Hello,
I am doing a project for automotive use and need to use a microcontroller of some kind to perform different functions. I am using a buck converter at the moment to take the 12v car down to 5v.
My issue is I am getting a load of radio interference with it! So does anyone know of a microcontroller or something I can program that will run off 12V? Failing that, does anyone know a circuit that will create zero radio interference and minimal heat?
Appreciate it!
Define "radio interference" as it relates to an arduino. It's not an sudio amplifier so what's the issue ? (your desription of the symptoms is vague)
Is the processor reseting ? Are you getting corrupted data ?
What's the problem ?
Atmel does not recommed the ATmega328 for automotive applications. PERIOD.
Sorry folks!
I am using an attiny85 with a buck converter setup, which I know isnt super low noise.
What I am getting is really bad radio reception when the mcu is on and it immediately improves when it is switched off.
Any ideas?
Anything with fast switching speeds (fast rise/fall times) tends to generate some interference - the faster the rise/fall times, the more RF issues are possible. Both the Arduino and the buck converter have the potential for generating RF noise. A linear regulator such as a LM7805 is clean as far as switching noise goes, however, being a linear regulator, you now have more heat to deal with if there is any significant current drain. You need to be more specific as to exactly what the RF "noise" issue is and then see if you can determine where it is coming from before you can fix it. (if you look at how a square wave is constructed, basically, it is the sum of the odd harmonics of the fundamental frequency and the faster the rise/fall times are, the higher the harmonics go and can extend well up in the mhz range or higher).
Assuming the noise is coming from the buck converter, what else are you running off the 5V?
I've got an Arduino in a car running the Arduino plus some LEDs and solid state relays. I'm just plugging 12V into the barrel connector and using the on-board linear regulator.
If that won't work for you, maybe a better switching regulator or try adding some "big" capacitors to the regulator inputs & outputs (maybe 1000uF so).
I am powering some LED strips and an optocouple.
Pretty sure it is the converter creating the noise as there isn't anything else that could apart from possibly the leds but they are at the back of the car so shouldn't affect the radio.
I will give the capacitors a go.
What do people think of figure 5 in this?
That power supply is enclosed in a metal box. Is your circuit and arduino in a metal box ?
It isn't but I need a decent power supply before I think about boxing it.
Fine.
Then don't listen to your radio.
Anyone got any useful suggestions regarding the actual circuit?
Is there anyway of combining switching and linear regulators to get a really low noise output?
blues1143:
as there isn't anything else that could apart from possibly the leds but they are at the back of the car so shouldn't affect the radio.
Try disconnecting everything including the LEDs. Get to a state where you can determine what causes the interference and what does not contribute to it.
Those LEDs radiate like f***k if you control them with PWM. It is the rate of change of current with respect to time that generates the RF and putting them at the back of the car is not going to do much to reduce this. Only when you has isolated the cause can you begin to address the problem.
Is there anyway of combining switching and linear regulators to get a really low noise output?
If the switching is coming from switching regulators , then how would combining switching regulators with linear regulators reduce the switching noise ? You can't take the switching out of switching regulators. You either don't use them or you shield it. Why don't you wait until you have properly shielded your circuit to ask for suggestions ? If it is properly shielded you shouldn't have to make any changes. It just seems like you have the cart before the horse. You are asking how to eliminate something that probably won't be a problem when it is in a metal box. Is that really necessary ?
I'm not seeing it. The normal design procedure is create a design, build a prototype and do your testing.
Revisions are only done if the prototype fails the testing. You haven't even completed your fabrication and your making revisions.
Only when you has isolated the cause can you begin to address the problem.
It's too late now , but FYI, next time you design something , build and test it in stages. Don't just throw the whole enchilada in their and test it. Build one stage, test it, and then build the next. Otherwise you wind up with fifteen unknown variables.
If the LED's are at the back of the car as you indicate and you are driving them with a square wave (PWM) and unshielded wiring, that is a BIG antenna that is going to radiate all sorts of garbage. Start by disconnecting the wires to the LED's - disconnect it at your board so you are disconnecting the "antenna". Does that solve the problem? There are a number of things you can also do to suppress noise such as ferrite beads in the power lines etc, but first, you have to identify the noise source.
If the LED's are at the back of the car as you indicate and you are driving them with a square wave (PWM) and unshielded wiring, that is a BIG antenna that is going to radiate all sorts of garbage. Start by disconnecting the wires to the LED's - disconnect it at your board so you are disconnecting the "antenna". Does that solve the problem? There are a number of things you can also do to suppress noise such as ferrite beads in the power lines etc, but first, you have to identify the noise source.
Am I to understand the cabling is not shielded cable ?
raschemmel:
Am I to understand the cabling is not shielded cable ?
Well a LED strip is not shielded so the cabling is some what irrelevant.
There are thing to try to fix it once you have identified that it is the LEDs. I think it is more chance than the switching regulator.
I once had a few engineeres who would not countenance using a switching regulator in an RFID reader because of the interference it was bound to generate. I made them test it and they were wrong, it did not cause interference at all. It is all those capacitor's on the output that reduce the noise.
Anyone got any useful suggestions regarding the actual circuit?
It is all those capacitor's on the output that reduce the noise.
Awesome! Some really great suggestions here guys! Will give them all a test this weekend! No doubt I will be back with more questions!