9V Extreme - first test results re inclinations and power supply

[Haddock51]

After a long summer break - and the construction of the new 40 sqm LEGO-room being finalized - I have now decided to go ahead wth the 9V Extreme Project (for more details check my topic "9V Extreme - Climbing Wall and High Speed Track").

With all valuable and constructive advice I received during last spring, I have now a pretty clear idea about how this is going to look like (with the exception of the wiring). However, prior to setting up this huge train track, a lot of preparation work - including a series of testing in full scale - still remains to be done.

I recently mounted a first test track with the following technical data:

* total track length: 11,1 m (70 straight and 16 curved tracks)

* two full scale ramps, 3,9 m long, with 30 straight tracks each and inclinations of 7,7 percent

* two 180 degree curves on level 1 and level 2

* 30 cm elevation between level 1 and level 2

* one power supply unit Voltcraft EP-925 (3-15VDC, max 25 Amps) in combination with one LEGO 9V speed regulator

* four power connections evenly spread along the track

In order to "smoothen" the edge where the ramps meet the upper level, I put 1x2 plates under the connection of straight track 1 and 2 before and after the edge, thereby obtaining a hump, spread over four straight tracks.

The main purpose was to test the new power supply unit in combination with the LEGO speed regulator and to test the uphill and downhill performane of some medium and large sized trains equipped with two 9V engines - such as the Metroliner, the extended ICE (6 units),the extended Emerald Night (6 units) - and heavy trains equipped with four 9V engines - such as the Santa Fe and the extended Horizon Express (8 units), each of these trains weighing approx. 3 kg.

Test results:

The Voltcraft EP-925 in combination with the 9V speed regulator works perfectly! To obtain 9V at maximum speed, the exit voltage of the power supply unit needs to be approx. 13V. Unfortunately, the Amp meter of the EP-925 is not digital, so I don't know how much Amps are provided and consumed. However, the tests proved that the internal LM317 unit in the regulator provides enough Amps to get four 9V engines up to full speed. So there is no more need to use two 9V regulators in synch when operating heavy trains with four engines. Using four power connections resulted in very smooth train operations.

All tested trains managed to get up and down the ramps - and over the hump - without any problems, and at decent speed. An interesting - and indeed very positive finding - was that even the Emerald Night with its rigid 2+1 wheel axles managed to get over the humps - at decent speed - without derailing!

Downhill speed control is easier with trains equipped with four engines compared to those equipped with only two. Both the Santa Fe and the extended Horizon Express could actually be stopped before the end of the slope.

It is indeed fascinating and impressive to watch these large and beautiful trains climb up and down! And this is still "minisize" - the total elevation in the final 9V Extreme track will be 160 cm!

In summary - so far, so good.

So what's next?

* In order to get the trains from level 195 (cm above floor level) up to - and down from - the High Speed Track at 215 cm a.f.l. over a short horizontal distance, I need to mount two double track 360 degree Altitude Adjustment Circles (AAC) - hanging 2 m a.f.l. These will be rather sophistcated constructions requiring bendable - but steady- material (most likely plywood).

The AAC concept is new and still unknown, at least to me. The circle shaped ramps will have inclinations both vertically and horizontally due to bending. Unfortunately, I don't have access to a 3D CAD software to learn more about the correlations between circle length, elevation/vertical inclination and lateral inclinations, so accurate full scale tests will be necessary, on floor level ...

* I still need to evaluate and decide upon various materials such as MDF, plywood, solid glass, and thin polycarbonate (for protection fences around the 180 and 360 degree curves).

* As for power supply, I feel rather comfortable with the EP-925. As originally planned, I will connect four LEGO 9V speed regulators to the EP-925 which will allow me to operate up to four trains (with up to four engines each) simultaneously, and independant from each other. I still need to figure out how to install (and hide) some 2x300 m of wire (1,5 sqmm) and 57 power connections (there will be no soldering of wire directly to the tracks).

* And last but not least - I need to prepare myself mentally for cleaning/polishing some 1200 tracks ...

Edited December 12, 2013 by Haddock51

[Phoxtane]

Build a robot to clean those tracks for you!

[Xris]

Hi Haddock51,

I am preparing a similar project, involving 3 levels at 200, 220 and 240 cm above floor level. The corridor I want to use is approx. 10m x 1.9m.

At the moment, the design has a total track length of about 200 m. I plan to have inclinations of at most 3.7% between the levels.

I also want to include an "AAC" from the second level down to the floor. The device will allow me to have "temporary" extensions of the design and to allow for additions by the kids. It will be mounted on a dedicated mobile rack, so I can store it separately.

For the AAC I am planning 16 curves plus 6 straights per turn (more doesn't fit into the corridor or through the doors). The trains and track support require a head clearing of at least 14 bricks (13.44 cm) height between two turns. This results in an inclination of 0.64 bricks per track segment (4.77%) and 17 turns in total.

I have not made any experiments with the pulling force of LEGO trains yet, but I am quite sure, this will bring them to their limits. The main problem I expect is the increased friction within the inclines curves of the AAC.

Do you plan experiments covering this problem?

Greetings

Xris

Edited December 5, 2013 by Xris

[Haddock51]

Hi Xris,

Quite interesting reading indeed!

Some questions:

Do you use 9V or PF?

What kind of track support will you use? Plywood? Masonite? Or something else?

What size of trains are you planning to operate? With what weights?

The double track 360 degree AACs that I am planning for are based on the following calculations and assumptions:

* Total elevation: 20 cm

* Minimum clearance: 15 cm

* Track length inner circle: 317 cm (9 straight and 16 curved)

* Average inclination inner circle: 6,3 percent

* Track length outer circle: 421 cm (17 straight and 16 curved)

* Average inclination outer circle: 4,8 percent

The track length includes 2 straight before the circle starts and 2 straight after the circle ends for the inner circle (3 plus3 for the outer circle). These straights are part of the inclination.

I think you need to plan for at least 1 straight before and 1 straight after the point where the inclinations start and end.

In my previous large LEGO train track, I had 90 degree curved 8 percent inclinatons. I did not perceive friction as a particular problem when operating trains uphill.

Downhill operations with 90 degree inclined curves however were more challenging, with several derailments initially. Therefore I mounted protection fences in plastic

Edited December 6, 2013 by Haddock51

[Haddock51]

Xris, I am not shure how you have calculated your inclinations.

If you plan an inclination of max 3,7 percent - and an elevation of 20 cm between the levels - the AAC track length needs to be at least 540 cm which corresponds to 16 curved and 26 straight tracks.

Will you have enough space?

Edited December 6, 2013 by Haddock51

[Xris]

Hi Haddock51,

I will use 9V tracks and motors, but I will additionally create custom-made power pickups and various PowerFunction motors.

Max inclination of 3,7% is for the connecting ramps between the fixed parts of the layout running along the walls.

I do not need AACs to connect the three levels of the permanent layout. I even do not need any inclined curves here,

since a can use the long walls to gain all the vertical space I need.

The AAC is needed to connect extensional layout elements on the floor with the permanent installation starting at 200 cm above the floor.

In the AAC I need the steeper 4.77% ramp of almost 1.5 plates heights per 16 stud rail segment in order to achieve the clearance of 14 bricks.

Each ramp ist started and ended by 2 - 4 segments with gradually increasing/decreasing inclination.

For the fixed parts of the layout, I intend to create horizontal shelf like elements from plywood on a lattice. The upper side will be covered with felt or cork to dampen the noise of running trains. The lower side will be covered with removable decorated plates with built-in elements for ambient lighting. Between the two layers, technical installation, as decoders, switch motors and electric wiring will be hidden away.

For the AAC, I intend to use some a cheap standard shelf (like "IVAR" series from IKEA) and mount thread bars on it, which will connect to the ramp made of plywood. On the both lateral sides of the ramp, "fences" of plywood or thin acrylic glass will add to the ramp's stability and protect derailed units from falling down.

I am planning with a maximum train length of 20 rail segments (=320 studs or 2.5 m). This should be enough, since the layout models a single track branch line.

Since I'll be on a vacation until wednesday, I won't be able to read the forum before then.

CU

Xris

[Haddock51]

How will you mount/fix the plywood fences to the track?

[Xris]

Hi Haddock51,

How will you mount/fix the plywood fences to the track?

I intend to use cramps/staples and hot glue to attach the fence to the plywood forming the bedding for the track.

Using craps could be tricky, if the plywood bedding is too thin. A local model railroading club reported some problems with this topic, so I will call them a visit to learn from their experience.

They also told me, that othagonal supports under the bedding often cause problems, because the plywood between two such supports tends to sagging.

Regards

Xris

[Haddock51]

After multiple consulting with various experts in the fields of polycarbonate, MDF, plywood (and other materials) and model trains, I am about to change several of my decisions re the AACs:

Serious doubts have been raised re 2 mm polycarbonate and the way I planned to attach it to plywood. The kinetic energy released by a 3 kg heavy train derailing in a 180 degree curve - at medium/high speed - would be of such a magnitude that a poorly mounted fence most likely would not hold up the derailed train. In worst case, the polycarbonate fence might even crack.

Comparing material properties, in particular lateral bendability, and price, I am almost sure that I will build the AAC circular ramps with two layers of 6 mm MDF, glued together.

As for the fences, I will use 3 mm polycarbonate with 2 mm angle-irons (25x25x12 mm), attached to PC and MDF with small bolts and nuts/washers that will spread the forces in a better way.

I intend to build a complete prototype of one AAC in the beginning of January.

Enclosed below is a track design with level 195 (cm above floor level) on the top, the two AACs and the double high speed track on level 215/210 to the left, to the right and at the bottom of this picture. Dimensions: approx. 7 m x 5,5 m

Edited December 12, 2013 by Haddock51

[Brickthus]

Interesting to compare notes.

My layout is 16ft x 12ft, a modular double-track looped-eight with 2 straights between each curve on the corners of the main lines.

So far I have the outer track working and have built the slope modules and fiddle yard of the inner track. The fiddle yards are under the station.

The total elevation is 48 plates at the moment, leaving a bridge clearance of 37 plates. It takes half a lap to do the elevation.

I use slopes of 1 in 40 (1 plate per track piece) in the curves.

This increases to 1 in 30 on the straights (4 plates in one 48M module of 3 track pieces). 1 in 30 is the maximum slope recommended for a model railway by the books I have.

I use 1 in 80 (half plate increments for the first couple of track pieces) at each end of each slope.

I have a see-saw module for use at home in the middle of the slope straights. It can change from 1 in 30 to 1 in 120, allowing insertion of 2 32M modules to go round the attic pillars.

All the track slopes, cants and ballasting are achieved with LEGO parts as part of the modules.

I have succeeded in doing this for flexi-track too, for the future, if PF trains improve.

In layout design I did look at helixes for elevation change but decided that with the slope rules I use they would be too big and also unrealistic on all but a "rabbit" layout (Swiss mountain prototype).

I run 8mm:1ft scale trains which are 8+ wide, using a dual 30V 3A bench power supply on the main line and 9V controllers in the yards.

This allows me to start one train as another is stopping in the fiddle yard, so that the next train arrives sooner for people to see.

The power supply has worked well at 2-day shows for both 12V and 9V layouts.

The highest single-train load is 1.3 Amps for my Hogwarts Express at the top of the hill. This has 4x 9V train motors, 2x 71427 gearmotors for the driving wheels and 3x 9V light bricks in the firebox.

I try to keep to 200-250mA per 9V train motor, 300mA for short periods only. This ensures long motor life.

I have 9V track power feeds at each end of the layout on each layer. I have divided each main line into 4 sections with a power feed at each end of each section.

I intend to use suitably-rated diode networks to drop voltage from the up slope to flat and from flat to the down slope. This will help to bias the power in favour of continuous climbing and not running away on descent.

On the previous flat layout, one power feed at each end sufficed for each main line.

I have 2 Power Functions trains so far, but only for yard work and maybe an occasional trip round the main line. I tested PF and found the LiPo 800mA current limit was no enough for some of my trains.

If PF gained a slave motor driver and wireless inductive charging then I would be more encouraged to use PF on the main line.

I use 6mm thick 30cm high clear Lexan sheet panels along the front and sides of the layout at shows, which will arrest the heavy trains and help keep little fingers out of the way.

In practice, with the previous flat layout, these have also taken the brunt of the germs and kids have tried to chew them without success!

I may buy some 500mm high panels now that the layout height has risen above the baseboards.

Lexan is tougher than perspex, which might shatter.

I have attached the tables, legs and Lexan sheets with coach bolts. The tables are covered in Sundeala board, which reduces noise. It is used by many model railway builders.

You didn't say whether the 9V controller was controlling the voltage for some power transistors or not. My power supply is effectively an LM317 chip with 2 2N3055 transistors in darlington configuration.

I expect a 317 chip on its own in the controller, with 4 motors for a long time, might overheat. The controller was designed for about 300mA mean load.

Have you tried using just the bench power supply on the layout without the 9V controller?

Mark

[Heppeng]

I intend to use suitably-rated diode networks to drop voltage from the up slope to flat and from flat to the down slope. This will help to bias the power in favour of continuous climbing and not running away on descent.

You might want to consider a feedback type of controller that measures motor speed and automatically adjusts its output to keep it constant - I use a cheap Hornby R965 which has got my PF motors running under perfect control.

[Haddock51]

You didn't say whether the 9V controller was controlling the voltage for some power transistors or not. My power supply is effectively an LM317 chip with 2 2N3055 transistors in darlington configuration.

I expect a 317 chip on its own in the controller, with 4 motors for a long time, might overheat. The controller was designed for about 300mA mean load.

Have you tried using just the bench power supply on the layout without the 9V controller?

Mark

I recently learned that the LM317 in the LEGO 9V speed regulator supports up to 2,2 A which is indeed quite a lot - and significantly more than what is provided when using a standard LEGO transformer.

A standard 9V regulator with a standard 9V transformer provides approx. 1A which in my case previously resulted in using 2 speed regulators in synch when operating trains - like the extended HE - equipped with four 9V motors. With the Voltcraft EP-925 (see picture below), I can do it with one regulator.

I also tested a heavy goodstrain with 1 Maersk and 2 BNSF locomotives, each equipped with two 9V motors, so in total 6 motors. It worked perfectly, and I didn't notice any heating in the speed regulator.

Edited December 18, 2013 by Haddock51

[1974]

I recently learned that the LM317 in the LEGO 9V speed regulator supports up to 2,2 A which is indeed quite a lot - and significantly more than what is provided when using a standard LEGO transformer.

A standard 9V regulator with a standard 9V transformer provides approx. 1A which in my case previously resulted in using 2 speed regulators in synch when operating trains - like the extended HE - equipped with four 9V motors. With the Voltcraft EP-925 (see picture below), I can do it with one regulator.

I also tested a heavy goodstrain with 1 Maersk and 2 BNSF locomotives, each equipped with two 9V motors, so in total 6 motors. It worked perfectly, and I didn't notice any heating in the speed regulator.

An LM317 does _not_ provide 2,2A, 1,5A max. A common misconception with standard linear VREGs is only to look at current output. One has to factor in the total wattage loss which must not exceed 20W <- and that's with some serious cooling to boot!

So if you want to draw 1,5 amps with an input of 9VAC, which is about 12VDC at the lowest voltage output, I'm guessing 1,5V here - you have a loss of 10,5V/1,5A = 7W .. and that requires a pretty decent heatsink. Does the 9V regulator have that? If not, it'll go into thermal protection

You could up the ante with a properly cooled bigger VREG but there's really no point in that when you got a decent lab supply right at your bench Use that to power the rails directly

[1974]

Also, 9VDC into the train regulator will give you a bad overall regulation. The wallwart is 9VAC, no? Put in 12VDC from your lab supply as the LM317 likes 3V above output, if you want to continue to use the 9V regulator

[Haddock51]

I am getting confused, most likely because my knowledge in electronics is rather limited, and most of it is based on empirical experiences - at least w.r.t. LEGO.

The recent information I received about LM317 was based on the following reference:

http://www.electroni...LM317/LM117.pdf

Have I misunderstood something? Or have I been misled?

My first experiences with running 9V trains with multiple engines goes back to 2006. In those days, I only used LEGO 9V regulator(s). The inclinations in this track were very much similar to those I tested in my test-track recently, i.e. 7-9 percent

It was not possible to run a Santa Fe train (equipped with 4 engines) uphill with only one regulator - the train would hardly move - which is obvious because the output was only about 1A. That's when I started using 2 regulators in synch - a little tricky but it works and you can actually come up to decent speed.

So let me ask the following questions:

* If the LM317 - including the wattage loss you calculated - "only" supports up to 1,5A, how come I can run a train with six 9V engines at decent high speed uphill with only one regulator, i.e. one LM317? Calculating approx. 300 mA per engine, that equals to at least 1,8A! If 1,5A were the upper limit, I should have made the same experience as I did back in 2006 when operating only with one regulator. Please help me to understand!

* One of the key requirements for the 9V Extreme and Highspeed track is the possibility to run up to 4 trains (with up to 4 engines each) simultaneously and independantly from each other. My plan is to use four 9V speed regulators (connected to the EP-925). (Eight high level single-crossover switch points - at 2 m above floor level - will be operated electrically with a separate power supply unit).

I don't understand how this can be achieved by connecting the EP-925 directly to the track (150 m long and approx 25 switch points). With such a solution, how would I operate four trains independantly from each other? And again - PF is not an option!

* I haven't experienced any signicant heat problems so far (maybe because the trains are not operated over long time periods). How should I solve "heatsink" on 9V regulators?

Below a link to the technical specifications of the Voltcraft EP-925:

http://www.produktin...de_en_fr_nl.pdf

Edited December 19, 2013 by Haddock51

[peterab]

Most electronic devices will operate outside their rated specifications. This will involve creating greater heat and stress on the component, which will greatly increase the risk of failure (most likely by burning out). On the first page you can see the design load current of 1.5A. 2.2A appears on page 4, the Absolute Maximum Ratings, along with note 1; Absolute Maximum Ratings indicate limits beyond which damage to the device will occur. In the same table you can see that 2.2A is the typical maximum. This implies some devices will fail imediately above the minimum of 1.5A, and others will be OK till 3.4A. The maximum maximum (3.4A) suggests that above this all devices will fail.

What this means for you in practice is, if you rely on your 9V controller to handle more than 1.5A there is a chance it will burn out. That chance increases the more often you do it and the longer you do it for. These things are inherently statistical, and therefore unpredictable. That's why the guaranteed performance specifications are provided, so you can design to avoid failures.

To address your questions;

* back in 2006 your set up was limited by the current provided to the controller by the transformer, once the train was using all the current the transformer could supply it couldn't go any faster. Your present test set up has overcome this limit which now means there is more current flowing through the LM317 and thus a faster train, but also more risk of burning the LM317 out. 1.8A is below 2.2A which should mean most devices will be able to handle it for a while.

* To safely operate four trains you'd need four EP-925, or four controllers which will handle the current (this could mean modified LEGO controllers or any others).

* My train club has modified the 9V controllers by adding larger heatsinks (these are pieces of metal designed for good thermal dissipation, typically with a lot of fins which bolt onto electronic devices such as regulators, CPUs and graphic cards) and fitting a small fan to the rear of the case. This allows as to use a larger power supply without burning out the controllers.

[Haddock51]

Thank you for your clarifications, peterab. Learning through knowledge-sharing, very much appreciated indeed.The recent replies to my topic have resulted in better understanding of some of the issues related to this challenging world of electronics and electric devices ....

My conclusions so far:

* The combination of the EP-925 together with one LM317 regulator makes it possible to run heavy trains (such as the Santa FE and the extended HE) equipped with four 9V engines without problems, and without stressing the regulator since the maximum current required will be 1,2A which is below the upper limited of 1,5A .

* Even though, I will consider to upgrade the TO 220 heatsink in each of the four speed regulators that will be connected to one EP-925 in order to get better control of the temperature and thereby increasing the safety margin - and hopefully avoid to shorten the lifetime of the LM317s. Given my requirements on future train traffic, I am not considering an upgrade from LM317 to LM350 yet. This decision may still become valid once I have more experiences when the entire track is set up.

Peterab, it would be interesting to see a picture of the inside of one of the controllers you have modified, i.e. to see which one of the almost 30 available (conrad) TO 220 heatsinks you have chosen and how you have fitted the small fan at the rear.

* To run a long and heavy goodstrain with 25 - 30 waggons and three locomotives with two 9V engines each uphill is a different story. This will not allow for running other trains independantly in parallel since I will use the entire tracklength for this particular exercise. Having 6 engines run with only one regulator might very well increase risks and cause damages as several of you have clearly pointed out.

These are my thoughts:

As I mentionned before, the 9V Extreme track will be "segmented" into four separate segments with one EP-925 and four regulators to operate each segment independantly. Segmentation will be achieved through adequate adjustments of the switch points. I would call this the "closed" track alternative.

When running a train with 6 engines, I would go for an "open" track alternative, i.e. the entire track (except the train yards) is one segment with still four regulators providing current to the track. Using two regulators simultaneously - and at equal speed - along various parts of the track should make it possible to run such a train without problem, having maximum 3,0A available all way long (theoretically 6,0A but so much will not be required/used - and difficult to manage with only two hands). Six engines will require at most 1,8A which still is substantially below the upper limit of two regulators used in synch.

An additional challenge in this case will be the lateral forces with such a long and heavy train in the 180 degree curves uphill. In order to prevent derailing, it will be necessary to spread the engines along the train, i.e. 2 in front (pull), 2 midtrain (pull & push) and 2 at the end of the train (push). Another measure of precaution will be inclined curves.

The "open" track alternative could even open up for more engines/loads but I guess other questions/questionmarks will pop up, such as: how much Amps/Watt can you run through 9V tracks before they get overheated - or even worse: before they start to melt?

It is of couse tempting to push things to the limit when you have so much current and engine power available. However, my primary intention still is to use these capacities to create safety margins, to make sure that the entire power supply chain is in balance, not to stress various components/parts unnecessarily (and thereby risking to shorten lifetime) - but still allowing for exploiting the challenges and opportunities with inclinations and high speed for all types of 9V trains.

The replies so far have clearly revealed some of the weak spots (bottlenecks) and risks in this power supply/enduser chain - with very different price tags: to replace one LM317 (TO 220) would cost about 3 Euros (assuming no other parts of the speed regulator have been damaged), to replace a 9V engine would cost in excess of 40 Euros, to replace several hundred damaged/melted 9V tracks (including power connections and wiring) would cost a fortune ...

Comments?

Edited December 21, 2013 by Haddock51

[peterab]

Given that your EP-925 seems to have a total output of 3A, at 9V it would be very unlikely that you would be able to melt the LEGO track. I think your most likely point of failure after the controllers is your connection wires. If you use a number of parallel connections or high enough gauge wire that should be easy to avoid. Obviously the higher the current the greater the potential damage a short circuit can cause so take care when connecting things up.

I'll try and see if I can get an internal photo of the 9V controllers we use but I won't have access to them till late January (after our next show). I'm sure whatever heatsink is in there is chosen on the basis of easy fit, as is the fan which was fitted in a hole made in the rear wall of the case.

Your plan for the very long 6 motor train seem fine and is similar to what train clubs around the world have done so should be fine.

[Haddock51]

As the picture above shows, I am using the connections to the far left of the EP-925 which provide 3-15V/25A, the main reason being that I need approx. 13V outgoing voltage in order to obtain 9V at the power connections through the regulator - at maximum speed. In addition, connecting four regulators to the EP-925 will require approx. 5-6A.

Making a mistake could in worst case result in supplying 15V/25A which reeds 375 VA(W) - and that would certainly result in serious damages. Most likely, the LM317 would burn before reaching the power connections and the tracks.

No problem with waiting for that picture until end of January. In the meantime, I will certainly be busy with building the AAC (climbing spiral) prototype ...

Edited December 23, 2013 by Haddock51

[lightningtiger]

Has a power transistor being biased by a LM317 to supply a higher current being considered ?

That's an old trick to raise current output of a regulator by having most of the current required flowing through a power transistor.

Man, I didn't think tonight I had to remember something from 1985 !

[Haddock51]

Interesting question.

I guess the answer is no, at least not in this topic or previous topics related to power supply discussions re my 9V Extreme project.

Given my limited knowledge about electronics and electrical devices suitable for LEGO 9V trains, I am certainly not the right person to explain why "power transistors biased by a LM317 to supply higher current" have not been considered so far.

What would your recommendation be? How would such a fitting look like?

I assume that temperature control/-management in this scenario would be even more important and necessary.

Edited December 23, 2013 by Haddock51

[Haddock51]

The prototype for the double-track climbing spirale is now ready (except for the security fences).

After some experimenting, I decided to build the spirale ramps with MDF 6 mm, mounted on strips of wood (21x43 mm). Since the prototype is built on floor level, I had to construct special pillars to cope with the lateral tensions.

The basic idea was to build a solid and robust construction that can support long and heavy trains such as the extended Horizon Express and the Santa Fe train.

Some technical details:

* Elevation: 20 cm

* Maximum inclinations: approx. 8 percent (inner spirale) and approx. 5 percent (outer spirale)

* Minimum clearance (measured from railtop): approx. 14 cm

* Ramp width: 12,5 - 13 cm

* Track lengths (measured from start to end of inclination):

* inner spirale: 323 cm (9 straight, 1 half straight, 16 curved)

* outer spirale: 466 cm (20 straight, 3 half straight, 16 curved)

* Horizontal distance (measured between start and end of inclination): 70 cm

* Lateral ramp inclination: negligeable

* Power supply: 1 Voltcraft EP-925, 1 speed regulator for each track, 2x4 power connections (total test track lengths:12,4 m resp. 14,9 m)

The test results showed that the climbing spirale concept works perfectly! Compared with the previous tests on straight ramps with 8 percent inclination,, more power/speed is needed due to friction in the curves - particularly for the inner spirale. Future tests will show if the trains equipped with only two or one motor(s) manage to get up the inner spirale. If the results would be negative, I will have to run these trains on the outer spirale.

Quite amazing to watch these long and heavy trains winding up and down the spirales!

What a spectacular view it will be to see these climbing spirales hanging in the ceiling - 2 m above floor level - when the 9V Extreme project is finalized!

Edited February 7, 2014 by Haddock51

[Ashi Valkoinen]

Hi,

we use power supply units instead of the supply unit sold with the 9V speed regulator here in Hungary since 2011. Using MW: MW7H380GTGS power supply unit, putting out 3 Amps to the track at 9V seems to be enough to drive trains with 4 or even with 8 train motors. At one of our exhibiton we left a switch in wrong direction, all the four engines with 8 9V motors and up to 2-3 kg accelerated quick.

The power supply unit can drive my Stadler FLIRT EMU, which uses four motors (two 9V, both of them attached their neighbour 9V motor) and has 4 kg weigh, if there are regular current support around the loop of track. Maybe if I put two FLIRT EMUs together, I have to change the power supply unit. :)

[Haddock51]

Hi Ashi,

Thanks for your comment re power supply.

This has indeed been one of the most discussed and elaborated topics re my 9V Extreme projects so far. I finally decided to go for the Voltcraft EP-925. This is by all means an oversized solution since it provides up to 25 Amps at maximum 15 V (which reads 375 VA ...)

My plan is to connect four 9V speed regulators to one Voltcraft unit and to operate at most four trains - each one equipped with up to four engines - simultaneously, which will require approximately 5 Amps at most (approx. 45 VA).

The LM317 provides approx. 1,5 Amps which is enough to operate trains equipped with four engines, each one requiring up to 300 mA.

I am quite happy with the Voltcraft. One issue that still remains to be solved is heat control of the LM317. I decided to exchange all existing heatsinks in the LEGO 9V speed regulators and eventually install small vans.

By the way, Peterab, don't forget to send me the pictures you promised me.

Edited February 7, 2014 by Haddock51

[Ashi Valkoinen]

Hello Haddock,

My plan is to connect four 9V speed regulators to one Voltcraft unit and to operate at most four trains - each one equipped with up to four engines - simultaneously, which will require approximately 5 Amps at most.

Is it not easier (and cheaper) to buy four power supply units (with current limitation 5 Amps), and plug each of the 4 9V speed regulators?