Welcome to the Tweaking4All community forums!
When participating, please keep the Forum Rules in mind!
Topics for particular software or systems: Start your topic link with the name of the application or system.
For example “MacOS X – Your question“, or “MS Word – Your Tip or Trick“.
Please note that switching to another language when reading a post will not bring you to the same post, in Dutch, as there is no translation for that post!
I am looking for help designing the electronics of a robot dog (quadruped). I have basic circuit design experience, but I want to get a second opinion before I purchase all the components (and potentially fry my Pi). Specifically, I'm concerned about the use of two voltage regulators in my design.
My proposed circuit design:
Components:
Raspberry Pi 3B+
MG996R Servo (x12)
Spec Sheet:
Purchase Link: Amazon Canada, Amazon US, Amazon Germany for those in Europe
20A 300W Buck Converter
Spec Sheet:
Purchase Link: Amazon Canada, Amazon US, Amazon Germany for those in Europe
Justifications:
- Each of the voltage regulators I chose is rated for 20 amps max draw, so I figured I would have to use 2 for this project. This might be a stupid idea (will this even work?), but I found it difficult to find any voltage regulators that could handle the total current draw of this system
- Each servo draws a maximum of 2.5 amps at stall, so in a worst-case scenario, they will draw a maximum of 30 amps. Split between the two regulators we should be well within the manufacturer's specs (they recommend sustained operation to be at 15 amps or less)
Question:
Ideally, I would like to be able to drive my servos closer to their max voltage (7V), but I am concerned that I will fry my Pi through the GPIO connections that control the servos. (I understand the power and control pins on the servos are isolated, but I have heard horror stories). Additionally, I have considered using a PCA9685 Servo driver, but I could not find information on the max current that this hat can handle and it seems kind of redundant given the Pi can control the 12 servos with the built-in GPIO pins.
Given all this, would it be possible to step down the voltage to 7V using my existing Buck Converter setup, then step the voltage down a second time for just the Pi?
More importantly, is my current design with two buck converters even feasible?
(updated by Hans: made PDFs permanently available for future reference, since new users cannot right away upload attachments, and added links to Amazon US and DE)
I hope you do not mind that I made the PDF's and Circuit Drawing locally available (for future reference - I know, as a new user you cannot post attachments yet).
Unfortunately my experience with Servo's is zero (for no particular reason, just never got to it).
As for frying your Pi; I suppose that is a real thing that can happen -- 2x 20A is a lot.
Not having any serve experience here, but are there any (affordable) servos that run on a higher voltage? (your Amps should reduce proportionally)
@hans thanks for making the PDF's locally available!
Increasing the voltage is an interesting idea I hadn't considered. According to P = IV it should work if we hold P constant.
However, the servos I am using increase their torque when we increase their voltage. Thus the P in P=IV is not held constant. For example, if you reference the data sheet for the MG996R servo I proposed in my design, you will see that the servo outputs 9.4kgfcm at 4.8V and 11kgfcm at 6V. Because of this the max current draw is likely unchanged.
I did shop around a bit for higher voltage servos, but I wasn't able too find much. It seems like the next step up are at least double to tripple the price.
Context on voltage and servos: https://www.helifreak.com/showthread.php?t=655292
I did see the prices going up pretty badly indeed when one tries to find higher voltage servos.
I wonder why that is. Considering some ridiculous prices, I'd expect the torque being much higher.
Maybe a crazy thought, not really knowing much of servos ...
So each servo needs 6V and pulls at most 2.5A. Thinking about voltage regulators like the 7806, then we can maybe find one that can handle 2.5A, so we can get one (cheaper) for each servo? Note: I know a 7805 only handles up to 2.5A, so I'd assume the same goes for the 7806 (never used that one).
p.s. interesting read related to this topic at Howtomechatronics.com. Bumped into that one when trying to find an answer for you.
I got some extra advice that I thought I'd share here:
Turns out the PWM signal produced by the Pi's GPIO pins isn't high enough to control these servos! I am adding in level transistors to up the voltage from 3.3V to 5V PWM and an additional voltage regulator to give the Pi and level transistors their own power supply (this will protect them from current spikes and noisy current fluctuations). Moreover, a third voltage regulator has the added benefit of letting me run my servos at their max voltage of 7V!
It seems like having multiple voltage regulators is ok as long as they are not connected to each other downstream.
Updated Circuit Diagram:
Re Hans: I did consider your idea of giving each servo its own regulator, but having 3 larger ones seems like the more economical and less complex method. Thanks again for your support in this, I'll add updates to the project here as it grows!
(update: Hans attached the new Circuit Design as a PDF file)
GPIO power:
Ai! That would be important to know indeed. I didn't even think about that. Nice fix though!
As for regulators for the servos;
Agree! If you'd have to get something like Buck Converters for each servo, then economically your initial option is definitely better.
If you could get away with a $1 7806 (for each servo), then this may change things.
You will however have to find a 7806 alternative though that can pull 2.5A (assuming you'd really using the "hold" position of the servos, in your application).
