PWM Issues on Gigabyte GA-Z87X-UD4H Motherboards

Ah yes, PWM fans can which is why I said..." Most anyone using PWM fans would not notice the difference between the two methods."



Correct you are, which is why I noted..."The fourth pin, which carries the PWM signal, does receive approximately 5V at any given time during my testing. But I do not have access to an oscilloscope to test signal pulses."

I don't think you're reading my posts all the way through. Just because the header has 4-pins and is stated to offer PWM capabilities, doesn't mean that fact true. It's only SUPPOSED to function as published. And since this is assumed, no one would notice if was operating either way unless measured.

Furthermore, as mentioned that while 5 volts are present on the #4 pin, I cannot verify if the PWM is modulating. The voltage present was verified using my meter, not the pulse widths. If the duty cycle is adjusted, the voltage on the 12V header pin remains constant.

In PWM mode, 12 volts are applied constantly to the header-then to the fan. Regardless of the modulated signal or even the presence of a load on the header. The components inside the fan motor housing are what processes the PWM signal, NOT the motherboard. Which only generates the signal and subsequent differences in modulation as a matter of a requested percentage value or to achieve requested a RPM. Basically acting like a rapid fire on-off switch for current to be allowed to flow across the fan motor coil. You've essentially just said the same thing in my previous post, but I needed to repeat myself so that we know we understand each other. A point was also addressed in a paragraph from my earlier post giving instructions for how to go about measuring the voltage on the #2, 12V pin. You did include something I didn't and that was to list the pinout. Unfortunately I assumed anyone familiar with board level ICs and other SMD components would know the pinout, or at least have the competency to learn or confirm.

As I stated previously, you'd need a scope to see the PWM signal modulating, which I do not own. Modulation, not amplitude, cannot be measured using normal multimeters. This is because the modulation frequency, and variants thereof, is to high to measure for most portable and bench multimeters. And yes, there are some that are a combination of scope and meter. Which is why I inquired if someone has a scope sitting around and the same brand and series of motherboard.

The crux is the issue is that while, as a matter of fact, the CPU_FAN header operates as normal by consistently applying 12 volts and simultaneously generating and manipulating the PWM signal, CPU_OPT & SYS_FANs 1-3 do not operate in this fashion. Ergo, the fault lies in the code or engineering. This is made evident by other users in other forums, and even here, complaining of the exact same issue.

So if only one header has been properly assigned and attributed to the IT8728F visa vie PCB trace manufacturing, BIOS code and chip programming; how would Gigabyte issuing an RMA for in warranty repair accomplish anything?


It is not my intention to be condescending but this issue has left me with no resolve for quite some time. Gigabyte eSupport hasn't escalated my support tickets to someone with the knowledge of this area. They have said yes, then no, then yes again.

What's even more disheartening is that if A) it can be resolved in a retooling and release of a BIOS update, there is not guarantee they will address this issue, it could be quite a while before they do or can; or worse even B) it's a matter of faulty engineering, for which there will likely be no solution. A fault in the hardware engineering would and could not be resolved until the release the next revision of this model/series. But there is a chance it could possibly be rectified by replacing the IT8728F control IC.

I don't know how many times I can say the same thing. Once again if you read the post, you'd know that I have verified the voltage issue using my multimeter, a Fluke 29-Series II . But I need confirmation from another owner-one with an oscilloscope.





An RMA is not a procedure. it's a control number issued for a piece of merchandise being sent to the manufacturer for a diagnosis to either be repaired or replaced. RMA=Return Merchandise Authorization

Have you used a scope to see if it's simultaneously varying the voltage while simultaneously modulating? If someone has a scope with SMD clips, they could even test this issue at the IC pins.

This is why I asserted that it's a BIOS code error or IC programming/tracing fault. Because if the CPU_FAN is OK, then what's stopping the rest?!?!?!? Fortunately for users, the PWM fans will work in either mode.

The CPU_OPT on my GA-Z87X-UD4H operates just as yours, as a variable voltage fan header. It and the rest may as well be 3-pin headers.

Yes, the people answering their eSupport tickets are not as technically inclined as others within the company. Here's a link to my ticket responses: http://www.tomshardware.com/forum/id-1844569/experiences-pwm-gigabyte-z87-ud4h.html#12986688

At any rate, as you and I have both mentioned that this will need to be solved within a BIOS update or worse, a hardware revision. But I guess since you've found this to be true of the course of time while testing the gamut, it will never be solved. Regardless, I'd like to see someone's findings published herein.

Would you happen to know what IC is used on a board within your reach that also has this issue?

Spoke with a senor technician at Gigabyte US in SoCal that had a conversation with the engineers in Taiwan.

THE CPU_FAN IS THE ONLY FAN HEADER WITH PWM CAPABILITIES ON ALL GIGABYTE Z87 MOTHERBOARDS. I then asked him why the hell are there 4-pin headers on the rest, he had no idea and the engineer also had no idea.

So if you're purchasing a Gigabyte GA-Z87-UD4H or any other Gigabyte motherboards on the Intel 8-series Northbridge architecture, the only PWM fan you need to purchase is the fan attached to the CPU_FAN header. Every other fan header only operates by varying the voltage rendering the #4 pin as a constant source of 5V, not a PWM signal.