Looking at powering 32 Nano's from a single power supply through the vin. I'm looking at a 9v 11A power supply. If I run these in parallel from the power supply are there any considerations I need to make since there are so many of them? From what I've researched so far It seems like that's enough power for everything, but I just want to make sure I don't need any extra circuitry or anything to make this work.
Thanks in advance.
I would add a decoupling capacitor across each nano.
If you can get a lower voltage that's better - ideal for powering off Vin is ~7v - though depending on what kind of load you're planning to pull from the 5v rail on each nano, it may not matter. With a linear regulator, the excess voltage (at the current of the load plus a few mA for the reg itself) is dissipated as heat. at 9v, you pull 100mA, then 4v at 100mA is being dissipated by the regulator - 400mW. At 7v on Vin only half that much heat would be generated in the regulator. I would be surprised if that regulator, on a nano, could dissipate more than a 1W sustained.
You can also power the boards with 5V, connected to the 5v/Vcc rail.
Decoupling cap across each nano isn't a bad idea, though they do have an input cap on the regulator, but who knows what value it is (assuming it's a clone). Whether you need it probably depends on whether any significant loads are being switched, and how long the wires from supply to nano are.
A few fuses on the power supply output would be a good idea. 11A shorting makes a pretty spark.
xxx
Thanks so much for this, one or two questions if you don't mind:
DrAzzy:
If you can get a lower voltage that's better - ideal for powering off Vin is ~7v - though depending on what kind of load you're planning to pull from the 5v rail on each nano, it may not matter. With a linear regulator, the excess voltage (at the current of the load plus a few mA for the reg itself) is dissipated as heat. at 9v, you pull 100mA, then 4v at 100mA is being dissipated by the regulator - 400mW. At 7v on Vin only half that much heat would be generated in the regulator. I would be surprised if that regulator, on a nano, could dissipate more than a 1W sustained.
DrAzzy:
You can also power the boards with 5V, connected to the 5v/Vcc rail.
5v would be easier to find than 7, so this seems like a good idea. If i use Vcc are there any considerations about having a high amp power supply connected this way? I've seen notes on this in doing research, but can't seem to find a straight answer. Theoretically it should only draw what it can draw, but I've read posts where folks say not to run high amp power supplies into Vcc. I just don't understand why not I guess.
DrAzzy:
Decoupling cap across each nano isn't a bad idea, though they do have an input cap on the regulator, but who knows what value it is (assuming it's a clone). Whether you need it probably depends on whether any significant loads are being switched, and how long the wires from supply to nano are.
So just thinking about this physically there's a chance that some of these will be fairly far from the power supply, any idea on lengths I should be concerned about? 10ft? 20ft? or are we talking about the difference in inches?
again, appreciate the help quite a bit.
If you are using an analog pin to measure a voltage, the result will be different for each Nano.
Paul
Paul_KD7HB:
If you are using an analog pin to measure a voltage, the result will be different for each Nano.Paul
I'm just talking about powering the units themselves, not measuring voltage. I'm not sure what you mean by this, am I missing something? Honestly asking, is there something I'm not taking into account here that I should be?
aintmichael:
I'm just talking about powering the units themselves, not measuring voltage. I'm not sure what you mean by this, am I missing something? Honestly asking, is there something I'm not taking into account here that I should be?
I know you are just powering them. But the A/D is usually dependent on the input 5 volts. If it varies from Nano to Nano, then the resulting measurements will be different by one or more bits. So, the program must be calibrated for each.
If not using A/D, then no problem.
Paul
Paul_KD7HB:
I know you are just powering them. But the A/D is usually dependent on the input 5 volts. If it varies from Nano to Nano, then the resulting measurements will be different by one or more bits. So, the program must be calibrated for each.If not using A/D, then no problem.
Paul
Oh, gotcha. Thanks, that's something i didn't think about. Appreciate it.
It would be more efficient to supply all the Nanos with 5V via their 5V pins as already mentioned. If you supply 7V or 9V, the extra voltage is dropped by the Nano's regulator, resulting in wasted power. With 9V, nearly half the power will be wasted by the regulator and turned into heat.
How much power/current should your 5V PSU provide? How long is a piece of string? Each Nano will need 30~40mA for its own power. But microcontrollers are pointless if they don't control something. What will your Nanos be controlling? How much will those devices require?
Now the most important point/question. What are you using 32 Nanos for? I am immediately suspicious when a beginner says they need to use 2 Arduino in the same circuit. Almost always that is a mistake due to lack of experience or knowledge. 32 Nanos? Well, goodness me...
Theoretically it should only draw what it can draw, but I've read posts where folks say not to run high amp power supplies into Vcc. I just don't understand why not I guess.
Even practically it will only draw what it needs. But if you make a mistake a power supply that can deliver 11A is very unforgiving. Where as lesser supplies will shut down if you attempt to take more power, like when you have a short circuit, with a high current power supply you will melt tracks.
So do you feel lucky?
PaulRB:
It would be more efficient to supply all the Nanos with 5V via their 5V pins as already mentioned. If you supply 7V or 9V, the extra voltage is dropped by the Nano's regulator, resulting in wasted power. With 9V, nearly half the power will be wasted by the regulator and turned into heat.How much power/current should your 5V PSU provide? How long is a piece of string? Each Nano will need 30~40mA for its own power. But microcontrollers are pointless if they don't control something. What will your Nanos be controlling? How much will those devices require?
Now the most important point/question. What are you using 32 Nanos for? I am immediately suspicious when a beginner says they need to use 2 Arduino in the same circuit. Almost always that is a mistake due to lack of experience or knowledge. 32 Nanos? Well, goodness me...
So I have 32 individual objects mounted on a wall. each one with a capacitive touch module and a haptic feedback motor. I have very limited room for wiring. I'm attempting to have very minimal code, so each nano (or possibly mini) has a single button press from the capacitive sensor, and a short buzz from the haptic controller/motor. I'd be using one output from each nano to trigger an input on a centralized Mega which in turn sends a string via serial to a PC. That part is required by the software running on the PC. I've considered many other ways to do this, if you have another suggestion I'd love to hear it. note that the haptic controllers do not have configurable i2c addresses, and using an i2c multiplexer isn't really going to work because of physical wiring. This scenario allows me to treat each object as just that, a separate object. The code on the mega and the code on each nano would be very simple. Again, if you think there's a better way to do this I'd love to hear it.
Grumpy_Mike:
Even practically it will only draw what it needs. But if you make a mistake a power supply that can deliver 11A is very unforgiving. Where as lesser supplies will shut down if you attempt to take more power, like when you have a short circuit, with a high current power supply you will melt tracks.So do you feel lucky?
11A is just what I've been able to find off the shelf. Because of space and wiring limitations I'd like to keep this to a single power supply. Trying to determine the power draw has led to to an estimate of 5-8 amps for all the devices. 11 just happens to be the size I can find.
Ok, so what you want is, correct me if I'm wrong, 32 touch buttons, 32 haptic feedback actuators and the ability to send the code of a button pressed back to a PC?
PaulRB:
Ok, so what you want is, correct me if I'm wrong, 32 touch buttons, 32 haptic feedback actuators and the ability to send the code of a button pressed back to a PC?
Correct. Capacitive touch, to be specific. and I need to send a string over serial to PC.
I should add that i need a very minimal amount of wires to each device. In this solution I'm thinking I only need three. power, ground and signal.
Look at the DMX protocol, it is normally for lighting rigs but it will do what you want. You need a transmit line and a receive line to make the protocol work.
A long run of 5 V cabling might not be a good idea. It might be more appropriate to use a number of 5 V supplies - aka "phone chargers" to supply a few Nanos each. Do include the capacitors.
If you are not using the USB functionality of the Nanos, you could use Pro Minis. One programming adapter could set them all up.
Now you describe
aintmichael:
I'd be using one output from each Nano to trigger an input on a centralised Mega which in turn sends a string via serial to a PC.
I suppose you think you need a Mega to handle all the connections. No way!
If you use Pro Minis, you chain the serial output of one to the serial input of the next. Each has an "identity" (just a number will do, or in fact an ASCII character) and passes on messages received or adds its own message. The last Nano - which may or may not share the same function as all the others - generates the message to your PC.
aintmichael:
Correct. Capacitive touch, to be specific. and I need to send a string over serial to PC.I should add that i need a very minimal amount of wires to each device. In this solution I'm thinking I only need three. power, ground and signal.
What type of capacitive sensors and haptic motors do you intend to use? Please post links (read the forum guide in the sticky post to find out how to post links correctly).
I suspect this could all be done with 1 Nano and no Mega. How many sensors could be activated at once, in the normal situation? How many haptic motors would need to be on at once, and for how long?
Paul__B:
A long run of 5 V cabling might not be a good idea. It might be more appropriate to use a number of 5 V supplies - aka "phone chargers" to supply a few Nanos each. Do include the capacitors.If you are not using the USB functionality of the Nanos, you could use Pro Minis. One programming adapter could set them all up.
Now you describeI suppose you think you need a Mega to handle all the connections. No way!
If you use Pro Minis, you chain the serial output of one to the serial input of the next. Each has an "identity" (just a number will do, or in fact an ASCII character) and passes on messages received or adds its own message. The last Nano - which may or may not share the same function as all the others - generates the message to your PC.
What's the reasoning for not wanting to do long runs of 5v? That's sort of a big part of my question. I can't think of a reason not to do it.
One reason for the mega is so that each device is independant. For failure purposes I'd much rather have one object not function if something breaks rather than the whole thing failing.