Hiorth brothers returns - waist mounted foil assist build - two approaches

Me and my brother was quite active on efoil.builders some years ago, building two regular e-foils. Now we are building wearable foil assists for ourself, with two somewhat different approaches.

One Pack raft (primary path) approach – using a Aluminium or Stainless frame glued to waterproof fabric with a zipper or just folding it to make it waterproof.

One more conservative approach (backup) using a off the shelf waterproof box.

ESC: flipsky 75200 Pro
Motor: Flipsky 6834 w. alu prop
Foil: Gong curve H XL and Axxis 1010 with 70-80cm mast
Power cord: Polyurethan 3G2.5mm2 or 3G4.0mm2 Custom coiled
Off shelf box: B&W Type 500
Battery configuration: 12s3p 21700 Molicel 42A or equivalent
Busbars: Custom from Wellgo Battery Parts
Controller: Possibly Wired, TBD
Cable glands: Likely customized - will be shared once design is ready
Harness: 3D printed TPU

We are found of illustrations, so we will just tell the story for each image as we move along. We have started prototyping and ordered some materials, hopefully will have something functional over the summer. As our previous builds, we will share 3D files and drawings of everything when we have something that works :slight_smile:


Here is the Packraft approach, Its a metal “backbone” made of sheet metal and Waterproof fabric (Packraft fabric and/or dry suit fabric). The fabric is glued onto the sheet metal then made waterproof with a zipper or just folding it over on itself (like regular waterproof bags). The main idea here is that the metal backbone will act as a large heatsink and dissipate heat the battery as well as the ESC. With plenty of room for wiring and perhaps some refreshments :slight_smile: etc.


Here is the off the shelf box approach. Its similar to several builds we have seen here on foil zone. Heatsink for the ESC is milled through the side, the green part will be milled in aluminium and bolted to an external heatsink, clamping the cutout, preferably, waterproof. the box is only 500g and can just barly fit the ESC and the battery. Which we think is nice for waist mounting. The box also has these external fins, that will be nice to protect the Cable glands.


Our Battery will be 12s3p 6x6 21700 cells. We ordered 5x custom copper busbars with pre mounted nicklecopper weld strips from a shop at aliexpress - they look great! For next order we will move the balancing wire attachment as shown with the red dot, to prevent the battery from being wider from the balancing tab. We ordered a K-weld with a car battery, and started playing with welding. Having almost zero experience with battery welding, we thought we found a decent setting of 45J. At least Gemeni told it was approporatly dialed in.


At the day of welding the first pack something went wrong :upside_down_face: After welding all the cells at 45J we found that at least 3 cells was punctured and leaking. It most definitively was user error, most likely too much force on the electrodes, incorrect/different calibration or different voltage on the battery (if that matters?). The battery was charged overnight, and the calibration was done holding the electrodes slightly different the next morning. In hindsight the welding should have been done immediately after doing the test welds. If anyone have some good tips here that would be appreciated. We are considering making a spring loaded mechanism of some sort, or weld under a microscope to make it more controlled. In any event - we have many more cells so we will just start over, and implement some improvements observed in the first build.


For balancing the battery our original thought was to have a charge only BMS inside the battery. To save weight and pack-build time we will make/get a balancing charger for 12s. Using a gold plated amphenol vga (15pin) connector as the balancing connector (+ temp sensor). The current idea is to just take the daly smart BMS and permanently attached to the charger and use it that way - with a customized connector to prevent improper connection sequence.

Cheers

3 Likes

Good to have you back with a new project, I still use the remote for which you gave me the file of the case.
As for spotwelding the well go copper nickel strips, I weld them with 25J with Kweld. You might have to re-sharpen the electrodes on a lathe or drill press with sanding paper. Do some tear off tests on spare or old cells.

1 Like

Thanks, will fun to build again! Awesome that the remote is still in use. We will definitely start lower on the K-weld next time :sweat_smile: tested all the way up to 80J without blowing through on some test welds on the minus pole, so was a bit surprised that 45J was enough to go through. Nevertheless a good but expensive lesson.

How many welds do you do per strip at 25J, and how do you hold the electrodes? Not that much space down in there.

1 Like

Hey, nice to see you around. I was looking in to your remote the other day actually I am looking to design one my self. Did you ever share all the files for it in one place? Cheers

Welcome back :slight_smile:
The post about your DIY remote back in the days was one of the inspirations for me to start the BREmote project :slight_smile: If you want to try the new BREmote V2 nano for your assist let me know! I think it would fit nicely in terms of size/shape and functions. @sat_be is also testing it on his remote currently

1 Like

I’m surprised you blew the cells with 45J. What material and thickness is inbetween the copper to which you weld the cells?

On 0.15 pure nickel I once blew a cell (VTC6) with only 30J

Damn! That is fast. Needed 50-60J to even weld 0.2mm. Might have been 0.22idk.

Which spotwelder are you using?
I assume the less expensive models may not be too accurate about the energy… and kWeld the energy is precise but peak current also changes the behavior, with bad input sources (peak currents <1000A) you tend to need more energy because the longer weld times mean the heat “runs away” from the weld if that makes sense

Makes sense. It was with a Kweld. Roughly 1150A.

Which weld do you recommend? Can be slow, but needs to be safe.

Odeer09 - I think its on here somewhere, if you cant find it, send me a PM and I`ll see if I can dig it up form the archives.

Ludwig_bre That is awesome, very cool to hear that you got inspired and are still going! We would be happy to test the nano. At this time we are leaning towards using a wired remote, as we already will be tethered to the board. Killing to motor logic mechanically through a wire seems more safe, also no need to charge the controller and deal with potential connection issues.

DutchFoiler - its a 0.15mm Nicklecopper fused sandwitch. We get 1500±100A from from a AGM car battery using the K-weld.

Newguy - I dont know much about spot welding, but I think the consensus is that K-weld is the best for small scale battery welding.

2 Likes

Looks like a great build - curious why you went Molicel P42A - a group of us have been using the P50Bs and they are much better thermally as well as offering longer runtime?

Junction box in mast mount, or in the board?

Printed a modified version of Andrew W’s zip-tie design found on Printables - great design! ( dont know if you are here on the forum)

I tweaked it a bit to have a junction box in the nose cone, but is this actually better? Currently planning on running a ø11 OD PUR jacket cable with 2.5mm² copper.

The thought was that having a single cable run up to the waist pack would be cleaner, but I kind of still need an anchor point in the board anyway for strain releve.

Having the 3-phase wires run directly allows for a sleeker profile, but those can be a bit annoying and more easily damaged than the tough PUR jacket.

Anyone gone down this road already?

also 3d printed tpu cable gland inserts?
i orderd a few but they came with the wrong compressino element so i wanted to try and 3d print a better one but is it actually better?

image

image

also here is some progress on the stainless steel back piece and TPU waist strap!


will try and bond the tpu fabric to it and insert the cable glands over the weekend.

tpu printed backpiece actually fits very comfortable. testing it here with a few rocks.
i also sunk it down to around 3m depth eventually most of the infill got satureated so i am considering making it so that it drains rather than having closed innfil.

Thanks :slight_smile:

We started with molicel 42A just due to its 50% cheaper than 50B with only 20% more capacity. It was also not in stock from Nkon at the time.

Also cheaper to mess up, so saved a bit already😅 will build multiple packs so will likely build packs when things are figured out with the 50B.