A rotary switch with a few different modes allowing for different maximum angle tolerances would be interesting, so that anyone could borrow your board ?
You may take advantage of a few ardupilot for APM drawings :
This is a good idea, but probably for the next rev. I’d likely put a BT link in so I can adjust all settings from a phone in realtime. Depending on how the first few test rides go it might jump up in the queue. If I find myself on shore all the time with a laptop re-flashing boards then it’ll bump up fast 
Yes, avoiding to have a computer on the field is exactly the reason why there are RC switches for those settings on Ardupilot.
For one of my esk8’s I just got a FOCBox Unity and I was pleasantly surprised to find out that it had BT built in and came with an app for your phone which did exactly what we are talking about - on the fly change of the settings.
It would be great if some of us could create a group to build the Phone app and interface to Bluetooth. I say this mostly because Android UI is not my strong suit
A couple of shots of the CNC mill cutting the slot for the mast…
The black sharpie lines have nothing to do with this project - this particular piece of aluminum was once used as a form for making curved diffuser lenses in the toaster oven.
The question is, was I accurate enough in my alignment that the nosecone will fit correctly?
I can’t print the nosecone until the water filter comes in and that’s not until late next week.
So I’m working on the battery. I’m going from 7P to 9P so I need different carriers and to order another $200 worth of cells.
2 Likes
I’m super-intrigued by these folding props mentioned in another thread.
I think Carbon Fiber blades mated to CNC’d brass gears, then encased in a Delrin base might be within my skillset (my skillset does not include casting nor does it include 4 axis milling - yet). But not until rev 2.
2 Likes
Doug,
I would like to know more about your batteries. What are they and where did you get them.? Everyone else is going with 18650 batteries. Those round things you have the batteries mounted in? Did you print those? So why round? Seems like it would waste space?
They are 20700’s. from Lithium Ion Wholesale. Only slightly bigger than the 18650’s but with 25% more capacity. In my case the Pelican case I chose was much bigger than needed to be for the original battery pack, so space efficiency wasn’t really a concern.
The round things are kind of a historical thing - when I made my esk8’s I built the frame out of extruded rail so it made sense to custom design battery carriers that could integrate into the design.
The round ones turned out to be a bad design mostly because connecting them together was going to be more trouble than if I had just built ones with normal tabs on them that I could screw together. As such I’ve abandoned those for a new design:
I like this method because it modularizes the build process and I don’t have to do fish paper or giant heat shrink tube or anything like that, and I can add, subtract or replace Parallel sets any time I need to.
The other, purely theoretical, advantage to these is there’s space between cells and between banks which might aid in cooling and might, if there’s a runaway, limit it to only 1 bank.
I CNC them out of Starboard (HDPE) from TAP plastic.
The downside to this design is it’s a double-sided cut, and I’m really only a single-sided CNCer. So I’ll have to figure out how to trim the mating tabs from the other side. Probably manually mill them unless I get inspired to actually learn how to register parts for double-sided cutting. Which probably isn’t that hard.
3 Likes
Thanks for the information. I need to crunch some numbers to see if those bigger batteries might work in my application.
1 Like
20700 are the best size vs capacity and weight but they are still a bit expensive
I tested the same ( not on a efoil) in 18650 size : GA , good capacity as long as you stay around 8-10a discharge constant rate i think
Don’t Forget that the capacity start to drop to much if the 18650 cell is used a high discharger amp , at a low rate 5a , with GA i get a better capacity than 30q , 25r , vtc5 for sure , but after ?
Yes, they are not cheap. All in I think the battery box is going to be right at $1,000. But it should last for a long session on the lake. Esp. while I’m learning and up and down a lot.
My plan was to build another pack so I could swap packs mid-day, but that’s a phase 2 thing as well.
Rev 2 of the receiver board. This one has the tilt-pitch-roll sensor and is very likely the one that will go in Phase 1. I will add a bluetooth receiver for Phase 2, but that’s just too much code for now.
This new board also has hardware to drive the alarm, and that tested fine as well. It’s painfully loud, but once it’s encased in the battery box it won’t be nearly so bad.
I also tested out the basic functionality of the display board (IE I lit up some LEDs) and it looks like it will work fine (the LEDs light up). I worked through all the information that has to get to this board and encoded it into a packet format. So the next task will be decoding the packet, and lighting the appropriate LEDs.
I was originally going to have the limit logic in the display board, but now I think that’s totally the wrong way to do it. Because I will need to act on some but not all events on the Receiver board I might as well just do all the limit logic on that board and then have a very simple packet that just tells the display board which LEDs to light up at any given point in time.
There is still a lot of thought to go into under what circumstances the machine should shut down, and how exactly the shutdown process is going to happen for each circumstance (I don’t want to just go to zero throttle instantly because that will toss me off the front of the board and really annoy me, so I’d rather it throttle down slowly).
At this point I think I’ll just have a beep and a red LED warning but not shut down for most warning events (battery box temp, driver box temp, coolant flow, etc.). The only events I’ll shut down for is excessive pitch or roll (in which case it does need to cut throttle immediately); lower battery limit and remote signal lost (both of which throttle down slowly). The rest will be some combination of beeps and lights but I don’t want it to be shutting down all the time.
Seems prudent to have another battery limit at which it throttles back, limits you to half-power to allow you to pooter back to shore. Limp-home mode.
The final rev of the prop extender is off to the 3D printer and I machined down the prop so I could bolt it on.
The only other really big 3D printer project is the nosecone, which I was going to print, but the printer estimated 1 Day and some number of hours. I might have to do the nosecone in segments.
2 Likes
Hi Doug, what mill is that?
I haven’t checked in in a while - mostly because I’m predicting a very early spring and I’m trying to get all my esk8 upgrades wrapped up before the weather warms up; and there’s a ride through the new waterfront tunnel that esk8’s are allowed in and I don’t want to miss it.
Anyway, I’ve been working on the battery pack. I did learn double-sided CNC milling. It’s not that difficult if accuracy isn’t too important. In my case being 10 thousanths out is ok. The trick is to drill mount holes the size of the cutter, then use them as registration holes for the other side.
So my battery cages are going well.
The bad news is because I had so much room in the battery box when I was doing 12S7P it didn’t occur to me that 12S9P would use up so much more space. But it does. Now I’m concerned that I won’t be able to get the rest of the components in the box.
I knew I should have modeled the whole thing before I started cutting stuff up, but I never thought it would be this tight. I might not make it. Might have to go down to 8P.
The mast is almost part complete except for the one giant 3 day 3D print that I’m afraid to start.
Oh - and I have a sort of an unfortunate distraction. The last few times I’ve used my battery welder (Sunkko 788H) it hasn’t been able to put down a good weld despite my dialing everything up to 11. So I’m headed two directions. First I want to see if @Mac’s strategy of buying the 220V 709A works, in which case I’ll follow his lead.
But I’m also assembling parts for one of those simple lead-acid-starter-battery-Meets-Arduino welders, which appear to work quite well. I’ve got some copper bar in, a starter solenoid and a battery scavenged from my dead (but not irretrievably so) scooter.
I can probably cowboy something together over the weekend and do some test welds to see if it really works.
Anyway, piccies this weekend hopefully. Unless the weather’s really nice in which case I’ll be testing my esk8 upgrades.
I love esk8 too. Got my Unity sitting in a box waiting to get used in one of my boards but I’m soooo close to having my efoil ready!
I too have a battery problem. I am probably going to buy the 2 6S batteries that pacificmeister uses since it’s proven. But man, I would love to build a battery pack but the work seems overwhelming.
1 Like
One of my boards runs a Unity. So far it’s a really nice driver. The mountain board had 2 flipsky’s but one of them died, so likely it’ll get a Unity as well.
I originally planned on the LiPo packs, but it didn’t seem like they were going to give me enough run time.
I’m hoping with the 9P of 20700’s that two people can spend a couple of hours learning - which means most of the time in the water rather than on the water If, after that, I can get an hour of time on foil then I’m perfectly happy.
I have 20700’s in two of my skateboards. I am actually using 10S4P of the 20700’s for my efoil testing. I have heard people mention to me before that the 20700’s didn’t put out enough discharge or something, but maybe that’s because mine are 10S. I would LOVE to put together a 12S 20700 battery pack, seems kinda expensive though!
The 20700’s I have are 15A max, so I can see why people are concerned about discharge current. On your 4P you max out at 60A, which might not be enough to get you on foil. I read somewhere that it takes 80A. But presumably that can vary widely with the efficiency of your drivetrain.
There are some beautiful 4200mAh 21700’s with 45A capacity, but they are really expensive.
I have 20700’s on my MTB, but I haven’t been able to test them much because about the time I was building that pack was about the time my battery welder was dying. So I’m suffering bad welds.
Yes, definitely not cheap. But I think battery prices are going to start coming down soon - hopefully before I want to build a second pack.