Final GIGABYTE X79 Boards with Digital PWM's Pictured!

[WeaponX]

Why does their low end board have more Ram slots then their High End board, still trying to figure that out.

[pazza316]

I would say both UD5 and UD7 are high end boards really. But I know what you mean you woudl have thought both would have 8 RAM slots.

[M.Beier]

Why does their low end board have more Ram slots then their High End board, still trying to figure that out.

Because the "OC" model was replaced with "UD7" name.
My guess is that this is in order to meet expectation of UD7 concept, the flagship, and therefore price it as UD7 traditionally has been priced, above UD5
TBH I dont mind one bit, and I am shooting in blind here, might be completely incorrect, I do not have inside information...

[antiacid]

I'm pretty sure I'm going with the UD7 again for s2011 but I was wondering, why do we care/want digital PWM's?

[zalbard]

I can't see any digital PWM on these mobos. Nice marketing again.

Indeed. Wtf?

[sin0822]

Um you guys know what a Digital PWM is? Right?

B/c If you see you don't see a Digital PWM, maybe you haven't seen the right pics. here let me clarify:

PWM is a pulse width modulator, a chip, a single chip, or more of there are more than 1 power supplies on the board, which there always are.

VRM: Voltage regulator module, in this case aa DC/DC SMPS

Good article on Digital PWM: http://www.petrovgroup.com/pdfs/Oct%...%20article.pdf


Now we have components:
PWM chip (maybe some doublers like on the Maximus 4 Genez, and the G1 Sniper 2)
Drivers
MOSFETs
Inductors/Chokes
Capacitors

PWM Chip- this is the only part that is analog or digital. The other components deal with voltage (analog), not 1s and 0s, which is what digital is. Digital PWM implies that the error voltage is convereted to 1's and 0's through an ADC, and then processed by a PID algorithm, that controls and changes the cuty cycle accordingly and other things to regulate the output.

Power Stage consists of the MOSFETs and the Driver, this part is nothing to do with 1's and 0's, especially not a 3 part system (Driver, H-Side FET, and L-Side FET), DrMOS are all 3 parts in one, ASUS doesn't use DrMOS, they use seperate analog components.

L/C Circuit- the output filter, made of inductors and capacitors. Proadlizer capacitors doesn't imply digital by any means.

Digital and Analog PWMs only apply to the PWM Chip, and the PWM is not the whole VRM, its just the PWM Chip.

These boards are as digital as they come, those fancy CopperMOS just look good, they don't imply digital by anymeans, which is what most people think. Most think that the CopperMOS and the Proadlizer capacitors on ROG means digital, but that isn't true those can be on analog PWM too.

BTW This PWM technology used is the same as ASUS's Digi+ II tech, its updated to VRD12.5 spec.

[Mechanical Man]

those look exactly like SAS connectors. I see that EVGA is doing sas on the sr3, maybe these are sas as well? Patsburg supposed to be UBER with sasness.

SAS would explain removal of those SATA ports..

[M.Beier]

Snip

Give up... The one you try to convince, also believes that AMD Bulldozer beats Sandy Bridge by miles.... I think you are reasoning for deaf ears.

[sin0822]

Here if they don't believe me they can just read this: http://www.hardwaresecrets.com/article/616

its not updated, but it explains the parts well enough

[abeeftec]

Not SATA power connectors for PCI-E?

Also, have they ever explained why the UD5 has more DIMMs then the UD7?

-PB

Probably because the original OC board for X58 was for overclockers who typically used less ram then a gamer or graphic artist.
I think its stupid too because all Asus boards have 8 dimm slots so far.
But I will never purchase an Asus board again after they sent me a bad board years ago on an RMA. I sent them a board with a Bad Ethernet port and they sent me back a board
that didnt work at all and then would not RMA it again when I complained that it didnt work 1 day after I got the board back!

Edit: Haha, I noticed someone answered this a few posts above!

HEY SIN0822! How you been?!

[K404]

Digital PWM implies that the error voltage is converted to 1's and 0's through an ADC, and then processed by a PID algorithm, that controls and changes the cuty cycle accordingly and other things to regulate the output.

Does it? Can you link us to an industry-standard definition of Digital-PWM? Maybe Gigabyte could also supply the community with a definition of a power phase?

I find it hard to believe that all the.... "D-PWM pioneers"....

examples:

http://www.legitreviews.com/images/r...950pro_pwm.jpg

http://i4memory.com/reviewimages/mot...-G_1000_tn.jpg

http://images.bit-tech.net/news_imag...xconn-pwm1.jpg

Did all this implementation work when the point was to get a Digital error voltage.

[sholvaco]

After seeing all the major players' offerings I have to say that overall I like Gigabytes the best... especially the assassin board.

Hope the BIOS/support is a worthy complement to the features/aesthetics.

Two platform launches in a row now.

[sin0822]

Does it? Can you link us to an industry-standard definition of Digital-PWM? Maybe Gigabyte could also supply the community with a definition of a power phase?

I find it hard to believe that all the.... "D-PWM pioneers"....


Did all this implementation work when the point was to get a Digital error voltage.

Digital and Analog are really almost marketing hypes, but there is some very big difference, and that in lies with the error feedback and control mechanism. A PWm compares ref to error and takes the difference and raises or decrease duty cycle to change the output to reflect whatever the PWM is instructed to produce as output. its the basis for what a PWm does.

Yes, that error feedback is how voltage is regulated. That error feedback has everything to do with voltage stability, and how precise a PWM is. Now analog PWMs use an error amplifier, and then a comparator to compare a reference to a error. its an "all hardware" method. It is fast.

Digital uses an ADC to convert the error voltage to a digital signal, which can then be processed by an algorithm (PID) which is more precise. Now in the past the ADC caused delays in the feedback loops, but that was years ago, and now Digital PWMs are very fast, and if they weren't they wouldn't pass through intel transient tests and get VRD 12 and VRD12.5 cert.

That digital feedback PWM is what causes the maximus 4 extreme to require less voltage than the P67A-UD7, as voltage is never constant, its always changing, the digital PWM allowed for better VReg, ill admit it.

now digital also has a few more benefits, total user controller, from programmer of the system to the end users. there is onboard NVM which can help a lot with putting in different settings for different situations. LLC and everything is much better implemented, there is no need for a GPIO.

How about some real articles: http://electronicdesign.com/article/...ht-to-the.aspx

here is a picture for you:



and here is the source, its a very good source:
http://focus.ti.com/download/trng/do...ic_7_Hagen.pdf

Now what I want you to do is google synchronous power stage, and you will see what a power stage is. you need to look for ones using buck topology. Anyways there are two kinds, either DrMOS, which are power stages all in one IC, Volterra uses the same DrMOS style, and IR even has its own, and it looks like the memory are using some type of integrated power stage, with the high-side MOSFET (connected to source), low-side MOSFET (connected to ground), and Driver. The high-side and low-side FETs are connected to each other, and at that point where they connect, the end of the inductor connects. The driver connects to the MOSFEts as well. From the driver to where the MOSFETs connect is considered a power stage.

With the Copper MOS, which ASUS used on their ROG boards, the driver and FETs are exposed, not in a single IC. Volterra uses single IC i believe, GB uses DRMOS which is drive, and FETs together like volterra, and IR has its own powerstages (all integrated like volterra).



To top it off look at the VRM for the rampage 4 extreme: http://cdn5.tweaktown.com/content/4/...eview_full.jpg


Now we just need to see how GIGABYTE's X79 UEFI is, I think GB is really starting to listen.

[Tor_2000]

Wow Sin, you really know your electronics Really impressed. Well, I do belive the DrMos and Digital PWM's are a great combo, and I don't think it has been done before. I'm looking forward to seeing the results.

[M.Beier]

Wow Sin, you really know your electronics Really impressed. Well, I do belive the DrMos and Digital PWM's are a great combo, and I don't think it has been done before. I'm looking forward to seeing the results.

Actually those two diagrams are very basic stuff.... You get that within 2 first semesters of any eletronics education.

Anyhow, best get back to that stupid RIAA Have to finish it soon

[El Mano]

Thank you very much Sin0822!

[jogshy]

Good MBs, however:

1. No thunderbolt ports.
2. No power / reset switch on the UD3. The UD5 have the reset and power switch too far one from each other.
3. Too few DIMMs, except the UD5.
4. Numeric LCD only present in the UD7
5. If USB 3.0 is retro-compatible... then why not put only USB3 connectors on the board?

[sin0822]

Actually those two diagrams are very basic stuff.... You get that within 2 first semesters of any eletronics education.

True its just schematic drawings. but finding the difference now a days is hard digging through all the crap, if i find more links or articles that i think you guys would like ill share them.

I am just an amateur in a sense, but I have picked up a few things, i have great deal of interest in DC/DC SMPS used on motherboards, and I have done a ton of research and have being doing some tests with deal boards. interesting stuff.

but the difference between digital and analog i didn't know until about 6 months ago, after i wrote my UD7 piece, which was not as good as it could have been.

[antiacid]

now that's what I call a thorough answer to my question (and the questions I didn't yet know I would have)!

[badbaker`]

10 sata on the UD-3? 2 down on the bottom corner.

[El Mano]

5. If USB 3.0 is retro-compatible... then why not put only USB3 connectors on the board?

THIS. This is a mistery for me too.
One could argue that it's the price what limits the number of USB3 ports we see on boards but. on high end boards? They are already quite expensive, I'm sure they can use USB3 only.

[paulbagz]

THIS. This is a mistery for me too.
One could argue that it's the price what limits the number of USB3 ports we see on boards but. on high end boards? They are already quite expensive, I'm sure they can use USB3 only.

Limitation of the chipset?

Doesn't USB3.0 use PCI-E lanes or is that only third party chips?

-PB

[M.Beier]

True its just schematic drawings. but finding the difference now a days is hard digging through all the crap, if i find more links or articles that i think you guys would like ill share them.

I am just an amateur in a sense, but I have picked up a few things, i have great deal of interest in DC/DC SMPS used on motherboards, and I have done a ton of research and have being doing some tests with deal boards. interesting stuff.

but the difference between digital and analog i didn't know until about 6 months ago, after i wrote my UD7 piece, which was not as good as it could have been.

Well to be fairly honest, it is not really the circles you have marked, but the..... 8? MCU' that makes the main difference.

[bamtan2]

THIS. This is a mistery for me too.
One could argue that it's the price what limits the number of USB3 ports we see on boards but. on high end boards? They are already quite expensive, I'm sure they can use USB3 only.

you guys are just inexperienced. so let me explain to you why the back panel uses some intel usb ports rather than all third party usb ports.

third party usb ports suck. savvy?

[cx-ray]

I find that the NEC/Renesas USB3.0 work well. The only problem is that I haven't seen a mobo yet that allows those ports to be bootable. Perhaps, that's among the reasons for also including USB2.0.