Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

Motherboard Design Process 87

MBRFan writes "MBReview has posted the latest revision of their 'Motherboards - The Designing Process' article. This article covers the design process for modern motherboards, and also goes over some of the most common components that can be found. Definite read for information lovers, though beware, it'll take a while to read!"
This discussion has been archived. No new comments can be posted.

Motherboard Design Process

Comments Filter:
  • Before I threw out my P2B-DS motherboard when it stopped working. I might have been able to do something about it. :\
  • by antifoidulus ( 807088 ) on Thursday September 16, 2004 @08:23AM (#10264667) Homepage Journal
    if they also gave a little insight into what goes into the manufacturing process as well. After all, you can design the best board there is, but if you have trouble making it in large quantities, it's not going to be successful. Cost and ease of manufacture should be factors in any engineering decision.
    • by bsd4me ( 759597 ) on Thursday September 16, 2004 @08:41AM (#10264782)

      PC motheboards aren't terribly complicated as far as PWBs go. The don't really have a lot of components, and they don't have a lot of layers.

      What is interesting about them, but I don't think it really complicates the manufacturing proces, is the effort that goes into reducing the number of layers (number of layers is directly related to cost). This compilcates both routing and signal integrity.

      • The don't really have a lot of components, and they don't have a lot of layers.

        Compared to what? I understand some motherboards are pretty simple looking, but the ones that tend to be in my systems can get pretty complicated.
        • Take a look at some embedded electronics (such as cPCI boards) if you get a chance. A lot of these will literally use all available PWB real-estate, front and back) for components and also sometimes have to resort to daughterboards for more space (especially for memory).

      • Sorry to disagree (Score:2, Insightful)

        by EightBits ( 61345 )
        PC motherboards typically have 4 - 6 layers. That's pretty complicated. Think about routing all of the signal lines of the PCI (64-bit) bus and memory bus. If you have AGP nad PCI-X, add those in ther too. Think of all the signals going from your IDE ports and SCSI ports (if you have them) to your chipset. If you have an opteron, you have in the neighborhood of 940 pins. They all get connected. That's a lot of nets and a lot of routing!

        And, it's a good thing to reduce board layers to a certain point
        • PC motherboards typically have 4 - 6 layers. That's pretty complicated.

          I guess I should have given a point of reference. I have worked on designs requiring 12-16 layers, even after conceding to use blind and buried vias and running the router for very long time (ie, days). That is pretty complicated.

  • w00t! Once I get $8mil and armed with this knowledge I can finally form my very own itty-bitty wittle motherboard...

    ... maybe I'll just check CompUSA

  • Not bad, but (Score:5, Informative)

    by jimmyswimmy ( 749153 ) on Thursday September 16, 2004 @08:27AM (#10264694)
    I used to design motherboard power system components, and the author spends a good bit of time talking about that. That is actually the most complicated part of the board design, as it is not at all automated. Most component vendors try to sell a complete solution to the motherboard vendor, easing their job somewhat while helping the sales of the component vendor.

    One particularly interesting item of note: all those capacitors the author describes are absolutely crucial, and together form one of the largest cost items on the board. The system is tested using a processor-vendor-supplied "load tool" which simulates the worst case load transients one can ever expect to see. Most of that testing is done by the power system component vendor and then provided as a block to the motherboard vendor. Most motherboard vendors have no idea what they are doing there.

    In fact, a lot of the foreign manufacturers (no names) cost-reduce their designs by simply pulling out caps until the system blue screens. Then they put the last one back in and sell it. Intel is the only manufacturer I am aware of that actually sells the worst-case performing design.

    Note that I am only aware of products related to Intel-type motherboards. I never worked on the othe stuff.
    • Re:Not bad, but (Score:3, Interesting)

      by GigsVT ( 208848 )
      I guess the question is, will we ever be able to go back to the days when the actual computer power supply did the supplying of power, rather than having a second power supply on the motherboard?

      Or are the extremely low voltages and high currents involved with modern CPUs going to always tie us to the motherboard, due to the necessary wires gauges to carry that sort of power at low voltage?
      • Nah- That shouldn't be a problem as soon as the light-based processors start...be-ing. (Cause they don't exist yet!)

        I imagine there will be a power supply that provides juice to a single light source and whatever switches are needed. Fiber will carry the light pulse signals instead of traces on a motherboard. You probably wont need that big ol heatsink on the northbridge chipset... And you won't even need to add any of those freakish cathode lamps and other lighting gizmos to your windowed case... Anyon
      • The low voltages and high currents are exactly the reason for the on-board power supplies. Distribution and regulation are very difficult if not impossible at such low voltages where the needed current is high and resistance/impedance varies even slightly from trace to trace.

        This is also the reason for so called high-tension lines from power plants. High voltage at lower current is less sensitive to distribution path variations and has less loss over a given impedance/distance than does a lower voltage/hig
        • The reason for high-voltage power distribution isn't load regulation as much as power loss. Wires are resistive and have losses that proportional to the square of the current (Pdiss= I^2 * R).

          By distributing electric power at higher voltages, the same total power can be transmitted at a lower current (P= V*I), so that lower line-resistance losses are incurred, increasing efficiency.
      • Re:Not bad, but (Score:3, Insightful)

        by Billy69 ( 805214 )
        Erm, what happens on the mobo is really just a case of regulation, not a separate 'power supply' as such.

        And board-level power regulation goes back at least as far as the IBM XT (yup, I had one, they had 7805 5v regulators on them)

        But to answer the question I think you are asking, which is why doesn't the PSU supply the processor voltages, you have to think about the number of different processor and interface voltages there have been over the last few years. Processors have gone from using 5v logic suppl
        • No, the components on the board really do form part of the power supply.

          Because of the way processors draw current - bursts of current when switching - the rate of change of current can be enormous, in the range of giga-amps per second. If high-speed digital circuits like motherboards did not have a heirachical power supply the inductance in the power supply network would completely eliminate the ability to deliver the AC current needed.

          Motherboards will use a heirachy of power supply components startin

      • Re:Not bad, but (Score:2, Informative)

        by jimmyswimmy ( 749153 )
        As another poster pointed out, until a complete shift in technology occurs, this will not happen. Power loss in a wire follows the law P=VI. The lower the CPU supply voltage, the higher the input current will be.

        We continue to move to higher CPU switching frequencies and lower CPU voltages. Why? Well, higher frequency means higher clock speed, and greater processing power. Another way to increase processing power is to add complexity to the circuit, which means a larger chip. BUT, you can reduce the
    • We had a common issue at our company where the IBM motherboards would have their capacitors pop. i.e. the top plate would bulge, and the system would randomly freeze. IBM replaced something like 12 motherboards at our company, various models with the same issue. So I dont think the noname vendors are the only ones.

      The few nonames (Athlon/Duron) motherboards that I do have at home work relatively fine, with the problem of badly implemented BIOS. Funny how almost all Pentium 1 machines still work and are pre
      • When did this occur? A couple of years ago, there were some bad caps on the market after an employee stole *part* of the dielectric formula when he left one cap company for another. The new company used his process/formula without knowing it was incomplete, an dthe result was bad caps on boards from quite a few vendors.

        I don't recall whether the vendor was found culpable, or just the employee.

        I know it was covered on slashdot...
        • I don't recall whether the vendor was found culpable, or just the employee.

          I know it was covered on slashdot...

          Here [ieee.org] is an article about the problem on IEEE Spectrum. I have had to replace many motherboards because of this problem, all MSI or Epox.

      • Actually, the electrolytic capacitors used by most companies are definitely in cents, but they are in the thirty to fifty cent range typically. If you think about that, for the recent P4-type motherboards, you're looking at around 6-12 of these, if they do it right, and depending on a variety of factors. That's $18 minimum! And no vendor can advertise that their motherboard is "rock solid," because something is eventually gonna cause it to crash. So, looking at two boards on the shelf, would you really
      • you mean like the VRM's on the PPro & Xeon's?
    • Re:Not bad, but (Score:1, Interesting)

      by Anonymous Coward
      I always leave 3 or 4 no stuff capacitors (Bulk and HF)on the motherboards I work on (and I AM a power engineer). That way when I do buy one of these systems I can add the extra caps fairly easily and I'll get a system that I can actually trust.

      I find that the industry is only now starting to appreciate how difficutly power supply design is.

      -Coward cause I don't want my boss to hear (Empty Caps locations means harder buss routing because the vias go alllll the way through the MB)
      • What kind of caps do you typically use? In other words, what specs do you find are most critical? Is it ESR of some amount or do you just go for bulk capacitance? Also, what about the inductors and MOSFETs? I find that many motherboards have inductors that get extremely hot. Heat = wasted energy and I've always wondered how easy it would be to upgrade some of the components on-board. Plus, that extra heat does no good for the neighboring CPU. I've got my computer down to one fan, and I could use a sm
        • I could use a smaller fan if I could reduce the amount of heat being generated. Quiet, reliable, efficient and cool computers are good.

          Bigger fans are quieter, more reliable, more efficient, and cool better.
          • True, which is why I'm using a 192mm fan. However, I think that most 192mm fans are fairly noisy even when PWMed or downvolted. They aren't built to be quiet. However, there are a number of 120mm fans that are built to be quiet. If I need less CFM for cooling, I could use one or two 120mm fans that are meant for quieter (and thus less CFM) operation.
        • In the design of a CPU motherboard power supply, I suppose my two biggest concerns were heat and transient response. In dealing with the heat issue, you have to select the "best" components to make the system work. In other words, the cheapest set that will work. That's not trivial at all. First you decide how many phases to work with, considering that the norm is about 25-30 A per phase maximum. Then you pick a frequency. The operating frequency of the multiphase converter determines the inductors yo
  • I've heard stories of crappy motherboards blowing capacitors. Which is why its always better to get the 150 dollar one than the 30 dollar one. Because the 30 dollar one is effectively made out of scrap metal...
    • by Anonymous Coward
      Nonsense! The capacitor issue you are talking about was due to defective manufacture of the capacitors. I see the effects to this day.

      The thing is, capacitor failures were not limited to "cheap" motherborads! Although the more expensive boards may have had more effort in terms of worst-case design or more parts added to increase timing/noise margins, in the end they all send it overseas to be manufactured and those guys go out and buy the cheapest parts they can find. For a while, those were the badly manu
  • manufacturer (Score:2, Insightful)

    by GMail Troll ( 811342 )
    The design process for motherboards in my experience seems to depend on the manufacturer. Some take a more sloppy approach where the first version is riddled with bugs - take a look at Via's KT133 and subsequent KT133a for example (in some ways reminiscent to me of Redhat's notorious .0 releases). In general I'm more in favour of a rigourous approach when designing hardware - bugs in software are easy to fix (download a patch), but with hardware you expect what you're buying to work.

    All in all a mother

  • by Alwin Henseler ( 640539 ) on Thursday September 16, 2004 @08:37AM (#10264746)
    I have always considered motherboards THE most complex, interesting, and difficult part of a computer to design, and design well.

    Every single component on a motherboard can be a magnificient piece of work, but it deals only with a limited number of variables. A CPU or chipset component ultimately only deals with bus protocols and internal design. The same goes for other components like memory or harddrives.

    But everything comes together through the motherboard, and that (in my eyes) makes it the hardest part of a system to design well, considering the number of variables. A truly well done MB design is really a piece of art.

    • A truly well done MB design is really a piece of art.

      Yes, but while mobos may be art, a piece of art is no necesarily a good mobo- The throughput on a Van Gogh is pretty crappy. Although, I've hear that mobos based on Much's "The Scream" can really move sometimes.

    • I concure full heartedly. Back in the early days of my computer career, I came across a 486 motherboard from a small company, I think was called Chico or Chimano, but anyways that little board with a 66DX chip could just rock and outperform first generation pentiums. Even loading it down in 3.1, it was just a solid performer. It was just the right combination of motherboard and ATI Mach 32 Vesa graphics card that matched well.

      Similarly, I later came across a board that I could overclock my Cyrix (Yes, I

  • Yes, but.. (Score:4, Funny)

    by m0RpHeus ( 122706 ) on Thursday September 16, 2004 @08:37AM (#10264749)
    Yes... but does your motherboard manufacturer follow best practices like Feng Shui? [bbspot.com]
  • I wonder... (Score:3, Funny)

    by genixia ( 220387 ) on Thursday September 16, 2004 @08:42AM (#10264788)
    Why do motherboard manufacturers make a big deal out of multi-layer PCB's?

    LOL, I'd like to see them try and make a single-layer motherboard!
    • Nah, make a wire-wrap version of a modern day motherboard.
    • Why do motherboard manufacturers make a big deal out of multi-layer PCB's?

      Roughly speaking, more PCB layers makes it easier to do a complex design, but harder (read: more costly) to do the manufacturing.

      The mobo market is very competitive, so shaving a single $ of the price of manufacturing a board, is profit/market advantage for a mobo maker. So the mobo maker will try to keep the number of PCB layers to the absolute minimum needed for the design. AFAIK, many mobo's use 4 layers.

  • by Kozz ( 7764 ) on Thursday September 16, 2004 @08:50AM (#10264856)

    ...comes with the wisdom of the Far-East, yessir!

    What, you haven't heard of the Feng Shui Motherboard [bbspot.com]?

    "We didn't rely on reference boards or schematics from chip manufacturers. Those designs had too many straight lines and sharp corners, which are unnatural and direct poison arrows at our soul.
  • Graphic cards hitting capacitors.
    IDE and Floppy conectors on top of each other.
    Fan connectors way at the bottom and not enough of them.

    I have to be a Chinese contortionist to pulg things up once the motherboard is installed in a case.

  • Because stage one of the design process is always going to be "what form factor we gonna go for?"

    if it's ATX then 99% of the components can only go where they go, PCI slots have to go where they do, CPU has to go at the top cos pci slots extend to the bottom on cases, ide connects towards the "front" of the pc, psu cables near the top cos psu makers save money by shaving cable lengths, ram is an awkward shape so it can only go where it does, and all the i/o can also only go where it goes.

    ok that leaves "s
  • £12.95 will get you a (Pcchips KT266A SKT A DDR266 ATA133 AGP Sound LAN USB 2.0 ATX Retail Box) ebuyer [ebuyer.com]

    This is not a recomendation of this particular motherboard, I simply took the item that was the top of the list. When you see something like this for this price it makes you wonder how they make their money. I know the quality is not great, it will probably die before too long, but thats not the point.

    For that price how can someone make a motherboard that works even once?

  • To save the other metric readers the trouble googling, 1 mils = 1/1000 inch = 2.54e-5 metre. conversion [simetric.co.uk]
  • I got as far as page 2 before leaving in disgust. I could follow it all easily enough (B.S. in ECE) and hit all the right technical points, but it was just so poorly written! Terribly sentence structure, terrible paragraph structure, several glaring spelling mistakes... I give it an A- for science&technology, but an F in English.

    And if the author uses the word "basically" one more time, I'll throttle him.

  • one of my friends graduated MIT when i was still a sophmore in HS twiddling with microcontrollers. i asked him what he could do with his CompE degree; he told me could build a motherboard. i was amazed, but then again i already knew i would be a CompE anyways. anecdote over.

    with more and more cores becoming more and more embedded, sometimes i think there's a small possibility users might be able to again develop their own systems. alright, so i'll probably never build a motherboard for an intel deskto
    • i'm just a computer engineer (may '05 baby), they dont teach us much about the black art of high speed signalling, but somehow BGA gives me faith. it may mean you have to have your 6 layer board shipped to you, but that seems like an OK tradeoff for being able to design some amazingly high speed hardware without being a signal-foo master.

      The high speed busses atually mean you've got to be more of a high speed design master- True, there aren't as many traces, but you have to be much more careful about the
      • to be sure, matching your line becomes very exacting. but manufacturers usually have plenty of reference materials on how to lay good traces, what specs your shooting for.

        getting paths down doesnt seem THAT hard. i want to work out each individual piece, lay my traces between two points and be done. everything short of 'my software will lay my traces for me'. thats not THAT sinfully bad. getting to spec doesnt look oh so difficult.

        crosstalk, on the other hand, scares the crap out of me. you've got a
        • with good solid connectors, good helpful specs, doing this stuff shouldnt be that hard. yes, its higher speed, but an idiot like me migh have a chance of making an uber-high-speed board if i use some liberal spacing.

          i could be dead wrong.

          Ahh... The innocence of youth... Good to set your sights high, but, really, this stuff is hard. All of it. Spacing is only part of it- usually you can get enough spacing if it is 3x to 5x the height of the trace above the ground plane. But if you're talking about di

Variables don't; constants aren't.