Spoke a little too soon!

Had used 3 x 40A glass fuses in a ganged fuse holder to fuse each motor circuit and then use the extra fuse holder for the accessories. The centre motor has a consistently high fuse consumption so was concerned that if a fuse failed, the redundancy of the installation would be compromised. The outer motors did not fail fuses (different style of motors).

A fellow modeler, who is an Electronics Engineer suggested using Automotive style blade fuses instead of the glass style. Blade fuses have a “slower-blow” failure characteristic. Suspected the fuse failures were being caused by high current peaks, not by a steady operating current. This modification could confirm that.

Decided to purchase a 6 blade fuse holder from Banggood, but try a 30A blade fuse in each motor circuit. The three extra fused circuits would operate the cooling pump, ESC control and accessory circuits at much current lower levels. If necessary, blade fuses can also be obtained in up to 100A capacity. Whereas glass seem to have a maximum of 40A.

On what is probably our last sailing day of the season decided to try this new layout. Over about 35 minutes of operation did not experience any fuse issues. She worked flawlessly.
The boat will also plane comfortably on either the wing motors or the centre.
Because of the small pond was not able to maintain top speed using all three motors for more that a few seconds, but am confident a scale 50knts is now obtainable. Task for the Spring.
An added bonus, evident from the pictures, is this new arrangement allowed a simpler, cleaner wiring installation.

Blog # 28

Used the Brave many times this summer and have made some significant upgrades.
The original Turnigy (Hobbyking) ESCs both burned out - the smell still lingers beneath the deck! Both were rated at 30A and were protected by 30 A fuses, exceeding the measured maximum current draw.

Have now had too many Turnigy ESC issues, so decided to replace the penultimate failure with an inexpensive Banggood water cooled ESC of the same rating. It had been previously purchased for another project.
After several frustrating hours trying to program it through a series of beeps got it to work as desired.
Was amazed at the difference! This ESC provides good low speed modulation and control throughout the motor speed range.

The second Turnigy ESC then failed, so immediately replaced it with a similar Banggood one – this also works fine!
Have used the model with these new ESCs for many hours and am pleased with their performance.

One of my friends has a GPS speed tracker, so could measure how fast the model actually goes. The max scale speed only equates to 37 knots - well below the 50 + a Brave could achieve! Nevertheless it still planes easily and looks great.

As the low speed control is so vastly improved, decided to replace the centre shaft brushed motor with an inrunner brushless purchased earlier. This should improve performance, perhaps even achieve the scale 50 knots!
The drivetrain layout would also replicate the actual vessel, where the wing engines can be used for manoeuvring and the centre to supplement them in higher speed situations.

Bought and fitted another Banggood ESC, to this same spec. However, this one came preprogrammed with usable default settings, no beeps to decipher!
Works fine and the performance has increased markedly; will get around to measuring it one day, maybe in the Spring as the weather has turned noticeably colder.

The weather has finally cooperated and have had one good outing at a local lake. The performance is consistent with the previous blog, however, observed over a much greater distance.
At full speed on all motors the plane exposes more keel than pictures of the real vessel; but the throttle can be reduced to slow the model and lower the bow. The centre shaft performance is adequate, which helps maneuvering and also serves as a “get home” back-up.
The pictures show the model at full speed with all shafts turning and at full centre shaft speed only. The brief video shows the model at full speed and then slowing as the outer shafts are stopped.
On 1 x 4500mAh 3S battery, life is around one hour at a mixture of speeds; with both single and three shafts operating.
In summary a great project, one that has proved both challenging and satisfying.

Did more testing and then reprogrammed the ESCs and Rx. These changes allowed the removal of the brushless motor selection circuitry and improved the tractability of the system.
As the videos show, a plane can be achieved with both 2S and 3S batteries. With the 3S the plane is flatter and faster. Suspect this is due to the increased vertical component of the propeller thrust forcing the rear of the vessel upwards, lifting it and lowering the bow. Of the two, the 3S resembles the original vessel pictures more closely, the bow lifts too far high with the 2S.
The videos show the vessel in motion. The first shows a run with 2S batteries, the next two show similar runs with 3S.
Just found cannot upload three videos on one blog. Anyway, they can all be seen on UTube under 'Brave Borderer blog'.
Anticipate further refining of the powertrain with a future running, but doubt any significant improvements will be achieved. Performance is great anyway and am very happy with it.
This will be the concluding post for this model.

---

After the trials, tribulations, experiments and frustration, can finally announce this model is sailing well. It has excellent performance and looks, both on and off the water. A build really for the more experienced modeller though. If the performance envelope is to be explored it should be done on a suitably large area of water with few other vessels around.
However, enjoyed building a model of an unusual class of vessels. It expanded my knowledge and skills, then kept me occupied throughout a Canadian winter!
Have also been lucky to have received considerable useful advice from other modellers, which was much appreciated.
Building from scratch off a proprietary glass fibre hull gives a robust model with plenty of opportunity for individual input. Intend to use this approach again.
Whilst am rather blinkered about building scale models to scale; considerable frustration, time and money could have been saved by using a single or even a twin screw layout, rather than the triple. Performance might also have been further improved and weight saved. Several of the U Tube videos of the similar Perkasa models show how well these layouts work.
There are several photos attached;
1) Stationary.
2) Max. speed using the centre screw brushed motor and then the outer brushless screws on the 2S battery
3) Max. speed using the centre screw brushed motor and then the outer brushless screws on the 3S battery.
Because the pond is small, was unable to capture the marked differences in performance between the 2 and 3 S batteries. It will plane easily on either, just rather more frenetically on the 3!. Suspect this was because of the difficulty in coordinating acceleration, photographer and deceleration. Will endeavour to get pictures and a video that make this more obvious and post when available.

---

Getting ready for the maiden voyage after the extensive winter upgrade and stumbled across an article on programming Hobbyking ESCs.
Have always been concerned about the “stuttering” of the brushless motors. This is more correctly identified as “cogging”, often caused by incorrect motor timing. Changed the timing to “Automatic” from the incorrect previous setting, which was for inrunner motors not outrunners. Now much smoother acceleration from neutral to full speed.
On the water, first, the brushed motor centre screw concept works well. It allows controllable maneuvering and powers the model nicely in a displacement mode. No planning, but with a reasonable bow wave. The model weight is a little high, as she rides towards the top of the boot topping.
Tried the brushless outer screws and realized my efforts to “detune” the motors were overdone. The model performance was adequate, but still had difficulty in getting both motors to run consistently and smoothly. Think the water drag was slowing the motors more that expected as they had worked satisfactory on the bench.
Adjusted the ESC output back to 100% - much better. After 45 minutes of running the battery alarm came on, curtailing trials. She planes very much as the real vessel did, lifting the forefoot clearly out of the water. Had a spare battery, but 3S not 2. Was apprehensive about the increased power this would bring, but as the model was operating satisfactorily decided to try it.
Better still! Motors now work consistently smoothly and in concert throughout the throttle range.
The top speed exceeds any picture I have seen of the Brave class though. Can readily get a third of the hull out of the water.
Will conduct more trails, but need a larger pond to give time to experiment with the various control settings. Hope to post a concluding blog with pictures shortly.

---

During our protracted wait for Spring, have had chance to reflect on the earlier tribulations with the brushless motor installation. My enthusiasm was sparked by an article on getting the best from your Tx in a magazine and the advantages of using the Elevon or V Tail facilities if available.
My Flysky Rx/Tx has this facility, so after some experiments, set the boat up to use the Elevon function. This puts both motor ESCs on the same control lever, so advancing the control accelerates both screws equally and moving to either side accelerates, or slows the appropriate motor. Think this will reduce the coordination previously required to sail in a straight line.
Fresh from that success decided to experiment further. The Tx allows reducing Servo throw, so tried that too so could reduce the max. ESC output, it also seems to work.
One problem with the original layout was that if the forward speed was reduced quickly it was very easy to move the lever through the neutral point, overshooting into reverse. This immediately reversed that motor causing erratic operation. The Tx also has a “Dual Rate” facility to adjust the lever response from linear to various alternate rates. This allows the neutral point zone to be increased, so the tendency to overshoot into reverse is reduced. Bench tests of these adjustments suggest the model will be easier to operate. Hope they will make controlling the models performance better in our closest pool, which is quite small.
A further precaution to prevent inadvertent operation of the brushless motors when using the centre, brushed one, was to add two small relays; one into each of the white ESC signal wires. This allows a another channel on the Tx to be used to select, or deselect the brushless motors.
Anticipate that by turning this channel and thus the brushless ESCs off, manoeuvring can be accomplished without the operation being overwhelmed by a brushed motor inadvertently being operated.
My patience for Spring is getting thinner!

---

Winter seems to encourage modeling, have spent many hours in hibernation working on the deck and superstructure details.
A supplier offers a full set of Perkasa fittings, most of which would work on the Brave B. At one point considered buying a set. They are made in both resin and cast metal. Eventually parsimony prevailed, so only purchased a small number of hatch covers and other intricate shapes that would be difficult to make well. The items duly arrived and the quality is good. Was surprised by the weight though, so am pleased had embarked on making the other items from the usual materials. There should be an overall weight saving, along with a reduction in my surplus styrene and wood stock.
One of the design tenants of the Brave class was flexibility. The vessel could operate as a MTB, MGB or Raider, or with a mixture of these capabilities. The weapon mountings were designed to allow armaments to be installed and moved around to suite the requirements of the role.
Have reviewed many Brave class photographs trying to establish a “standard” armament configuration, to reproduce. Not only does the configuration define the weapons installed, it also establishes the ammunition and flare storage cabinet arrangements. Eventually decided upon the 2 x 40mm Bofors gun arrangement with 2 x 21” torpedoes and 4 x extended range fuel tanks.
The model is now essentially complete. No doubt as I keep examining it will add further small details and refinements.
Only disappointment so far is that it does not achieve the original weight target of 6 lbs, it is 9.5 lbs. The 6 lbs may possibly have achievable using one screw and motor etc., but once three are installed, not likely.
The real test is when finally back on the water. Will close this blog then with a concluding report.

---

After completing the cowl, turned to the rear structure covering the gas turbine and other engine spaces. This can readily be made from styrene sheet. The sides and top were cut out, reinforced with “L” shaped angle and fitted together with CA glue. No particular challenges, other than determining where the various section transitions occur. Luckily had two different sets of plans to compare, so the nuances could be established.
It was not until the rear structure was fitted into the cowl, the assembly fitted to the removable deck and placed on the hull, realized just how important this milestone was. Once everything is firmly located the accuracy of build becomes readily apparent. Any inaccuracies show up as an obvious misalignment.
Was able to check the alignments and squareness using eye, rules, squares and a spirit level and was pleased with the outcome. A subtle sanding of about .020” off the base of one side of the superstructure and everything became square, parallel and correctly aligned. Quite a relief!
Have always stressed the importance of accuracy throughout a build. This supported that recommendation.
Once the superstructure was completed realized my plan to lift the deck off to gain access to the electrical control switches was impractical. Have thus cut a small access hole in the rear deck to facilitate access. Still undecided how to best disguise the hole, but at least access is now relatively easy.
From now on, until the test program can be continued on the water, will add detail to the model. Doubt there will be much to describe is that of interest, or that has not been covered by others.
Will continue this blog once there is anything significant to report.
In the meantime, best wishes for Christmas and 2019,

---

Back to the build. Next milestone, to complete the superstructure and engine covers. The superstructure is essentially a cowl that supports the open bridge and serves as the air intake for the gas turbines. The engine covers fit into the rear of it.
The superstructure is full of curves and will be interesting to make. Still trying to save weight, decided to make it out of glassfibre. Rather than first make a plug then a female mould and finally the cowl, wanted to try the technique of making a plug out of styrene foam sheet, then covering it in a glass fibre matt. Once the glass fibre is set, the foam is dissolved out using a solvent and the cowl remains – inshallah!
To ensure the foam did not react to the glass fibre resin, painted the finished cowl with enamel paint before sticking the matt down. See pictures.
What a mess! The resin had crept under the paint and into the foam dissolving it. When the
resin dried the plug had shrunk slightly and had the surface finish of a quarry. First thought was to hurl it and start again, this time in wood.
On second thoughts, wondered if the plug could still be used. Decided to build it up with wood filler and from it make a female mould, as originally intended. The cowl would then be made from the mould. Built the damaged plug up and sanded it smooth. As the plug would be covered in fibreglass, the surface finish was not critical. Brushed a coat of fibreglass on the plug and, after drying filled any defects with glaze putty and sanded smooth.
Once the finish and dimensions were satisfactory, applied a thicker coat of glass fibre to the plug. This was again smoothed down, waxed with carnauba polish and then covered in mould release. From it the cowl was made.
Picture shows plug, mould and cowl placed side by each. The cowl requires reinforcement; the fittings and various mountings then adding before installing. A trial installation showed that it fitted properly the deck and was accurate.
A lesson for the next time is to make the plug and mould much deeper than the finished item. That will allow any rough edges, on either the mould or the component, to be trimmed off leaving a smooth fibreglass edge.

---