It’s incredible the coverage generated by yesterday’s blog article on the further iMac upgrade restrictions. I’d like to personally address some questions, context, and provide additional technical detail concerning this issue.
I want to be very clear that I think these are absolutely the best iMacs ever. These machines up the game considerably and provide performance that can even match up with the Mac Pro for a lot of applications. We’ve been excitedly covering these new iMacs starting with an unboxing and teardown blog post just hours after they were first introduced.
Most iMac buyers (and buyers of anything Apple in general) are more than satisfied with how things are right out of the box. The vast majority will never even think about after purchase options that Apple does support, such as installing additional memory. These are great systems right from the get-go… But pardon us or anyone who wants to make them even better. :)
What is our testing environment/parameters?
In our lab, we have all the different flavors of iMac 21.5″ and 27″ 2011 models currently shipping. This includes one or more of each model processor option and/or video option. Apple’s current latest 2011 iMac models have Apple Model Identifiers of iMac12,1 for the 21.5″ and iMac12,2 for the 27″ regardless of what base model/options selected.
The OS installed is 100% up to date as of all updates available today including the recent EFI update which
enabled 6.0Gbp/s SATA 3.0 capability with two of the iMac’s three SATA ports.
From the iMacs, these model drives were found present in the systems we received:
- Seagate ST3500418AS 500GB 7200RPM Barracuda 7200.12 7200RPM 16MB SATA 3Gbp/s
- Seagate ST31000428AS 1TB 7200RPM Barracuda 7200.12 7200RPM 32MB SATA 3Gbp/s
- Western Digital WD1001FALS 1TB Caviar Black 7200RPM 32MB Cache SATA 3Gbp/s
Note: You don’t get to pick what brand hard drive, cache, model, etc are installed at the factory. Apple iMac Specs do confirm all factory hard drives are 7200RPM. Of further note, the base $1199 iMac model doesn’t have any factory drive configuration options – it is only offered with the 500GB capacity drive.
The following non-Apple Drives are new units pulled from OWC’s drive inventory:
- Hitachi 0S03208 3.0TB 7200RPM Deskstar 7K3000 64MB Cache SATA 6Gbp/s
- Hitachi 0F12115 2.0TB 7200RPM Deskstar 7K3000 64MB Cache SATA 6Gbp/s
- Hitachi 0F10311 2.0TB 7200RPM Deskstar 7K2000 32MB Cache SATA 6Gbp/s
- Seagate ST33000651AS 3.0TB 7200RPM Barracuda XT 64MB Cache SATA 6Gbp/s*
*Unaltered, current drive – but also used for other previous testing.
- Seagate ST2000DL003 2.0TB 5900RPM Barracuda LP 64MB Cache SATA 6Gbp/s
- Western Digital WD1002FAEX 1.0TB 7200RPM Caviar Black 64MB Cache 6Gbp/s
- Western Digital WD2002FAEX 2.0TB 7200RPM Caviar Black 64MB Cache 6Gbp/s
- Any SSD (tested ours and competitor 3.0Gbp/s and 6.0Gbp/s models)*
*Unaltered, current drives – but also used for other previous testing.
***and a ‘zinger/wildcard***
*Unaltered, current drive – but also used for other previous testing. Same Western Digital WD1001FALS 1.0TB Caviar Black 7200RPM 32MB Cache 3Gbp/s model as we found in one of our factory iMacs, but this one was pulled out of one of our 2010 Apple Mac Pros. This is an Apple Factory/Apple Rom unit from before these new iMacs rolled out of the same drive model p/n of one of the 1TB models we’ve so far seen used in the new iMac 2011.
Note: Systems were booted and then left with sleep off, display on. We did not put these iMacs under any kind of stress load that could otherwise impact the system temperature and result in fan speed changes. The ambient temperature of the room being tested in was between 72 and 76 F.
Apple Rom/Factory Drives from received iMac 2011 models:
Regardless of what drive brand model was factory installed, any of the different drives we received in any of the iMacs worked fine in any other…
WD where Seagate 500 or 1TB was and vice-versa. Apple Rom/Factory drives worked interchangeably and no fan rev or Apple AHT failure. Even after a couple hours, the reported HDD Bay fan speed remained right about 1100 in both 27″ and 21.5″ 2011 Apple iMac models as reported by SMCFanControl.
Results from all other drives we tested in those iMac 3.5″ drive bays:
(including the same model Apple Factory Rom 1.0TB WD drive from our Mac Pro 2010)
Didn’t matter which drive, no drive, or which iMac. Consistently the same result in each and every one of our iMac model variants. AHT test = FAIL due to thermal sensor failure.
After startup, the HDD fan would start to ramp up from the 1100RPM base and into the 5600RPM range as reported by SMCFanControl. This didn’t happen instantly, but over the course of 30-40 minutes typically. Within just a couple minutes though, the fan speed would be up over 2500RPM and it’s a steady rise from there before the peak.
Apple AHT – why this not working can be a problem:
When you have a problem with your system, Apple Hardware Test is often suggested for determining if the problem is… well… with a hardware component. If you can not run this test, it could impact support or even warranty response for an iMac with an issue unrelated to the hard drive. Certainly in a production environment, you want to know your system is 100% go-go-go. Not being able to run Apple AHT can be a detriment in this way, although admittedly something that many can do just as well without.
Depending on your environment, this may not matter. While the fans are definitely noticeable in an quiet space, Apple does use fans that are very quiet even at higher speeds. The noise I observed is more from actual air movement through the channels vs. noise from the fans themselves.
In my opinion, the following should be an industry standard thing…. and maybe it will be. Currently, each Hard Drive manufacturer has different pins for the available thermal sensor line found on today’s hard drives. As an example of this: owners of 2010 iMacs can more easily replace their factory hard drive with a retail available drive, as long as they stick with the same make (it’s still a major job just to open up these machines, independent of other challenges). They could also use a different make, but they would need to make some modifications.
What I am trying to say is that Apple didn’t make any barrier with the 2010 (or 2009) models with that respect, the barrier to plug & play drive options in this case is a lack of industry drive standards on the extra drive pins. So, perhaps use of this additional line on the power connector (which is standard on all SATA drives) might be an industry standard in the future.
And if so – then it’s only Apple today, with a firmware set that enables this feature, which just isn’t standard on retail drives yet. If so – it might not be as big a deal. There is still a question I’d like to know the answer to – why isn’t Apple simply using the S.M.A.R.T. reporting feature of today’s drives for the drive temperature information? That doesn’t require/use any extra pin out/line out and I am not aware of any disadvantage from utilizing this option. Maybe someone out there has insight on Apple not using the thermal data via S.M.A.R.T.?
What’s different between the power connection/leads for the HD bay vs. the other bays:
First, note that the HD bay has a different power connection with seven total leads, two of which are used for thermal feedback. Normal SATA power connections, including those going to the SSD and optical bay in the 2011 iMacs, utilize five total leads.
Depending on the Apple factory drive brand, voltage on this line is held at a steady rate with a pulse about every two seconds during which the voltage falls within a consistent range before coming right back up. What’s interesting is that the voltage range, and this is contradictive to my suggestion this could be a future industry standard thing, is that the different brand drives have a different ‘hold voltage’ and ‘drop pulse’ range observed. Furthermore,when we connect up a drive that was not originally shipped in one of these iMac 2011 models, the pulse pattern is no longer present.
Although the differences between drive brands here raise questions about possibility of industry standardization, the fundamentals appear similar – just a different starting voltage possibly. Noting that external thermal sensors provide the thermal data by voltage passed – with resistance on the line affected by the temperature of the sensor that results in voltage read change, so still could be a standardized industry function future forward…. Either that, or Apple has trusted firmware/data it gets from the installed drive and unless it’s the Apple drive – it doesn’t recognize/do the thermal profiling via the power line…. Ok, that’s pretty confusing…. We have done other testing with external options, but it’s so far inconclusive – so can’t say anything is 100% definite yet other than: either Apple has custom firmware in its drives that enables something, or Apple has firmware on the drives that it looks for to use what is there anyway.
*NOTE – NOT FULLY TESTED/Approved/Recommended, etc.* I strongly caution, we’re looking at a number of scenarios and doing a lot of testing to determine what is going to make sense and be reliable and implemented in a way that best maintains the Apple warranty on the system as well.
If we short those two lines vs. connect to the drive (factory or otherwise), the fans don’t rev and Apple AHT does pass. We have been aware of this and see upgraders in the field now reporting this as well. When you do this though, our best guess is that the thermal reporting in the iMac is being told that the bay is empty. When we see an Apple stock iMac with SSD only, we’ll be able to see if this is how Apple handles that bay being empty.
The problem with this solution is running a drive in the bay means that heat isn’t being measured. Of which, a consequence could ensue for obvious reasons. The bottom line is that the iMac HDD bay can heat up and the iMac doesn’t know and doesn’t do anything about it.
HDD Fan Control appears to be another option and a solution to the above.
This utility uses the S.M.A.R.T. Data to monitor drive temperature and allows you to set the temperature parameters for what speed the HDD bay fan runs at. It will hold the fan speed at bay, based on temperature parameters you give it, and prevent the fans from revving up to unnecessary and noticeable levels…. Again, th0ugh, why doesn’t Apple use S.M.A.R.T. Data?
By itself, this software can solve the fan rev up issue. Combine that with the short solution above, and you appear to have an iMac with fans under control and that you can run Apple Hardware Test (AHT) with normal results.
One concern with using any utility to limit the fan speed is what if your settings are too low for heat? We don’t know Apple’s exact parameters or collected data points from the HDD bay for thermal, and there is a potential for issues should the bay be allowed to become too warm without the proper fan speed engaging.
The bottom line is that work around options are in play now, and with further testing, we should be able to be confident in one solution of another. Whatever the solution – the best solution, in my opinion, would be not needing to find a solution in the first place. Certainly it’s a very small percentage that are even going to want, or need, to replace drives in their iMac… Apple made that hard enough as it is… But for those few, it didn’t need to be this hard.
When we have more on this, we’ll share it right on the blog. And…
I absolutely welcome any feedback and details concerning work arounds you’ve implemented, especially if you have installed a non-Apple drive and had AHT pass. I’ve read various posts out and about, but most don’t have a lot of detail (and I suspect really didn’t check fan and/or AHT results and also seen an initial all good report and then a follow up post after noting the fan and AHT issue).
We’ve developed a workable hardware solution to the fan sensor problems and are now offering it as part of our Turnkey Program for 2011 iMac.