Digital vs. Analog PWM

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
For those of you wondering exactly what the difference is between the older analog PWM controllers and newer digital PWM controllers we are starting to see on enthusiast motherboards, we have this handy illustration to show you exactly what has changed. Image courtesy of abit.
 
so what's the difference between those and the PWM in fans? Man all this new tech makes my head hurt lol.
 
Abit IN9 32X-MAX ftw!! lolz

This is just one reason why I chose this motherboard for my latest build.
 
To be truthful, it's not really "analog" vs. "digital" PWM. All the components are the same. You've still got the chip generating the PWM signal, the transistors switching the current, inductors and capacitors smoothing the power. The difference is in the construction of those components. The MOSFETs are smaller and more efficient, the inductor is solid-state instead of wirewound, and the capacitors are in a ceramic surface-mount package instead of electrolytic radial package.
 
Boards for testers have been using this tech for years. I don't understand why it took this long for a motherboard maker to get with the program.

With the parts available today the companies could make an outstanding low profile board. None of us would have to worry about cooler clearance again. I just wish we could get back to real solder instead of this rohs crap.
 
I agree with Mohonri 100%.

The circuitry is the same, the only difference is in the type of components that are used. The Digital vs. Analog thing is just marketing BS from Abit.

However, that's not to say that the new components aren't more reliable, efficient, and/or accurate... it's just not Digital vs. Analog.

I found this article on the net (probably authored by Power Electronics, the manufacturer of the Power Bead Inductors) that highlights the advantages of using THT Power Bead inductors over the traditional wound toroid inductors for CPU voltage regulators.

http://www.tmcnet.com/usubmit/2007/04/15/2509129.htm
 
"electro coils and solid capacitors"
"mini sized ceramic capacitors reducing a capacitors own high frequency"

Wow. Its as if they had a 5 year old help with the marketing on this. The translation on that was either poor, or whomever wrote it in the first place had no idea what they were talking about.

Though, wasnt it also abit that once had "High quality 100% japanese capacitors" advertised on their motherboard boxes?
 
Biodune said:
Boards for testers have been using this tech for years. I don't understand why it took this long for a motherboard maker to get with the program.

With the parts available today the companies could make an outstanding low profile board. None of us would have to worry about cooler clearance again. I just wish we could get back to real solder instead of this rohs crap.
Click to expand...

They didn't do it before because it takes more time to design a digital PWM vs analog and it used to be a bit more expensive. Its more common nowadays as board makers need the space more than ever, and the board designs are so complex its not a big deal to do a digital PWM at the same time.
 
this has nothing to do with fans i dont think...

in their pictures of "analog PWM" they are showing voltage regulators...

im guessing they are going from LINEAR regulators to SWITCHING regulators...

its a good thing if you can reduce switching noise... its significantly more efficient... which is also why you can get 1000w out of a thing the size of a computer psu, meanwhile a similar linear regulated psu would probably be bigger than most computers

linear reguators dont use PWM, they turn whatever they are dropping into heat usually... so inside a computer, not so great... switching regulators do generall use PWM that gets smoothed out using inductors and filtering capacitors... the only thing is that the switching causes high frequency noise, or switching noise, which can be problamatic in some uses... they are also significantly more complicated to implement than a similar linear regulator... minus the heat issue...

but ya, this is by far the largest bastardisation of electronics and electronic terms i have seen in a while... and i am disapointed at abit for making that graphic... but it was probably the marketing dept, and the engineers are like "what the hell are these marketing guys on?!"
 
Motherboards aren't really that complex. I don't think they even go past twelve layers. On top of that the components are not that bad of a price. Components were actually cheaper before everyone had to retool for rohs. Consumer boards also have massively loose tolerances compared to tester boards. You could replace most parts of a motherboard with smaller better parts without even having to redesign the board.
 
Biodune said:
Motherboards aren't really that complex. I don't think they even go past twelve layers. On top of that the components are not that bad of a price. Components were actually cheaper before everyone had to retool for rohs. Consumer boards also have massively loose tolerances compared to tester boards. You could replace most parts of a motherboard with smaller better parts without even having to redesign the board.
Click to expand...
Actually, no motherboard I've heard of has more than 6 layers. Most have either four or six. A 12-layer board would be super expensive. But yes, RoHS has made things a little more expensive, although it's safer now, supposedly.
 
Mohonri said:
Actually, no motherboard I've heard of has more than 6 layers. Most have either four or six. A 12-layer board would be super expensive. But yes, RoHS has made things a little more expensive, although it's safer now, supposedly.
Click to expand...

RoHS is a huge pain in the ass thanks to those goddamned tree huggers that bitch about *everything*. I can't stand them/environmentalists. Lead-free solder is bitchier to work with...melts at a higher temp, you might need a new iron, etc. Not to mention that the joints looks uncomfortably close to cold joints as well, making visual inspection a PITA as well. So far only one of the lab station here (RKE/automotive systems lab) has been set lead-free, and we don't plan on going all lead-free until someone comes in and forces it. It sucks :<

Back OT though, seems like this move could have been made awhile ago; the rows of electrolytics & inductors near the CPU socket always angered me during cooler and waterblock installation. :(
 
I love the caps burst under high temp, true if you use crappy electrolytics. So don't use crappy caps...problem there is solved.
 
You must log in or register to reply here.
Back
Top