Precision 7530 & Precision 7730 owner's thread

Hello to everyone on this thread.

I am writing this to put into perspective my experience so far with Dell as well as to ask for ideas and suggestions on how should I proceed forward, given the issues mentioned below.

Last year, I have purchased a Precision 7730, with the following specs:

- Intel Xeon E-2186M, Six Core Xeon 2.90GHz, 4.80GHz Turbo, 12MB 45W, w/Intel HD GFX
- UltraSharp UHD IGZO 3840x2160 AG,Cam/Mic, w/Prem Panel Guar 100% color gamut LCD
- Full Scan FIPS finger print reader and Smart card
- 64GB, 4x16GB, 2666MHz DDR4 ECC
- 1 x M.2 256GB NVMe PCIe Class 40 SSD
- 6-cell 97Wh Lithium Ion battery with ExpressCharge
- Nvidia Quadro P5200 w/16GB GDDR5

I also purchased PRO support from them:
- 5Yr ProSupport Plus and Next Business Day On-Site Service
- 5Yr ProSupport Plus Keep Your Hard Drive
- 5Yr ProSupport Plus Accidental Damage Protection
- 5Yr Diagnosis On-Site Service

For this config, I've purchased 4 additional NVMe drives (Samsung 960 Pro, 2TB each) and I've replaced the factory NVMe drive. The device has 4 NVMe slots:
- One near the battery
- 2 near the right fan, just a bit upwards to the one near the battery
- One slot near the RAM slots

For over a year I've been struggling with high temperatures of this system and this is truly visible while using the NVMe slot near the RAM. The ridiculous thing is that the NVMe slot is placed directly under the PCH diode and while using that slot, the diode reaches temperatures up to 130°C which will cause the laptop to cut its power in order to prevent burn-in. Somehow, the NVMe slots will have current flowing through so they will always be warm, thus creating a temperature lock in-between the NVMe drive and the Motherboard's PCH diode on the other side (which is not linked to any heatsink at all, nor any cooling solution).

I have been trying to get this sorted out with Dell Support as I have a PRO support with them, but the issues I've got with their support is absolutely ridiculous. Yes, this is a rant about Dell PRO Not-So-Pro support.

I called Dell Support to have this sorted out. They replaced my motherboard about 4 times and then they said they need to further look into this into their lab and because I have a PRO support, this shouldn't take more than 5 business days. Apparently, those 5 business days turned into 6 weeks long and they returned it to me with the note "Too expensive to repair" although the issue was kind of solved. The PCH was not getting that hot anymore, but it would still stay at around 100°C in IDLE.

They gave me a brand new laptop with the same config, however, the new machine had a flaw, which is that Bluetooth was completely disabled in the Motherboard. I later found out the PCH temperature issue can easily be fixed if you open up Advanced Power Options on Windows 10 and on the PCI Express and Link State Power Management options, I needed to set everything to "Maximum Power Savings". That fixed the PCH skyrocketing to a lava state.

I called them again to have this fixed, they scheduled a technician to replace my motherboard on my new system. I was not at home at that time, so I had a relative take care of this and to my surprise, when I got home, the bluetooth was working, system was operational, however, I didn't notice until later in the evening that my system was reporting that the dGPU was not connected. The technician who was supposed to handle this actually broke the DGFF connecting cable. I said "Ok, this must be just the cable". I still had my old Precision back then, so I opened it up and swapped the broken DGFF cable with the one that was intact on my old system which was sent for RMA. To my surprise, with the perfectly looking DGFF cable, my dGPU is now visible in the system, however, I am getting BSOD with "Invalid power state" for the dGPU, so the issue is not just with the cable. My hunch is that the technician who broke the cable also scratched or did something to the connector itself, thus rendering my dGPU unusable.

I called Dell support again, told them about this, they booked another appointment to replace both the motherboard and the dGPU. So far so good, except that was over a month ago, due to the availability of parts (mainly motherboard). In the meantime, I've had a short vacation and the engineer that was scheduled for this called me during my vacation telling me that they have the parts. I told him that this needs to be postponed because of my vacation and agreed to do this on Friday, November 1st. On Friday, the technician couldn't give me an ETA on when he could be there and I also needed to handle some personal issues on the afternoon so he specifically mentioned that the appointment needs re-booking.

I said that is ok and later in the afternoon, to my surprise, my dispatch and ticket got canceled, without even getting consulted on this.

I called them again, told them everything about my case and asked to have it reopened. They had to open a separate Dispatch case for this and now the "fun" begins as the status is again "Parts not available".

I'm living in EU (Netherlands) and I find this utterly disappointing in terms of support and I've never had such issues with any support team ever (I mean it). They didn't even send me an email to ask if it's ok to close my tickets. An agent sent me an email later in the day, after they canceled the ticket and the dispatch asking for my availability, but not mentioning that they got my case closed.

I am now looking towards this community for a solution as this has become a real nuisance for me. This issue wasn't supposed to even happen and the dGPU problem was caused by one of their technicians. I am reluctant to open up the laptop and re-attach all the cabling as I don't want to break anything there (Even if I have the accidental damage protection).

What suggestions can you give me for this? I really don't want to wait for one or more months to get this fixed and when I call them, they say they will get back with an update, however, they are mostly ghosting me after this incident with the ticket closure.

 

@Soromeister, that's a disillusioning experience indeed, primarily because it was a combination of bad timing, your not being present (not your fault—you have things to attend to rather than preside over a technician fiddling within your notebook) and poor handling of your issue on Dell's end.

There are a few options you can now consider, in increasing order of 'fang length', to coin a phrase:

1. Hang in there, and wait for things to proceed
2. Try to fix it yourself: take a magnifying glass to the DGFF ribbon cables and the power cable, and see if any pads or lines have been damaged
3. Call Dell again and again, and email them about your problem (with evidence from the above, if possible)
   
4. Raise up this issue with your country's consumer rights protection agency/bureau—this is sure to get some gears turning, because as far as I am aware, the EU is excellent at this sort of thing.

I hope everything turns out OK, because I've had excellent support thus far on my machine (barring weird problems like the battery life issue and such).

 

First of all, thanks for the prompt reply. My next step is to wait one more week. If that's not done, then I'll go with option 4.

Have you managed to find a solution to the Lightroom power draw drop? I mean, did Dell manage to provide you with a solution to this? Alternatively, I guess you can also direct this to Microsoft, as it's already possible to get this low power draw just by using that Lightroom trick so I'm guessing they could issue a patch as well for the OS. It's not only a BIOS-related thing here.

 

All the best with this—I hope you get your laptop back to 100% once all this is done and gone.

Not quite: I'm still on an email thread with a local Dell technical support engineer who's liaising with engineers in China about the whole debacle. If I really wanted to get my hands dirty, I'd say I would involve the developers and manufacturers of the entire product stack to take a look at this issue: starting with Dell, then NVIDIA for the GPU and their drivers, and then Microsoft for the OS. That being said, I still do end up getting a neat 8+ hours of screen-on time with Illustrator open and minimised (I've switched to keeping Illustrator open now, as Lightroom tends to hog my memory).

I've tested it with a straightforward stopwatch, too: if I don't breathe on the notebook and maximise every single power-saving measure (networking and connectivity off, screen brightness down, updates switched off, extraneous background processes closed), the power draw is so low (~3-4 W) that it takes a full day to completely drain the battery. With said Adobe app closed (aka nothing is running, 'true idle'), the battery lasts a quarter as long, and the notebook dies in six hours flat.

This honestly needs more publicity, as it's a seriously weird hack.

I haven't tested this with Linux yet as I haven't managed to completely install a distro with the necessary drivers and configuration for Optimus.

 

I have noticed a few of the newer M.2 NVMe PCIe SSDs are gen 4. I think the 7730 only has M.2 PCIe Gen 3 slots correct? If that's right, I presume there is no benefit for me to buy a gen 4 SSD when the bottle neck will be with my motherboard? Thank you!

 

None whatsoever. The only chipsets that support PCIe 4.0 now are from AMD, anyway.

 

Future-proofing, maybe? If you think that you might want to carry the drive forward into a future system.

 

So I finally signed up to this forum to participate since I am reading a lot of the stuff going on in this thread. And big thanks out to @Aaron44126 (for this Thread all the tips you provide and the amazing DellFanCmd), @Mazed (for the P3200 Dump of the 7730) @Ionising_Radiation (DGFF Upgrade Tutorial and benchmarking the P3200 with the two bioses).

I am also a proud owner of a Dell Precision 7530 (Nicely specced - Intel Xeon E-2186M (with Liquid Metal Thermal Paste upgrade - running a stable CoreV Offset of -0.150V), 128GB DDR4 ECC RAM, Quadro P3200 (with Liquid Metal Thermal Paste upgrade - vBios upgraded), Killer WiFi 6 AX1650 upgraded, DW5811e LTE, 3x 2TB Samsung 970 Evo Plus)


Since I love to push everything to the maximum (1363 Cinebench Points with this 7530). I ended up in this thread many many times. But I think I never found someone complaining about the Power Limit Throtteling. I currently run Bios 1.5.2 (which is newer than the 1.11.0) but the number is lower I don't really get Dell there... and I run into Power Limit Throtteling issues all the time (even with my strong undervolting) I don't know if it is the newer Bios or if they changed something else. But I can remember doing the Liquid Metal Mod back in November last Year and doing Benchmark after Benchmark after Benchmark with no Power Limit Throtteling.

Things get even worse if you have your Precision 7530 hooked up to the TB18DC... for some reason it messes around with the Turbo Boost Power Max and puts it to 5W?!? Or so...

Temperature has never been an Issue on this notebook since the Liquid Metal Mod (I attached some pictures for those who love nerdiii tech like me)

 

Welcome...
I've seen some recent posts in the 7540/7740 thread regarding CPU power throttling from @Hopper82. Maybe the situation is similar with the 7530/7730.
http://forum.notebookreview.com/thr...40-owners-thread.830037/page-29#post-10956566
http://forum.notebookreview.com/thr...40-owners-thread.830037/page-29#post-10957053
http://forum.notebookreview.com/thr...40-owners-thread.830037/page-31#post-10959234

Haven't run into this myself, I haven't done any messing with the thermals in the 7530 so it hits the thermal limit first. I do totally plan to get into this when I replace my personal system (about two years out from that, waiting for Intel to release some workstation CPUs from their new architecture). Liquid metal makes me nervous though

 

Okay, I will check this out.

Funny, before I did the Liquid Metal Mod I was running into the Thermal Limit all the time. Now I never ever hit over 80°C even on permanent benchmarking. And I could bet I did not have that Issue when I first did the mod back in November last year.

"Liquid metal makes me nervous though" I totally agree with you on that. Killed two quite expensive Asus Zenith Extreme Boards on my first Liquid Metal encounters, doing crazy workstation build, ended up doing a world record attempt https://hwbot.org/submission/3934452_loadit_cinebench___r15_ryzen_threadripper_2990wx_6683_cb).