It depends. That's a pretty wide spread of voltages, so even an active balancer could take a long time to finish. If it was me, I would put all four in parallel and connect a constant-voltage charger set to 3.65 volts, and let it charge until the charge amperage dropped to under 1% of the cell capacity. So if they're 300Ah cells, I'd let the charger go until the charge current dropped under 3A. Even then, I'd let them absorb for at least a few hours.

Welcome, and thanks for watching. Sounds like a fun project you've got there.

Emulators are amazing for that. And ya, the world of IT has changed a LOT. I started in the 80s, and today is totally different.

Hah! Ya, it's a long slog, but it'll be worth it in the end.

Thanks, Joe! KZfaq is really strict with comments, and will auto-delete them if there's URLs or what have you. I need to decide if I want to replace these connectors, but if I don't, I'll look for those crimps. Thanks!

Thanks for watching!

Thanks for watching, Lou!

^_^

Hi Rick! The reason for the single controller is that the condition when I'd switch throttles and when I switch motors are independent. With the throttles, there will be one on the port and stbd side, so being right handed, I'd use the stbd side mostly, and only switch to the port throttle if there was a problem with the stbd throttle. Someone who is left handed would probably do the opposite. In either case, we might be running the 5kw if we're maximizing range, or the 10kw if we need extra power. Having one central controller makes this flexibility possible. That said, I'll make spare controllers, so if the controller failed, it should be a quick fix by just swapping the boards out.

Aye. I've already started pulling the rats nest apart. I'm working on adding the multiplexors now, and will make a prototype panel. I'm going to redo which i/o pins are used for what in the new version, and will wire things up to the prototype panels as I do so. I can't hope to debug the software when the hardware is so precarious.

Howdy! I spent a lot of time thinking about the throttle redundancy earlier in the project, and even considered the Apollo approach of three sensors (well, Apollo had three computers, but...). The idea would have been the discard the pot that was most out of range and average the other two. In the end, I decided to just have two independent throttles and a selector switch. This would mean that left-handed people at the helm could choose the throttle that felt best for them, also. If one throttle acted up, a flip of a switch and the backup throttle would be ready to go. So I went that route. The efficiency of the stm32 over the arduino is ... yeesh. I adore Arduino for now accessible it made embedded dev, but I also now understand the power and control of the stm32. They're both great, but for very different reasons.

@@TheDigitalMermaid Nice, neat solution. Yeah stm32 is better if you can get over the leaning curve, and now its getting easier since they're supported on arduino IDE you could port an arduino project onto it without issues and benefit from better specs and lower costs. That was my justification for getting a bunch, albiet the low space versions, they were 1.40 each and 1 arduino was 5+ so why not, space being equal it was 72mhz vs 16mhz, sold!.. But wow it took me some time to get over the learning curve of just getting a progrma onto one, and I have expereince with C and programming, so kudos to you whos learning all this from scratch, thats quite an accomplishment! Look forward to see how it turns out, especailly the graphical display. Theres some new stm32's that have vector graphic GPU's built in(STM32U5F9/U5G9) you know the old school vector graphics before pixels took over, that would be cool if you could make graphical displays with that and what that would evel look like.

@@mikejones-vd3fg So funny thing is, I'm designing an 80's retro-futuristic console, so the red 16x2 character display will be the "fanciest" display I'll use. It's mostly going to be a LOT of LEDs :)

@@TheDigitalMermaid Cool, yeah nothing like navigating the water with a retro terminal screen, maybe get an atari joystick for the steering? nah jk those gpu loaded stm32 do fancey displays it looks like but they'e expensive a$15-25 and thats just the bare chip, so lots more work. They just came out 6 months ago, if you were going for the Tesla dashboard look that wouuld be the chip to power it. Character displays are just fine too though, Marty McFly travelled back in time with LED displays, it should be good enough.

@@mikejones-vd3fg OK, so, I've actually be planning for that (using an Atari controller for the boat). I've not talked about it because it's so far away, but it's absolutely something I plan to do. It's one of the main reasons I want to learn more about the ESP32.

I've been poking around digikey looking at connector options, and that actually did cross my mind, as I'm not using 9 pins on any of the connectors. I dismissed it pretty quickly through as I really want to pick something that keeps the air out.

​@@TheDigitalMermaid Ethernet cable does come in an "outdoor" flavour that has sealant between the pairs which keeps water (and other things) outside of the cable. I am not familiar with if digikey keeps something like that in stock, and I cannot claim to know if that sealant will keep air out. I enjoy watching your videos, just trying to pitch in with the thinking process and hope you succeed in transforming the ship to all-eletric!

Twitching being the most apt part of that, haha! It's so good seeing it come together!

This is what it's all about! Unless you've tried something hard, you can't really explain that "oh crap, oh is it going to work?" feeling. But when it does (possibly on the tenth try), the euphoria is ... damn.

Oh, I absolutely plan to use terminal blocks. These though, the wires are too small guage and too short (coming off the controllers and motors). On the side I'm making, I'm using tinned copper wire and pins. I figure that I'll be doing annual inspections of all connectors as part of my maintenance on the boat, and if any connector shows signs of corrosion (for small signal wires like these), then I can just cut the plus off entirely, solder a connection and seal with resin-lined shrink tubing.

@@TheDigitalMermaid I think in the engine room, you will be mitigating a good amount of risk. It’s the 4 meters of extension and in the helm that will be really critical to mitigate the corrosion as best you can.

@@jimduke5545 Aye, and also, I was on the boat yesterday and I misjudged, I think it'll be closer to 2~2.5m, but still, tinned copper and then sealed everything!

That was part of it, but also, I was soooo tired by that point, I was scared something was going to go catastrophically wrong and put the project back. There's been so many delays... I just want to be making progress again. :)

Another thought.. To multiplex your switches, how about a USB keyboard PCB ? Many available switch contacts (Matrix) and its a HID device with Arduino library.

@@NeutronModulator I have zero doubt that a canned solution would be more refined and have more features than what I can build. What started all this though (and applies to the keyboard question) was that when I was looking for electric motor options, I was put off by the closed off nature of the options. At first it was being annoyed that they wouldn't charge third party batteries, but then the real kicker was that they'd only allow authorized service centres to fix their controllers and electrics. Well, that's going to leave me stranded is something breaks in the arctic or on some pacific island. So I decided that I wanted to build a hackable electric boat, and decided the only way to do that would be to develop a ground-up open source option.

If the control grounds are connected to the battery grounds, then when those wires get up to the STM32 command unit and come together it’s a point of connection between two large batteries.

@@jonathanbradshaw8473 All of the batteries go to common bus bars, and the loads connect to the same bus bars, so I figured all the grounds were connected already. With the motor controllers also going to the bus bars, their grounds are connected, too. The one thing I've not solved yet will be the onboard 5v supply (when it's no longer the laptop), but I figured I would ground the PCB / stm32 board to the system ground as well. Would doing all this not remove the concern? I'm still new to this, so pointers to docs / guides that help explain this is always appreciated. :)

I’m sorry I can’t do a good explanation. It’s a tricky concept.

Oh, some of my comment got lost. Possible references: en.m.wikipedia.org/wiki/Ground_loop_(electricity) help.campbellsci.com/CR1000X/Content/shared/Maintain/Troubleshooting/ground-loops.htm

All right, I’ve tried twice to put some links in a comment and had then disappear. Time for some typing… You can search for Ground Loop on Wikipedia for an introductory article. Let’s assume you’re using the 10kW motor at full power. In your specific case, bringing all the big battery negative terminals makes a lot of sense. You’ll have a length of chunky power cable going from the bus bar to the 10kW Kelly controller, carrying about 200 A. This will make some voltage drop. Additionally, this may carry some spiky switching noise from the motor drive switches. Meanwhile, the 5 kW Kelly controller uses basically no current, and hence no voltage drop. That means the ground on the 10 kW Kelly controller is at a different potential to the ground on the 5 kW Kelly controller. Next you connect your 10 foot (or whatever) control lines to ground at each Kelly controller and bring them together at the STM32 controller. According to the ground potential difference between the 10 kW and 5 kW Kelly controller, some current will flow (limited by cable resistance) along your control wires. The local ground at the STM32 will sit somewhere between the grounds of the two Kelly controllers. There are two things to watch out for: this unexpected ground current in the control wires could maybe get both enough to damage something OR when the STM32 I2C DAC delivers 2.5V out, the Kelly controller sees 2.5V plus some offset due to the grounds not being the same. I looked up a manual for the Kelly controller and saw that it showed +5V and a _separate_ ground connection to the throttle potentiometer. This separate ground would preclude the formation of a ground loop.

Oh, I very much plan to get everything cleaned up, but I've been really focused initially on a minimum viable product, so that once I can get the boat in the water, I'll be able to control it. It's the old "make it work, then make it right" approach. Of course, the danger with this approach is that you never get around to the second part. I'm hoping the community here will keep my feet to the fire so that doesn't happen. A benefit of open source code. ;)

AHAHAHAHA! Oh my god.

A statement that is not at all meaningless, but reflects the real state of affairs.

I just looked them up (assuming you meant Iwiss), and I see examples of the crimps they do. That looks exactly what like I needed.

IS THAT WHAT IT IS?! Holy shit...

@@TheDigitalMermaid Yeah, especially masking autistics have to watch speech cadence like a hawk when doing any kind of public speaking. The "weird" pauses always want to creep in. Lots of things to pay attention to at once and normies really pick up on it.

That's awesome to hear! It's the biggest reason we do DIY... All the frustration melts away the moment something works.

Aye, especially as I plan to do regen off the motors. If the load disconnected while the motors were generating power, the diodes might help there also. Now that I've got an idea of what they're meant for, I'll know what to read up on. I already had back EMF (I think) blow my buck converter and catch fire on my last year, so I should know better. Trick with learning so much so quickly...

@@TheDigitalMermaidthe contactor when it is de-energised produces quite a voltage spike that can destroy connected electronics, connecting a diode across the low current side of the contactor with the band to the positive side will help clamp the back emf generated by the coil field collapsing. Hope this makes sense

Duuuude, when I heard Kepler was sick, a pick part of me was like "are you serious!?". I'm kinda hoping I've paid my dues and can go without any more medical emergencies for a while cause I am so done.

@@TheDigitalMermaid I legitimately thought it was Catpler and not Kepler and Mr catters I am... Not good at smart or accents

@@tylerjw702 Hahaha! Ya, named him after Johannes Kepler. Also, I love the idea that I've got an accent to other people. ^_^

@@TheDigitalMermaid just past the part where the motors spun 👏 legendary!

Thanks for watching, Lou!

I never thought about using diodes, but now I'm extra glad I went with two ULNs!

Absolutely! The issues I've got are minor and fixable, I was very happy.

I think you're right, actually. The kit I got (and linked) even says "Dupont" on the case. I'm not sure where or why I started calling the JST.

ME TOO! Didn't think you would jump on it, with your focus on access and overview, so didn't mention how to. You like woodwork, so make a frame, holding the water block, that the controllers in turn mount on. In the shape of the frame you can integrate a mount for whatever structure you want to secure the controllers to. Like someone else mentioned, when you get time, you should sheld some of the wires from EMFs. I don't like the sound of those motors. Nice progress 🎉

Haha! Deservedly so, I loved that idea. I mentioned your comment to a friend and he said "You missed an opportunity to say "Quoth the Raven..." and now I hate myself. Haha

@@frejaresund3770 Aye, a wood (white oak) frame is what I've got (loosely) in mind. Some sort of cam lock to allow them to be separated, but when closed to press in with decent pressure (and using a thermal compound).

Using diodes makes a lot of sense, and I'm kinda annoyed I didn't think of that. Well, thankfully I had two ULNs and that worked.

Aye, I can see that. My friend has a pick and place machine and a solder bath, so the board I make for the boat will hopefully be well soldered. In any case, using the liquid flux should help as well. I really worry about lead, but I do know it was used for a reason...

Oooooh, I've got some serial to parallel shift registers I bought to play with. I originally planned to use them to drive LEDs, but ya, they might be perfect to also drive the A0 pins! I may have enough pins when I'm done, but if not, I'll remember this and pull it out to free up a few more. :D