really like with this porsche, is that TLG has made a perfectible work, so many mods are possible to balance it, and it is very pleasent to see differrent solutions ^^

Edited June 6, 2017 by DayWalker

[Didumos69]

2 hours ago, DayWalker said:

I dont understand. The black 11L beam (and the 4x2 L beam)  is fixed by two connectors, on the long and short orange pannels respectively. Just look at the LXF ^^

You are right. In the LXF-file the 7L panel and the second pin are missing. I still do not really like the long unsecured end of the 11L beam though, but I guess there are more volnerable spots in this set . Btw, the 7L panel is available in LDD now.

3 hours ago, Paknaloid said:

Didumos - as a neutral third party, which of the two door locking mods do you prefer? I'm undecided which one to add but am a way off being at that point anyway , interested to hear if you favour one over the other

I do not really favour one. I would have to actually build them to be able to compare them.

Edited June 6, 2017 by Didumos69