Chokes/Inductors and their Purpose?

This is a pretty general question, I suppose.

Something that always struck me when looking at computer motherboards is the amount of chokes on board. The computer I built last year must have 50+ chokes/inductors on the board.

As far as I know, their function is to reduce interference of a sort.

When you look at an Arduino board, there are no inductors in the schematic, as far as I am concerned.

Where do you put inductors in a circuit? What do they do? Why doesn't Arduino use them?

Maybe you know what I am trying to get at.

Thanks.

The computer I built last year must have 50+ chokes/inductors on the board.

That seems unlikely. Perhaps you mistook electrolytic caps for inductors?

Chokes resist changes in current similar to the way that capacitors resist a change in voltage.

So one uses is to prevent current spikes from one part of a circuit from getting into other parts.

Another consequence is that inductors can change voltages (similar to the way you can charge a capacitor at low current and discharge it at high current.) So inductors are used in high efficiency voltage conversion circuits (switch mode power supplies); a modern computer typically requires several different voltages, and it's pretty common to have converters on-board to convert a supply voltage to appropriate values, rather than requiring a power supply with lots of different voltages...

Okay you are right. It wasn't 50. Apparently I have no sense of numbers. I actually counted them and there were 22 ferrite chokes on the board.

Well I really wish I knew how to implement them into a circuit. Before becoming interested in the electronics field, I was pretty much a computer power supply fanatic. This like http://blog.antec.com/power-supplies/technical-finesse-the-high-current-pro-hcp-1200/ make me genuinely happy.

Well, the point is, I guess I have more learning to do. Building power supplies is probably what I've done most of so far in my electronics career. Little breadboard mountable power supplies. I've built lots of them...but now, I am curious to see how to add a choke in to the situation.

Thanks for the help.

The use of inductors in power lines is part of supply decoupling:- http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html

The use on signal lines is to reduce electrical emissions to meet emissions standards. There is so little going on electrically on an arduino that it doesn't need inductors to pass emissions tests.

iirc, the ATMega328p datasheet specifies an inductor between the Vcc and AVcc pins, yet only a very small number (at best) of Arduinos or clones implement this. Is there ever a “good reason” to not follow that spec? I’m wondering if it might be a simple matter of “we probably don’t need it, and including it increases cost, so let’s not use it.” If that is the case, and I’m building my own clone, how would I decide if I needed it or not?

In fact the chokes/inductors that are in the motherboards are used in the dc/dc regulators that can pump the 100+ Amps that one CPU use.

iirc, the ATMega328p datasheet specifies an inductor between the Vcc and AVcc pins, yet only a very small number (at best) of Arduinos or clones implement this. Is there ever a "good reason" to not follow that spec? I'm wondering if it might be a simple matter of "we probably don't need it, and including it increases cost, so let's not use it." If that is the case, and I'm building my own clone, how would I decide if I needed it or not?

I believe it's more of a Atmel recommendation rather then a requirement to have an inductor feed the AVcc pin. Anything that reduces power supply noise will have a positive effect on A/D performance, but if that is important to your application only you can determine. The A/D performance of a Atmel chip is not really 'instrumentation' quality to begin with, the specs a useful A/D but not high performance by any means. External A/D chips can be had that offer much better resolution and accuracy if one really requires that. The ± 2 LSB Absolute Accuracy specification stated in the 328p datasheet kind of defines the limits of the built in A/D feature.

Lefty

Lefty,

Thanks for the info. That's about what I had expected but I don't QUITE have a grasp on what "high quality" vs "low quality" requirements/results are for A to D. Hence it's not clear to me how you'd even make that sort of decision.

That's about what I had expected but I don't QUITE have a grasp on what "high quality" vs "low quality" requirements/results are for A to D. Hence it's not clear to me how you'd even make that sort of decision.

It's all about what your application requires. The built in A/D is a 10 bit conversion so it can only resolve a measurement between 0-1023 counts. If the application requires a larger measurement dynamic range then a A/D converter with 12,14,16 etc bits of conversion is required. 24 bit A/D is the largest I have seen and requires a lot of attention to board layout, power filtering etc to take advantage of so much measurement range. Another key specification for A/D converters is how fast they can do a conversion, as it sets a limit to have fast one can sample a analog signal without losing information about that signal.

Lefty

The ATMega328 datasheet does not mention inductor anywhere. Perhaps there is an application note somewhere concerning that. It does recommend a low-pass filter on the AVcc pin when it is connected to Vcc for ADC use. It also has a caution regarding EEPROM programming with heavily filtered power supplies, with Vcc changing slowly on powerup/powerdown & possible EEPROM corruption.

The ATMega328 datasheet does not mention inductor anywhere.

Sure it does:

21.6.2 Analog Noise Canceling Techniques : b. The AVCC pin on the device should be connected to the digital VCC supply voltage via an LC network as shown in Figure 21-9.

Figure 21-9 shows a 10uH inductor before the usual 100nF bypass cap.

There seem to be different revisions available then.
I was looking at Rev. 8161D-AVR-10/09, which is a searchable .pdf format document, which I downloaded in August this year when I started Arduino-ing. The word inductor does not appear. AVcc filtering only says low-pass filter. Fig 21-9 is an overview of the TWI module.

Huh. I'm looking at Rev. 8025I–AVR–02/09. Note that this never mentions "inductor" either, but 21-9 ("ADC Power Connections" (section 21 is on the A-D converter)) has a schematic with an inductor symbol labeled "10µH"

I guess I should update my docs ...

(Rev. 8271C–AVR–08/10 has this in figure 23-9.) (wow. They keep getting bigger. My older rev was 444 pages and 8.1MB. The newer one is 556 pages and 21MB)

Hmmm...maybe I should try putting in a choke then.

What kind of inductor is to be used. RF or a regular power inductor?

10uH, low current. not much needed.

RF? And in theory, if you put a 100uH choke in there...would it make a difference?

I think anything with a max current of more than 50mA will be fine; that probably means an RF choke would work (this IS, after all, an RF-like application.) I couldn't find a TH 10uH choke at digikey with that low a current rating...

The danger going to higher values is that the DC resistance and other current-related parameters would get "worse." Still not likely to be an issue.

Yes I was originally thinking RF and then I saw Mouser had other types of chokes…

I’ll just keep the value in spec. I was just curious as to the effect a different value would have.

Thanks.