I was hoping for some advice on driving a bunch of servo motors for my mobile hexapod robot project. The problem is that the hexapod will be quite heavy and so I need powerful servo motors to be able control the system. I initially looked for bus servos as they seemed like a more reliable servo motor for use in robotics applications, but I couldn't find any that met my needs of a minimum of 70kg.cm (6.9 Nm) of torque, but I did manage to find a hobby servo that met those requirements. I especially like it because its available in my country, South Africa, which is not always very cheap (or simple) to import to. The servos I found are cheap Chinese servo motors and can be found at the following link:
I have absolutely no idea how much current these would draw, but I know that it will be alot and I need a way to drive them. After struggling to find something that meets the needs for that, I am considering just using an esp32 for managing the servos (I might need more than one esp32 to do it properly) with separate power coming straight from the LiPo battery that we intend to use.
Does anyone have any suggestions on what I could use to drive these. If so, could you perhaps give your suggestion as well as a simple explanation on how you could go about setting it up.
Thanks in advance, I would greatly appreciate any help that you could give.
I would presume that, like most servos, this device will take care of 'driving' the motor using the pwm signal. I don't see a spec for that signal, but it may be as simple as a 0-5V logic signal with negligible load. You need that signal specification. You'll need to insert a signal booster to raise the output signal of the ESP(3.3V, I think), to the level required by the servo.
The separate power connections from LiPo to servo are where the power flows. You need to tie the (-) of the LiPo to the (-) of the servo, and to the GND of your Arduino.
So I think you're looking for a power supply for your servos and I think your application is mobile (which is more expensive, to be sure). What lipo? 2S? I've never driven a servo straight from a lipo so I'm not sure how well it will work out for you. I don't want to trust what I say as according to Hoyle truth, but I have some experience with servos and such and I don't mind telling you that, at 500 rand taxes in (about $27USD), maybe you've found the Chinese servo diamond in the rough but I'm not convinced. Claiming 80kg/cm (1100 ish oz/in) and features like
a bottom cover and misspelled "precition" steel gears, and not brushless, I gotta say I'm skeptical.
Consider that Savox, a premium brand (I'm a satisfied customer in demanding rock crawler applications) for waterproof applications that get tested hard right away by their customer base (upgrade to hobby grade as in pricey as in picky customers RC rock crawlers) their servo that boasts similar specs, the SW2290SG-BE claims 972 oz/in of torque @ 8.4V, regulated by a BEC of some kind (probably by Castle Creations, they're the best in the RC business afaik, again I'm a very satisfied customer).
Now I'm not trying to tell you what to buy or anything like that, and I don't know what you can get in South Africa, I just hate seeing guys spend hard earned rand on equipment that won't perform as advertised.
I had a similar situation with a treasure chest I built (first one but still works, well works again) that was mechanically designed by a moron of mechanical engineering (me). Point is, I swapped out a 3500 rand servo (200 USD) because I was asking more than the 1/10 scale could deliver for my application. I ended up using a 1/6 scale Savox and it's wicked, has been for nine Hallowe'ens now.
How mission critical is your robot over time? Is it just a demo or school project? If so, I'd roll the dice on the less pricey option. But if you want the thing to be durable, I'd look for, shall we say, more reputable servos.
If your pockets are deep (either way, none of my business), and they deliver to South Africa, Holmes Hobbies in the USA sells monster class gear that can take a beating for ages. Consider this (at 1800 rand, I know, I know), direct battery powered as you indicated. https://holmeshobbies.com/servos/tug-buddy-shv500lp-brushless-winch.html
Thank you so much, youve given me a bit to think about. As you say, the servo is probably not going to be highly reliable. I was kind of hoping that because the seller is reputable, that the product would be too, but im inclined to agree eith you that thats an amazing produvt for the price.
Ill take a look at the Savox and do some calculations to determine how I can adjust the servos torque because I sort of just when for slightly overkill to be safe. But thanks for the detailed reply, it was quite helpful.
I've seen variations of that circuit multiple times, my issue is that it seems to have a limit of 1A. Now I just assumed that meant Icc, but maybe thats a limit to the control current, but I doubt that. Nonetheless, even if I were to use only one of those servos at half the torque rating, it would still use around 1.5 to 2A. Thanks for the suggestion though.
@jesseboise Back to your original post. To answer your question as posted, an Arduino Nano is perfectly capable of outputting 6 "servo" pwm signals you would need to "drive" these servos, provided the signal voltage level is compatible, which is what I was getting at earlier. The question that should be asked of the vendor - "what is the voltage and current requirement for the control signal?"
Now there is a separate question - what is the complexity of the application running within the Nano, or other controller, to accomplish what you envision? That may necessitate an ESP32 due to speed or computational complexity, certainly. But not because the 'drive' signal needs heft.
To @groundFungus point, there are controllers where you don't need to worry about servo current draw through the controller, such as the Maestro (someone in another forum thread was looking at these, maybe suitable for your needs?).
The Pololu Maestro says this:
If you connect a power supply to the servo power terminal and connect another power supply to GND/VIN, then the Maestro’s processor will be powered from the VIN supply while the servos are powered from their own supply. The VIN supply must be within 5–16 V and be capable of supplying at least 30 mA to the Micro Maestro or 50 mA to the Mini Maestro. The servo power supply must output a voltage within the servos’ respective operating ranges and must be capable of supplying all the current that the servos will draw.
The DS5180 80KG servo @jesseboise identified is sold on Amazon in the US for $40usd. One of the reviews provides some test data, including measured torque and current.
Fair. I also see that the reviewer with the luggage test said the size wasn't exactly what he was expecting, ie slightly wider.
As I said, maybe it's the deal it advertises, maybe my criteria for distrust in products is too low or misaligned, could be. I've been burned on shoddy products a few times so maybe that's what's informing my bias.
According to another review on the same page:
Works well to wave the arm on my Donald Trump mannequin I have set at the road
so I mean, on second thought, surely this servo can take a bit of punishment
I have to laugh about this because, I asked the vendor about a recommended driver board and they suggested the following board which is a 12-channel pololu maestro. I find this board interesting because it seems like I could also program certain walking paths into the board to reduce the processing requirements of the main board, although it might be difficult because the project will need to have an adaptive walking style.
I will probably use the pololu driver board, for my servos because it think it should be better capable of handling multiple servos simultaneously than the ESP, although I do see that it is more expensive than an ESP32.
are you suggesting that the two servos are the same? If so, great, at least that means I can put a bit more faith in them. I would personally be skeptical of simply assuming that the two are the same, so your confirmation is appreciated.
To further explain, I would be skeptical because I assume DS stands for Digital Servo, which would mean that these are "unbranded" servo motors. And I can imagine that if a Chinese company isn't even willing to put there name on it, then it definitely can't be a good products. But the two do seem to be the same.
My vendor also claims that they haven't had any issues with the servos, which does not necessarily indicate that they are reliable, but its at least something. So, thanks for the showing that, I'll take a look at the test data. I kind of expect it to perform below its rated stall torque.
Funny this particular company comes to mind for the second time today (someone else using a delta wing RC plane in another thread), your project makes me think of this one:
I'll point out that the Amazon listing has the same image @hallowed31 posted with the same spelling error.
The Amazon test sample as tested peaked at 2.8A and fell a bit short on max torque.
What are you confused about?
Before you answer, take a look at this:
Also, if you're trying to power this off the 5V pin on an Arduino, don't. Servos need their own separate power supply (and don't forget to connect all the grounds in your circuit)
