Electrical:
I am powering 4x M100 motors with 4x 20A ESC (Little Bee FVT 20A Pro 2-4S OPTO BLHeli) and a 11.1V Lipo battery (7800mAh, 40C) connected to the ESCs via an automotive relay (14 V 30 A / 12 V). I am using at least 16 AWG thick cables, 3,5mm bullet connectors, and have a fully charged battery. Wire length from battery to motor is around 60-70 cm. When running at full thrust in fresh water, one motor consumes max 1.2A and 12 W. I get the same output when running the motors dry. Is this nominal?
Mechanical:
All parts of the thruster are 3D printed. The propellers blades are sanded smooth. The propeller design is based on the T100 propeller (2 blades, 78mm diameter, around 20-25 degrees AoA).
Control:
I am using an arduino nano and the Servo ESC library to send out a 1.5ms - 1.9 ms PWM signal.
I was expecting more thrust like in this video. So I’m suspecting that something is wrong with my setup.
Looks like a fun project, that you’ve made quite a bit of progress on
That’s definitely not the expected behaviour. The T100 is from well before I started, and info on it is a bit sparse (which I’ve asked about internally), but even in the initial kickstarter page it had a rated max current of 11.5A, and max power of 130W. Your relay says it can support 30A throughput, so assuming all of that goes to the motors you should be able to get ~7.5A per motor.
On the thrust side of things, low voltage means low thrust. Using our Voltage Drop Calculator for the full 30A through a single 16AWG cable yields only a 0.54V drop, so that’s unlikely to be the issue. I haven’t used automotive relays before, but I searched for a 12V 30A one and found this one, which has transition times (operate and release times) on the order of 1.5-3ms, which I’d guess could be too slow to handle the power fluctuations for the ESCs (which are capable of changing the motor’s speed every 2.5ms, and are activating and deactivating the coils much faster than that).
It’s also possible this is causing issues. I’m not sure what firmware settings have been used for your ESC, but our BasicESC comes pre-flashed with bidirectional control, and “low RPM power protection” and “demag compensation” turned off to properly run the low kV motors we use. Our firmware settings are available on that product page, but they may not be directly usable for your ESCs since they have a lower current rating (20A instead of our 30A).
As with all projects, I’d suggest you start with a minimal setup and add components in as you confirm the basic functionality. Try starting with just the battery connected to one ESC + motor and the buck converter powering the arduino, and test the motor speed and current draw. If the thrust is decent there then your ESCs are perhaps not the problem, but to confirm that you can then try with more thrusters at once. Once you’ve confirmed that you’ve got thrusters and ESCs that seem to be working properly, then you can start adding other switches and relays and whatnot into the power path and seeing whether they adversely affect your performance
On something of a side note, you may be interested in this performance expectations breakdown I did a little while ago on using T200s for dive propulsion:
Success! The ESC firmware was the problem. Thanks for the input.
I changed the startup power from 9A to 13A and turned low RPM power protection OFF. I will try the Demag setting OFF aswell.
After these settings aswell as temporarily skipping the relay, I got a significantly better thrust on one motor and 13A current draw in max thrust. So now I can unlock up to 570 W of thrust.
Now I will do the same for the other ESCs and solder back the relay. I bought another relay to be able to run 2 + 2 motors with 26 A through each relay, since they are rated for 30A.
Are relays not usually used for an ON/OFF function for machines like this? In that case, please advise on other ways of doing it.
Personally haven’t required ON/OFF functionality beyond what can be achieved in software or physically unplugging the battery. I’m aware that some people use a physical switch, and others use things like the solid-state relay that @k-deboer suggested - the main challenge is making sure anything in the current-path has the capacity to handle the required current, and doesn’t excessively drop the voltage
It may also be possible to use a power-MOSFET, although that may be harder to attach conveniently. I’m not sure why I haven’t seen others trying that - perhaps there’s some issue I haven’t considered, or it’s in use but I don’t know about it, or maybe it just hasn’t been thought of.
