Hello, for my next project I'm going to build a mirror which consists of 9 smaller ones. For that I've bought 9 TowerPro SG90 servos. I need your experience to check If I planned everything. Please tell me If all components are suitable.
All the parts that I want to go for: (I provided links to polish shop, from which I want to buy from)
data transmision through CH340. I've heard that people have problems with this clone, but choosing arduino nano (old) from the board manager in the IDE solves this problem
I want to go with a board that has some sort of usb to upload code fast
this board uses only 4 pins from the arduino. It's better than using 9 digital pins straight forward from nano, and I know that pins have limited amperage
It has radiator, which has to be mounted when using high amperage
9 servos consume max 9A
this guy has also 8A converter, but it is momentary 8A
I suppose that lower amperage will occur in servos being slower or something. Servos won't be operating all at once, so i think 5A will be okay. Moreover servos will move max few degrees.
Servos will operate at 5V
[b]Things to make clear:[/b]
Arduino will use power from 4 alkaline batteries
Servo controller will use power from the 18650 cells
Here's a look at my 3D model of the case:
Thank you for helping me, I really can't tell how much this forum helped me.
The stall current for the standard SG-90 is about 0.75 A, and it briefly draws that every time it starts moving.
Multiply that current by the number of servos that will be moving at once to get the minimum acceptable current rating of the servo power supply. It is best to use a separate power supply for everything else.
On a side note: My experience with the plastic geared servo is that they last 7ish days under constant torque conditions. I have 2 metal geared servos that have been running for 11 months continuous.
I power the metal geared servos with their own 5V power supply, I account for 1 Amp per servo.
You can make, with your spare batteries, a servo power supply.
You could use several of these Amazon.com to generate 7.5ish volts and then run the power to several 5V regulators, especially if you have a few 18650's hanging around, and the regulated 5V to the servos.
The topic has been moved to another section. After I edited my question, I need to reply so my topic is not on the 3rd page. All the informations are in the topic. Thanks
Not exactly knowing the ultimate goal of your project, it sounds interesting.
I think the development part shouldn’t ne too hard, you’ve picked up most of the initial questions..
I’m curious about the mechanical construction and requirements.
Particularly the precision of the angular servo movement translated to the mirrors.
If using the Arduino servo library one can use servo torque micro seconds, where 500 uS represents 0 degrees and 2500 uS represents 180 degrees, about 11.11 uS per degree, which would be more precise then using degrees to torque the servos and more resolution than a servo can respond.
lastchancename:
I’m curious about the mechanical construction and requirements.
Particularly the precision of the angular servo movement translated to the mirrors.
Idahowalker:
Oi! Good point about movement precision.
If using the Arduino servo library one can use servo torque micro seconds, where 500 uS represents 0 degrees and 2500 uS represents 180 degrees, about 11.11 uS per degree, which would be more precise then using degrees to torque the servos and more resolution than a servo can respond.
I don't have much experience with servos and I don't quite uderstand what you're talkng about. Can you please provide me with any tutorial or explenation Sir?
I actually wanted to use Adafruit servo library. Before the parts come, I'm gonna study the documentation and watch some tutorials. Before puting parts together, I want to experiment with one servo, test the code etc.
I understand that this little, cheap sg90 servos won't be precise:/ But since I'm interested in this kind of stuff, I want to develop my skills. Maybe someday I will do a mirror consisting of 500 servos
Here's how I'm gonna join servo and mirror: (using bracket or buckle, don't know how to call this. The tricky part is to join it with servo, I will test some strong glues)
The point I was making, apart from repeatability of the servo positioning accuracy, was the ‘gearing’ between the servo angle and the mirror displacement.
To examine this, you need to understand the relationship between the servo arm radius (arc), and the resultant movement of the mirror (depending on the method of mounting and fixing).
If you need very precise mirror alignment and matching, this will be critical to your mechanical assembly and programming.
If a servo doesn’t reach your spec, perhaps a worm drive- which will be slower, but can provide near infinite accuracy.
Servo feedback may then become an issue, which may require a separate encoder - rather than the one included inside the RC servo.
I have another question.
This servo controller uses I2C protocol. For future development, Is it possible to connect I2C display at the same time? Or do I have to search for a display that uses SPI or GPIO?
But still SPI will be used my nRF24L01 module, does these two will interference?
lastchancename:
yes, both can be connected. they have unique i2c addresses.
That's great news!
Below is the final scheme of the connections. Display is only an option, but it's nice to plan ahead, right?
I'm gonna order all parts tomorrow. I will update you about progress or obstacles. Cheers!
4 fresh alkalines will put out about 6.4V, if you connect that to the Nano's USB, the onboard reverse polarity protection diode will drop about 0.4 volts, still putting 6V on the processor, NOT good. Can you put an external diode (1N4004) between battery + and USB socket to drop another 0.65V?
It may be better to use use 4 x AA NiMH rechargeable cells (Eneloops etc). 4.8V nominal, around 5.6V fully charged. Better voltages and you don't have to keep throwing them away.
slipstick:
It may be better to use use 4 x AA NiMH rechargeable cells (Eneloops etc). 4.8V nominal, around 5.6V fully charged. Better voltages and you don't have to keep throwing them away.
JCA34F:
4 fresh alkalines will put out about 6.4V, if you connect that to the Nano's USB, the onboard reverse polarity protection diode will drop about 0.4 volts, still putting 6V on the processor, NOT good. Can you put an external diode (1N4004) between battery + and USB socket to drop another 0.65V?
Seems reasonable, thanks! I thought also of powering nano with 9V block battery. I found this in google: Given that a typical Alkaline 9V block has a capacity of approx 450mAh, this means it drains the battery in just 30 hours or less two days (450mAh/15mA)
Also I need to keep in mind that nano will give out power to the display and NRF module, so I need to find proper source of power. For now I will test nano using 5V from PC. First, I need to order parts :>
