Gigabyte GA H87-HD3 System Fan Header Issue?

[Mantrid-Drone]

Gigabyte GA H87-HD3 System Fan Issue?
Just powered up my first completed build with the GA H87-HD3 MB.

Everything working, HDD idling, all fans going, GPU fine etc, pretty pleased. Went through the BIOS menus seeing what's what and all looked good except I noted that under "PC Health" only the CPU fans and System Fan 3 (the rear exhaust) were reporting RPM info.

System fans 1 and 2, set to the default Normal ie. PWM control like the others. Both are pairs of Akasa Vipers (2 x 120mm/2 x 140mm). They're connected to the MB by Akasa PWM Y splitters and both pairs appear to be functioning correctly. What is wrong is that the MB BIOS is not reporting their RPM status, it just shows "0" alongside each entry.

I tried them on Silent and Manual (x1.0) modes and they changed speed fine but still no RPM report. However when Full Speed is selected immediately the RPM for both appears.

Obviously I've tested the connections and I've also disconnected one of each of the pairs and re-tested with the same results. About the only thing I haven't done is remove the Y splitters and tested each of the fans plugged directly into the MB headers.

If this were a particular problem with the Akasa Y splitter why is the Full Speed RPM being reported and why isn't it affecting the two CPU coler Akasa Viper fans which are also connected through another Akasa Y splitter to their header?

Questions:-

Is this a MB/BIOS issue and something that is likely to be fixed after installing the OS and then the drivers from the disc supplied with the MB?

It just seems a bit unlikely that would fix it since the other fans are reporting RPM in all modes OK straight out of the box.

Is there any cause for concern?

It does appear to be a specific reporting issue only, otherwise the fans seem to be working as intended.

What I don't want to discover later is that there are some known problems with the 1 and 2 sytem fan headers with this MB and have to dismantle the whole thing to get it sorted. I've searched online and can find no similar recorded issues so that's why I've come here for advice.

MB is marked as revison v1.1 and BIOS version shown as F5 if that is any help.

 

[coffeejunkee]

About the only thing I haven't done is remove the Y splitters and tested each of the fans plugged directly into the MB headers.

Well, start with that

Pretty sure it's a problem with the splitter. Or rather, those fanheaders are 4-pin but not pwm, they use voltage control to change speeds. Connecting 2 pwm fans via splitter might confuse the header. You can try and see if Speedfan can readout the rpm.

But if the fans are working and properly speeding up and down I'd probably just forget about it.

 

[Mantrid-Drone]

Yes, I'm going to test each fan without the Y splitters at all ie. just use an extension lead instead. Also I'll try both other 4 pin and 3 pin fans in case it is some odd issue with the Akasa Vipers.

The system headers are all 4 pin and so one assumes PWM as the manual says they are. But it is a dual MB manual covering the GA Z87-HD3 too so there might be differences in function not covered.

There is a distinction, which may just be an omission, made in the manual between the System 1 Fan header and 2 and 3. The fans using System Fan 1 can apparently be controlled by something called "EasyTune" whereas 2 and 3s' function are identically described but don't include any mention of "EasyTune".

But that doesn't look to be any part of the cause of the descibed problem as it is fans connected to 1 and 2 which are not reporting RPM, the exhaust fan on 3 does.

If after more thorough testing I can't track down where the issue is I think I'll try opening a ticket with Gigabyte themselves to see if they can provide an explanation.

 

[coffeejunkee]

Check page 23. Cpu fan has pwm control (4th pin speed control) the others have voltage control (2nd pin 12V/speed control). 4th pin is vcc which is an electrical engineering term for saying it doesn't really do anything.

Now that shouldn't really matter because all headers have rpm readout pin. But the combination of those headers, pwm fans and a splitter might mess things up. It's also not unusual for splitters to give incorrect readings, although usually not 0.

In the end it doesn't really matter, rpm is just a readout, it's not used to control fans. If they work properly you can forget about it (i can understand if you want to figure it out but it's not necessary).

 

[Mantrid-Drone]

Thanks for the info but I'm not sure that's right about the GA G87-HD3. In the manual all three system fan headers are stated to have Manual "Slope PWM". SF2 and SF3 speed control is linked but that's the only difference between the CPU and SF1 fan speed control functions. All it says on P23 of my manual is that the "speed control function requires the use of a fan with fan speed control design". All my fans are PWM Akasa Vipers either 140mm or 120mm.

The relevant page is P43.

Anyway after long and fairly thorough testing I've tracked down the culprits and it is 'just' a reporting issue but it actually raises more questions than answers.

All fans are functioning mechanically 100% perfectly the problem lies with the RPM reporting from two particular Akasa Viper 120mm fans, apparently identical to the three others I've used in the build with no such problem.

The two misreporting ones behave identically and mysteriously also identically to a Scythe 120mm PWM fan I also tested. All the bog standard 3 pin fans (Coolermaster, Corsair) I tested have no similar issue and neither does the 120mm PWM Coolermaster fan that came with my Hyper 212 Evo Cooler.

1). None of the misreporting ones show RPM readings in Normal, Silent, Manual PWM (default settings) modes on any of the three system fan headers.

2). All report RPM correctly at Full Speed, all headers.

3). All report RPM correctly at the highest Manual PWN mode, all headers.

4). When connected to the SF1 header no other Manual PWM setting provided any RPM details.

6). When connected to the SF2 and SF3 headers all these fans reported the RPM in Manual PWM mode with any setting 2 and above.

The only inconsistency with these test findings was that occasionally one of these fans, particularly the test Scythe, would report grossly exaggerated RPM in Normal mode but often this would return to 0 again on reset.

It appears to be a specific reporting problem between these particular fans and the MB. But surely this can't be coincidence. Three fans, one of a different make, which in identical fashion do not report the RPM correctly in three out of four modes but all have no trouble doing so at Full Speed. Seems very unlikely to me.

Any further ideas/comments very welcome.

 

[coffeejunkee]

"Slope PWM" is just a bios term, speaking of which have you updated the bios? Usually a good idea with new boards. And have you tried Speedfan? If that works and reads out the rpm correctly it's likely a bios issue. Or what if you connect the Akasa's (and Scythe) to the cpu fan header, does that give the correct rpm?

 

[Mantrid-Drone]

Thanks.

No, I've not even installed the OS yet. I didn't want to go through all that hassle just to find out it was MB problem and have to start from scratch. What I will do is open that tech enquiry ticket with Gigabyte I mentioned to see if a BIOS update does address this issue.

However I do think I now might now have tracked down what the problem is. It is with the fans or, more precisely, the fan design/spec. This post on the Corsair forum describes a very similar issue to the one I've detailed:-

http://forum.corsair.com/forums/showthread.php?t=119379

It is a high resistance problem on the fan tacho channel. Why it should affect some fans and not others even of the same make and type isn't easily explainable but it does fit the facts.

The fans exhibiting this issue all report RPM correctly in High Speed mode and Manual PWM at the highest settings. The tacho signal in these modes is increased enough to overcome the resistance, that's the theory. It also suggests that when the system is under load and the fans speed is increased in Normal mode they will start reporting RPM too.

It could still be a MB problem and maybe one that could have been fixed by a BIOS update (I haven't ruled that out) but what do you think is more likely?

Of the ten fans I tested both ordinary 3 pin and 4 pin PWMs, 4 different makes, 6 different models, all individually tested on all three headers only the three mentioned exhibited this problem. So MB or fan design issue? My money is on the fans.