Precision 7510 Owner's Thread

There are parts available to convert the onboard SATA slot to M.2 slot.

WPTND - M.2 Interposer Board for 7510/7710

Y1WJX - M.2 bracket

HR8P8 - Thermal cover

Its just that the second drive would not work in NVMEmode, but in SATA mode.

For me also. As far as I understand, the fans are actually running too low. I cannot even log in, before the computer shuts down. I have HWinfo taking over the fan control after login, so if I can keep the machine cool enough ( I just take the laptop and move it up and down one meter quite fast, so that more air would get in until login screen appears and then quickly log in) then HWinfo takes over the fan control and everything resumes normal. But I noticed I have to make the fan curve much more agressive to keep the temps reaching upper 90s. I also noticed that HWinfo logs 98 degrees as Core MAX. so there is definately something either making the chips run too hot or messing with temperatures, so that they appear hotter than they actually are. Right now I can keep my laptop running if I set manual fan control 100% from HWinfo and respin period to 1ms (the faulty fan control tries constantly to take over and the fans spin down sometimes for a brief second)

Right now I can only start up in safe mode and then disable all services except from Microsoft or Dell. If I let it load any more, it will never reach login screen.

I also noticed a weird black screen pause before login screen appears.

I also got thunderbolt installed and shown in bios (although grayed out), even though I have no thunderbolt ports or controller.

Anyways, I will be trying to roll back BIOS or somehow take over fan control before login and wait for next update.

 

1. (Obviously) You have a major problem if your system is running this hot. If everything is right with the heatsink assembly connection (thermal paste applied properly and everything screwed down) then the Intel CPU should throttle itself to an appropriate level if it gets too hot. It should be able to run at a low clock speed even if the fans are not running at all. Have you done anything to override the CPU throttling behavior (i.e. ThrottleStop)?

2. DellFanCmd (see signature) can override the built-in fan control, you could use it with a Task Scheduler job to force the fans to max much earlier in the boot process and prevent the EC from trying to spin the fans back down.

 

While great, it wasn't what I was asking. Thanks anyways!

 

1. I have now come to a conclusion that its not running actually that hot. Its the system reporting wrong temperatures. It has been a while from the last dust and thermal care though and I will do that today. If it will help even with 5 degrees, I can save a lot of time troubleshooting the issue, avoiding critical thermal ID86 events. I had everything smooth and normal until the Intel-Dynamic-Platform-and-Thermal-Framework_74P16_WIN_8.3.10209.6897_A12.EXE update. I also let the Dell Precision Optimizer update BIOS to 1.19.3 and it stated, while updating, "Sending Intel(R) Management Engine Firmware Update" and "Updating EC Firmware" I now downgraded BIOS, and downgraded the Intel ME Thermal-Framework, to A03 as suggested in one of the posts, but with no luck. I am also running Windows 1909 that was said to be affected with these issues.

2. I will check that out, if I won't find any other solution, to use it as a workaround. But the scripts you made later for apps, seems very useful, even after the issue will be gone. I really like the idea that when I start certain app or rendering, I can just tie the executable of the rendering or calculation engine to the app, so that I can ramp up the fans before the CPU gets hot and throttling. It is good for both performance and lifetime of the cpu. I will also chip in for the certificate pot when I do.

3. I have now noticed abnormal temp spikes. They shoot sometimes from 40 to 88 degrees in two seconds while fans are spinning 4800 RPM.

 

Not all, but most WWAN slots only have a USB connection, power and a couple traces for the SIM slot interface.

 

Might not be that abnormal. I've noticed that the CPU temperature in my 7530 can spike very quickly if the load goes from basically nothing to full (even briefly). That doesn't seem to happen in my older M6700. I wonder if it is just because the turbo CPU frequency a good bit higher.

(Though also, I have applied good paste in my M6700 and I have never touched the cooling solution in my 7530.)

 

I opened the system and applied real paste as soon as I got the 7510. Just to make it last a bit longer. The cooling today was, however blocked around 20% and the paste was not in good condition. Takes around an hour if you collect your screws carefully. After cleaning I got temps in low 30s in idle while fans are maxed. so around 15 degrees. It might just be that it helps me to work around the issue. If I find a solution, I will write here.

 

Got it fixed. All I did was install previous versions and then back to current + cleaned cooling:
1) Chipset_Driver_NWWR5_WN32_8.2.10900.330_A03_03.EXE
2) System Bios to previous version: Precision_7x10_1.18.5.exe (Installing previous verisons of BIOS must be turned on in settings)
3) Cleaned fans, radiators, applied new thermal paste (temps went down ~15 degrees)
4) Latest Quadro New feature drivers from Nvidia page, 436.30
5) Intel-Dynamic-Platform-and-Thermal-Framework_74P16_WIN_8.3.10209.6897_A12.EXE
6) Windows update for 1909, KB4550945
7) System BIOS back to current: Precision_7x10_1.19.3.exe

I have now been running Aida64 stress test with both GPU stress for 45 minutes, with CPU settled at 84 and GPU at 55 degrees.

Is it possible, that the new thermal management is fine tuned to CPU loads and the actual solution was cleaning? Maybe.
While issues started with the latest thermal management FW update, I have no thermal spikes after cleaning, that indicates I was too sure about my thermal paste application 2 years ago.
That is why it feels like the fans do not react to actual temperatures anymore, but to CPU loads. More realistic though is, that the fan curves are not that agressive. On that AIDA64 test, even while the CPU was over 80 degrees, the fans hovered around 4200. And these are maglev fans, there is no point to "save the bearings" The fans are levitating.
I will be using Speedfan to get fans tu ramp up 100% after 78ish degrees without disabling the built in EC. that seems like a sweet point to avoid constant on and off.

I hope this helps someone else one day. Most important takeaway- clean your machines.

 

The problems returned.
It seems like the new thermal framework thermal throttles less, lets the CPU be more agile and therefore have huge temp spikes. The Dell EC, however not designed to handle it.

Well I discovered that Dell has a special application to adjust our cooling performance. I was looking for it in BIOS and therefore also in Dell Command, only to discover disabled options- not available in 7510, presumably in other laptops.

The solution is- Dell power manager application

If the application is able to control your CPU throttling and fans, there is "Thermal management" menu in the left:



I chose "Ultra Performance"

This solves the long-lasting problem of not being able to choose cooling strategy.

You can download / read about the app the app from Dell support here:
https://www.dell.com/support/conten...ase/software-and-downloads/dell-power-manager

 

I am sorry for hijacking the thread.

It appears, that this did still not actually help. So I found a workaround.
It endured AIDA64 stress test, but when I started gaming, the PC still got thermal shut downs again.

The issue seems to be, rogue temperature reports by cpu with Turbo boost turned on.
I noticed, using Throttlestop, that whenever Turbo boost is disabled, the temps drop 15 degrees.
Throttlestop changed the profile with Turbo boost enabled and without turbo boost 2x per second.
And while the CPU was cooling down the temps dropped nicely, but with enabling / disabling turbo boost, the temps dropped and rised 15 degrees every time, twice a second.
So I suggest the issue is False temperature readings with turbo boost enabled.

As a temporary or (permanent, if it ever gets fixed) solution I have now done some more things and I am able to game now:
To get better fps, while gaming I Installed HWinfo and triggered it to fans max above 65 degrees. So whenever I game I switch it on.

To not get shut downs, I Installed Throttlestop, to:
1) Undervolt the CPU with FIVR -80mV
2) Disabled CPU Turbo from BIOS
3) Enabled Turbo boost on main performance profile
4) Turned on Speed shift - EPP 64 instead of Intel Speedstep
5) Enabled DTS 20 alarm to fail safe profile with no turbo max freq multiplier and without Turbo boost

This results turbo boost being shut off above 80 degrees and CPU clocked to base frequency (I plan to change this to 90, since its 15 degrees faulty and that would mean cozy 75 degrees). Turning turbo boost off, immediately gives me back the 15 degrees fault to prevent shut down. so the profile will keep changing from then on twice a second and shut down will never happen.

While this is a workaround I started communicating with Intel support. Its definately the chipset firmware.
Throttlestop has also a possibility to run a command with each profile change, so it is possible to run Aaron44126 dellfancmd commands on each profile change. I have yet to try that out.
http://forum.notebookreview.com/thr...ual-custom-control-of-the-laptop-fans.827106/

I had alot of help, following these guides:
Lower temps by disabling Intel cringe-worthy graphics features / tweak nvidia control panel / Use Speed Shift instead of Intel Speedstep with throttlestop - https://bradshacks.com/tweaks-laptop-faster-more-efficient/
Undervolt and set up Throttlestop - https://www.ultrabookreview.com/31385-the-throttlestop-guide/
Set up Throttlestop to run at each startup - http://forum.notebookreview.com/threads/the-throttlestop-guide.531329/page-453#post-10247294