Any interest in a custom e-foil ESC?

That’s a lot tighter, that’s perfect.
If I can get one of these I will cut a hole in the metal tube the size of the vesc and weld a flat metal in it, to mount the vesc direct to the outside water cooled plate.

1 Like

Here is a pic of a unit that Nick sent me for testing. I have attached it to one of my custom watercooled heatsinks. I have the Metr pro so will be able to get full data from this and the 65161 120KV motor it will be tested on.

7030f7803a50a26f4f09b581cd2bbf9db7055522_2_690x388

8 Likes

Good looking for sure! Do a report as soon as possible!

What kind of data are coming out of the ESC ? I see at least 6 data wires. 3 for pwm, what’s the purpose of the others ?

Tx and rx , for uart probably

5v, pwm, gnd, 3.3v bec, rx, tx, canh and canl.

thx ! :slight_smile:

I had a successful first test with @nickw1881’s VESC today. No logs yet but the metr is in the mail. I could monitor temp/A/V on the remote and I didn’t see it much above 40C during our 2x 40min sessions today. I had to remove Nick’s pretty VESC case to fit it into an IP68 aluminum case which is mounted on top of the mast. Am very happy with that new setup so far. @peter’s 80mm direct drive prototype motor (Any interest in custom Direct Drive Inrunner - #45 by overdrive - Propulsion System (Motor, Gears) - FOIL.zone) is a quiet beast with that VESC in FOC mode. I will post more build details and logs when I get my metr.

8 Likes

That’s good to hear… Great you are back on the e-foil :+1:

I’m testing a similar setup, Nick ESC and maytech motor, first thanks to Nick for your ESC and help to configure. Well, I think my data is good but I have problems in 30-40min of ride. My esc stops and I need restart to continue, I dont have any problems with Maytech ESC in the same setup. I agree Merten, this FOC setup is incredible silent.

[34|690x456]

4 Likes

Not sure what is going on with @virus board. My guess is software issues with FOC and his particular motor. There are 6 other esc out there with various other forum members but @virus unit was not tested with this batch, and so I can’t say for sure it doesnt have a hw problem.

Several people have asked about getting more of these, and here are my thoughts: I designed for small size, packing in the parts as densely as I could. Because of this trade, the cost of these units is around $250. 3 pcb, billet aluminum case, cnc brass busbar, xt90 pair, and that doesnt count labor to integrate it which is about 30min per board once ai am practiced. Then in every batch one of them doesnt work, there is shipping to pay etc. This is all to say the design itself is too expensive since no one wants to pay over $300 for an esc when flipsky is selling them for $200.

Then, VESC software doesnt work for every design as we have seen in some cases. How could I recommend someone spend significantly more money on an esc when it may not work in their setup?

Maybe it makes sense to make a different design, one that is cheaper and easier to build. Maybe it’s a little bigger, but has the same heat via concept. That seems to work really well. I’d keep the busbar as well. But then the cost of a new dev effort is prohibitive at this point. Maybe the existing design is good enough (after I beef up the input capacitor board).

3 Likes

Following up with my train of thought, I designed a board per @pacificmeister specifications to fit in the enclosure he picked out. He mounts his motor controller on top of the mast in an IP68 3.5x4.5x2.25 enclosure. Since it is in the water when not foiling, it always starts cool, and then it can conduct heat down the mast when foiling. To simulate good conduction to the water, I used a large aluminum heat sink.

Here is my first test of this design at 200A motor (150A battery) for 30 seconds:

This design can handle 48V and 200A to the motor, but is much cheaper and easier to make than the last design. I believe I can sell this for $300.

12 Likes

Awesome work! why not 56 to 64 volts so rough estimation you reduce motor current roughly by 15 to 25%, This means also less loss on cables and higher efficiency, less loss on fets, easier thermal management. Yes high voltage fets are expensive but you will require less fets and can save on battery pack size (better efficiency) space and more …

100V fets are whats on there. The volt limit is only due to the 63V capacitors. At 60V, 200A the switching spikes put the caps just over the 63V limit. In the test I am using 5x 12V batteries in series, so it’s actually 60V at low current! But voltage drops by 10V when you put 150A through 8awg wire and 5 SLA batteries.

These are the biggest value capacitors I could get in this size to be most conservative for 1st test. I need to try some smaller value, larger voltage capacitors to validate the ripple at 14 and 16s.

Higher voltage isn’t a free lunch on efficiency, it just gets you more top end speed. Nature takes the same cut. Of course, speed is why we do this so I will be testing different capacitors and pushing the envelope soon.

This spike issue has been with vesc and I have thinking how it could be possible to damp it without going for higher voltage expensive parts. Whenever I get some time I will look in to it. I had similar issue on bms high side fets and used zener diodes to limit spikes across the pack + and pack - at the load side not sure if this is a similar issue

Have to clarify something for health and safety: This is not a VESC, and not just because I dont want Frank Trampa to sue me. VESC has lots of parts that are barely suited to the application. I am only using a modified version of the software. The gate drive and switch circuits are all my design.

Wavebreaker parts are all well rated for this application, and the board has been specifically designed for thermal conduction cooling in an ip68 case. It is immune to the sort of switching spikes that kill VESC drv830x drivers.

I only say what I have tested so far. This board will do 72V, 16s, I just need another golf cart battery and some time to swap in some more capacitors.

2 Likes

Even better :+1::muscle: iI didnt look vesc to see under what kind of license it has been released. It would be great to branch out the hw and fw to adopt it it completely for watersports

Did you have a link to the enclosure?

(I got all the answers by reading the whole post)

It’s definitely a long thread, I wish there was a way to edit it down some.

My last post may be confusing, so TLDR:
Over on eskate builders, now forums.eskate.news, Trampa has been very vocal that if it’s not made by him or vedder, dont claim anything VESC. Dont use the name, dont provide links to compiled versions of VESC tool OR ELSE.

I am using modified version of the vesc project open source code. That means it’s basically just like a vesc in every way (except way more powerful) but I will not always be up to date with the latest vesc tool, and my code is not included with the vesc tool defaults. So slightly less user friendly until I for vesc tool as well and call it something else