In my case, iStats seems to show mA when you hover over the graph (the menu bar icon always shows A). I was recently prompted to install an update, so maybe improvements were made.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Upgrading 2013-2015 Macbook Pro SSD to M.2 NVMe
- Thread starter maxthackray
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
In my case, iStats seems to show mA when you hover over the graph (the menu bar icon always shows A). I was recently prompted to install an update, so maybe improvements were made.
This seems more plausible to me than 0.00A necessarily being bogus (there may be cases where it is, my machine is but one data point). Mine shows 0.00A in times of extremely low (practically none, e.g. when all apps are closed) disk activity, but never shows this low under any disk activity more than a few tens of KB/s, and this behavior is the same whether I am on battery or mains for any length of time. I suspect it's more a data precision issue as your graph would appear to indicate. Unfortunately, even with the latest version of iStat Menus, my graph is stilll showing in A rather mA. No idea how to change that... so I have no way to confirm.
The presentation in iStat Menus indeed seems changed a bit in latest version (?) but I have reservation on the conclusion of a precision issue.
I think the values on the dropdown menu are raw and instant values from SMC. Those values displayed on the graph are data points derived from some sort of averaging or interpolation algorithm. If so, extra precision is not meaningful. Hope someone with more insight in iStat Menus could comment.
_However_, if people follow my instructions in post #8111, and see very low or even 0.00A values when connected to _AC power_, then it's a very encouraging sign. There is some chemistry among the kexts + Big Sur that I haven't seen before.
Keep your observations coming, and post raw data i.e. screenshots instead of your interpretation (unless you really know what you're doing).
I think the values on the dropdown menu are raw and instant values from SMC. Those values displayed on the graph are data points derived from some sort of averaging or interpolation algorithm. If so, extra precision is not meaningful. Hope someone with more insight in iStat Menus could comment.
_However_, if people follow my instructions in post #8111, and see very low or even 0.00A values when connected to _AC power_, then it's a very encouraging sign. There is some chemistry among the kexts + Big Sur that I haven't seen before.
Keep your observations coming, and post raw data i.e. screenshots instead of your interpretation (unless you really know what you're doing).
Last edited:
Hi kvic,
Without SsdPmEnabler the Samsung 970 Evo Plus boots correctly, I'm just not happy with idle (0,42A) and temperature. So that why I bought Corsair 510 and using Evo Plus as external drive.
With ssdpmEnabler installed and Samsung 970 EVO plus my Mac had exactly the same panic error as with Corsair 510 last week. And also I couldn't reboot even with disabling ssdpmEnabler via Terminal, I used to reboot from external drive and delete SsdPmEnabler from /Library/Extensions manually.
To be 100% sure I did try twice.
Yeah.. I could tell. I think the screenshot of one hour session in post #8115 from @oryan_dunn was on AC power too. I think he also had NVMeFix additionally installed during that experimental hour.
I would put my money on that no 0.00A at idle was observed in both cases during those two hours. lol
Hi BoPI, I've created a ticket to better track this issue.
Ideally we would prefer a stock AppleSSD to troubleshoot. Samsung sticks are properly the least candidates for troubleshooting an issues that's already quite a myth. But I could understand AppleSSD won't be readily available in your case. In the meantime, I think your best strategy is to wait for some members to step forward and report more data points on 2014 13-inch MBP in combination with different brands of SSDs.
So your Samsung 970 EVO Plus failed to boot into Recovery Mode after KP. Did you verify later if the disk has data corruption? If so, it's another indication, the SSD and/or this specific MBP model are incompatible.
Kvic,
I will try to install your SsdPmEnabler with my original Apple SSD in and test.
No data corruption. But couldn't reboot back, even with disabling your kext via Terminal. Rebooted from external drive and manually deleted SsdPmEnabler.kext from ../Library/Extensions folder, what helped to reboot Mac normally.
Please check what fault report I get straight after panic error and reboot back
Last edited:
Searched through this thread but I couldn't find a definitive answer. Is the NVMeFix compatible with High Sierra (10.13)? Has anyone tried it?
https://github.com/acidanthera/NVMeFix
My screenshot was definitely on battery, and I've never installed nvmefix, only your SsdPmEnabler.
As for iStats, it seems the 0.00A readings are either bogus (and should be 0.01A or higher) or a result of precision issues (current is actually less than 0.01A), either in iStats or in hardware. I highly doubt the system is completely powering down the SSD while the system is running. Precision issues somewhere seem to be the most likely cause. Are there any other programs other than iStats to get this info? Any built in command line tools to directly query the power sensors?
Edit:
Here is a screenshot from this morning. I had ran it down to dead last night before plugging it in. This hour is on battery, with no kext loaded. For most of this hour, I was using RDP and had Safari open, but not really surfing the web with it. But as soon as I started really using Safari, it bumped up to that plateau you see. I saw this same behavior yesterday without SsdPmEnabler enabled. It would sit at 0.00A, I'd be getting great battery drain while using it (say 6% per hour or so). But once it gets in this state, battery drain increases, and I don't see it pop back to that 0.00A. level.
Last edited:
@oryan_dunn Looks something interesting. What's the lowest value in the above graph? Could you check multiple places along the flat line at the bottom, and perhaps do a screenshot with the tooltip pop-up?
EDIT:
Thanks for the updated screenshot below.
So looking at the graph above basically between ~10:37 and ~11:20, iStat Menu shows your SSD almost consumes no power. One thing you could check is counting the disk activities such as bytes written/read etc during such a period. If there is non-trivial amount, you perhaps will be convinced the power consumption can't be zero. You can also fire up BlackMagic (like I did months ago) and run for a few minutes and check the SSD 3.3V current readings during the benchmark.
EDIT:
Thanks for the updated screenshot below.
So looking at the graph above basically between ~10:37 and ~11:20, iStat Menu shows your SSD almost consumes no power. One thing you could check is counting the disk activities such as bytes written/read etc during such a period. If there is non-trivial amount, you perhaps will be convinced the power consumption can't be zero. You can also fire up BlackMagic (like I did months ago) and run for a few minutes and check the SSD 3.3V current readings during the benchmark.
Last edited:
The flat line is 0A. And in the screenshot below, I've gone back to only using RDP, but the power never did go back to 0.00A like after I first booted the computer this morning. This is with no kext loaded.
Edit: perhaps it's related, I just realized by looking at the chart, the fan turned on around 11:20am, and hasn't turned off since, sitting around 1300rpm.
Last edited:
Good evening,
I own a Late 2013 MacBook Pro Retina.
Just to let you know, I finally made the switch from my 512Gb original Apple SSD to a 1Tb NVMe SSD.
I used the small Gigabase adapter with a Western Digital Black SN750. My operating system is MacOS Big Sur 11.1.
Everything works like a charm
I just have to install what's needed to lower the sleep power consumption and I'll be done.
I own a Late 2013 MacBook Pro Retina.
Just to let you know, I finally made the switch from my 512Gb original Apple SSD to a 1Tb NVMe SSD.
I used the small Gigabase adapter with a Western Digital Black SN750. My operating system is MacOS Big Sur 11.1.
Everything works like a charm
I just have to install what's needed to lower the sleep power consumption and I'll be done.
SmartMonTools or other SMART tools should provide more exact reads from the actual device. I guess that’s is what istats is reading as well.
I recall you had the same machine as me... That's the behavior I've noticed on my 2015 13" MBP - the SSD 3.3V line reads 0.0V from a cold start, until the fan turns on. Then it starts to fluctuate as one would expect.
I looked through a boardview schematic of the laptop - the 3.3V current sense circuit is shared with the fan control, which is why I suspect this happens on our model.
My sisters laptop is an early 2015 13" MacBook Pro. I'll keep an eye on this behavior when I reinstall SsdPmEnabler. I don't recall seeing this same behavior when I had that installed before, but I hadn't paid close attention.
Here are a couple screenshots that show this, top SSD 3.3V, and bottom is the fan rpm:
The fan graph y axis is terrible, at 2:54:12, the fan kicked on, and remained on (the flat line before is 0rpm, the flat line after is 1290-1310rpm, despite it looking flat).
Edit: Right now, it was reading 0.00A and 0rpm on the fan, I kicked off a disk speed test, and was getting normal numbers, but it still showed 0.00A and 0rpm. About 10s into the test, the fan kicked on, and the SSD powered showed 1A or so.
Also, at some point, I'll pop the original SSD back in and watch the behavior under as close test conditions that I can. Before swapping it out, I briefly watched it, and I don't recall seeing any behavior like this.
Last edited:
You really should look into actual kernel panic logs inside Console.app before jumping to the conclusion that the installation of 3rd party SSDs is the culprit.
For area to apply Kapton tape, look at post #1 as a starting point. And exercise common sense from high school science class. Personally I don't think it hurts if you apply a bit more than you need.
The kernel panic is in the NVMe driver which isn't good news.
If you show AppleSSD is fine, then the machine (2014 13-inch) should be safe to assume okay. Next step will be borrowing a SSD listed in this table and see if it works or at the same time other models from 2013 to 2015 could report success (or failure) stories with these sticks:
- Corsair MP510
- Samsung 970 EVO Plus
- Silicon Power A80
Device: Mid-2014 MacBook Pro 15" Retina (MacBookPro11,3)
OS: 10.13.6
Boot ROM Version: 429.0.0.0.0
SSD: Western Digital 1TB WD Blue SN550 WDS100T2B0C-00PXH0
I installed the Lilu+NVMeFix on High Sierra, after first installing and testing SsdPmEnabler. The power consumption seems unchanged as compared to using just SsdPmEnabler. At idle, the lowest that the power consumption will go to is 0.18A. I'm wondering if the Lilu+NVMeFix is not working as expected in High Sierra. Does anyone have a similar or different experience?
Also, the overnight power loss during sleep on a brand new battery is 5%, which no notifications to the stock sleep/hibernation settings.
OS: 10.13.6
Boot ROM Version: 429.0.0.0.0
SSD: Western Digital 1TB WD Blue SN550 WDS100T2B0C-00PXH0
I installed the Lilu+NVMeFix on High Sierra, after first installing and testing SsdPmEnabler. The power consumption seems unchanged as compared to using just SsdPmEnabler. At idle, the lowest that the power consumption will go to is 0.18A. I'm wondering if the Lilu+NVMeFix is not working as expected in High Sierra. Does anyone have a similar or different experience?
Also, the overnight power loss during sleep on a brand new battery is 5%, which no notifications to the stock sleep/hibernation settings.
@manu3569 Thanks for posting your result.
What's your SN550's idle current without any kexts?
High Sierra might not have the best NVMe driver. If you're adventurous, try to port the driver from a newer macOS version. You may try your luck asking how-to in MacPro hackintosh threads.
I can't speak for High Sierra. I could share my brief experience with Nvme + Catalina + SN550 + 2015 13-inch MBP. So our SN550 idles at 0.26A without any kexts. With NvmeFix, we see 0.03A drop regardless of the existence of ssdpmEnabler.
Now the weird thing that I still don't understand is that this 0.03A drop is consistently observed right after every reboot. And then after some use (perhaps a few days with multiple sleep/wakeup but no reboot), this 0.03A drop will mysteriously disappear. Since it doesn't hurt, we still have NvmeFix installed. Perhaps that would make you less bother with it working or not.
With stock sleep/hibernation, 5% battery drain overnight is very good I would say.
What's your SN550's idle current without any kexts?
High Sierra might not have the best NVMe driver. If you're adventurous, try to port the driver from a newer macOS version. You may try your luck asking how-to in MacPro hackintosh threads.
I can't speak for High Sierra. I could share my brief experience with Nvme + Catalina + SN550 + 2015 13-inch MBP. So our SN550 idles at 0.26A without any kexts. With NvmeFix, we see 0.03A drop regardless of the existence of ssdpmEnabler.
Now the weird thing that I still don't understand is that this 0.03A drop is consistently observed right after every reboot. And then after some use (perhaps a few days with multiple sleep/wakeup but no reboot), this 0.03A drop will mysteriously disappear. Since it doesn't hurt, we still have NvmeFix installed. Perhaps that would make you less bother with it working or not.
With stock sleep/hibernation, 5% battery drain overnight is very good I would say.
That's something else I was testing. On my system, overnight (12hrs), battery dropped 20%.
I looked at what pmset showed, and it seems the default for Big Sur (and perhaps others, not sure), is 24hrs before hibernate kicks in when battery is above 50%, and 3hrs when battery is lower than 50%. Even then, the hibernate mode does not power down the RAM. Is this common for default settings on a MacBook? It just seems very strange it'd be 24hrs before hibernate kicked in. I know you can change it, just still quite odd for a default, as it would likely drain half the battery or more before actually going to hibernate.
@kvic - Thanks for your response and insight about potential driver related current differences.
With all 3 kexts removed, the lowest I see the current go is 0.31A.
And when loading the SsdPmEnabler.kext again, the current drops back down to 0.18A.
@dannytang - What macOS version are you running?
With all 3 kexts removed, the lowest I see the current go is 0.31A.
And when loading the SsdPmEnabler.kext again, the current drops back down to 0.18A.
@dannytang - What macOS version are you running?
Yes, perfectly normal. Sometimes turning off Bluetooth helps, as well as disconnecting any peripherals.
I set the standbydelay settings on my machine to 30 mins and 60 minutes to help prevent the overnight drain.
is there a guide to installing a new ssd with big sur? i see a lot of posts about bootrom and nvmefix and im not sure what it all means, sorry
I have a macbook air 2015, and am looking to upgrade the ssd to a 500gb kingston a2000. my bootrom version is 425.0.0.0.0, is that a good thing?
i know i have to backup my data through something that isnt time machine, but then how am i supposed to access the data after installing ssd?
I have a macbook air 2015, and am looking to upgrade the ssd to a 500gb kingston a2000. my bootrom version is 425.0.0.0.0, is that a good thing?
i know i have to backup my data through something that isnt time machine, but then how am i supposed to access the data after installing ssd?
Last edited:
I'm running Mojave, 10.14.6. And if it matters, the Sabrent Rocket Pro is the Phison E12 running ECFM12.3 firmware.
@manu3569
thanks again for the follow-up. Those graphs are very pleasing to look at and IMO their jigsaw shapes are representative of normal systems.
@dannytang
thanks for posting your result. Concise yet detailed.
So you shucked a Sabrent Rocket Pro. What do you end up re-purposing the enclosure? I might be completely off track.
I've added both results to the table in the User Guide.
On our 2015 13-inch MBP + SN550 1TB, I measured 2% per hour battery drop during sleep under stock sleep/hibernation settings. You got 1.7% per hour. That's even better, and expected as herb2k said.
I believe Apple's choice of the defaults is to minimize wear & tear on NAND chips (less durable & expensive but replaceable in older systems) and again NAND chips (more durable & cheaper but soldered to logicboard in newer systems). In both cases, NAND chips lasting longer serve Apple's interest, and perhaps people who strongly believe in Apple's way.
Apple won't tolerate 2% per hour drain during sleep. I think their proprietary SSDs in these older machines are capable of power-off and this facilitates the whole system entering very low power consumption during sleep. I believe that's not the case when 3rd party SSDs are installed.
Battery drain during sleep is easiest to fix by entering hibernation earlier. Idle power consumption is harder to improve. Here comes ssdpmEnabler in an attempt to address this aspect, with further savings from NvmeFix/Lilu in some circumstances. OTOH, high power consumption during active/read/write has no way to workaround for end users.
Hopefully people know how to prioritise when picking a new SSD for these older machines.
thanks again for the follow-up. Those graphs are very pleasing to look at and IMO their jigsaw shapes are representative of normal systems.
@dannytang
thanks for posting your result. Concise yet detailed.
So you shucked a Sabrent Rocket Pro. What do you end up re-purposing the enclosure? I might be completely off track.
I've added both results to the table in the User Guide.
On our 2015 13-inch MBP + SN550 1TB, I measured 2% per hour battery drop during sleep under stock sleep/hibernation settings. You got 1.7% per hour. That's even better, and expected as herb2k said.
I believe Apple's choice of the defaults is to minimize wear & tear on NAND chips (less durable & expensive but replaceable in older systems) and again NAND chips (more durable & cheaper but soldered to logicboard in newer systems). In both cases, NAND chips lasting longer serve Apple's interest, and perhaps people who strongly believe in Apple's way.
Apple won't tolerate 2% per hour drain during sleep. I think their proprietary SSDs in these older machines are capable of power-off and this facilitates the whole system entering very low power consumption during sleep. I believe that's not the case when 3rd party SSDs are installed.
Battery drain during sleep is easiest to fix by entering hibernation earlier. Idle power consumption is harder to improve. Here comes ssdpmEnabler in an attempt to address this aspect, with further savings from NvmeFix/Lilu in some circumstances. OTOH, high power consumption during active/read/write has no way to workaround for end users.
Hopefully people know how to prioritise when picking a new SSD for these older machines.