Dell Precision 7540 and 7740 Owner's Thread

Well, I still don’t know what caused the screen to go black in the first place, but after running a bit under direct output, I gathered my courage and switched the BIOS back to switchable graphics. I was more patient this time and allowed the computer to continue boot up even though it appeared to still be a blank screen. To my relief, my password screen did eventually come up, and since then, everything seems to be normal and I’m running under the Intel processor graphics. In the BIOS, I have switchable graphics selected and direct output unselected (when I first changed back to switchable graphics, it appeared that direct output was checked as well, which made no sense…I went back into BIOS and made sure that the direct output was unselected).

Sure wish they would at least include a splash screen or something telling you what’s going on (“GRAPHICS PROCESSOR RESETTING” or something) because the way it is you have no idea why there’s a blank screen and no apparent way to get out of it other than the power button. I’ll have to remember the Win+Ctrl+Shift+B thing - thanks!

 

It appears to be an issue with the Intel Graphics Command Centre.

I'm also having this issue with a new Precision 7540 with 4k display and a T2000. Latest BIOS* and drivers.

Although its for an XPS 10th gen Intel, this link seems relevant:
https://www.intel.com/content/www/u...ics-for-10th-generation-intel-processors.html

So in the Intel Graphics Command Cente:

• Try to create a new color profile? *boom* black screen!
• Try to "restore original colors"? *boom* black screen!

The solution I found was to restart many times using the power button until the machine gets tired of it and deigns to offer the Recovery Options, and then go revert to a recent windows restore point. (I hear that going into safe mode and uninstalling the intel uhd drivers should also work).

I had wanted to adjust the saturation because the colours on this machine are ridiculously oversaturated and videos look terrible.
I can't use Dell Premier Color because it claims not to work on this machine - Dell have just got back after a few months of investigating saying they want to try replacing the LCD part as its *supposed* to work but ¯\_(ツ)_/¯

So I'm left with cartoon colours on what is supposed to be a high-end professional machine.

---

(* and updating the BIOS to 1.11.2 was a nightmare. First try it got so stuck I had to pull the battery cable. Only got it to install by using the USB-C to HDMI with external 1080p monitor trick that some wonderful people outlined earlier in this thread (Thank You!))

 

Seeing as you guys have pupmped ram up high, whats the consensus on getting 2933 and 3200 ram speeds?

16 i have ATM is not enough when i start running several VM-s so i'm looking into 2x16GB sticks which i would use in place of existing 2x8GB

Is it enough just to find sticks that have JDEC rated speeds over 2666 or could XMP rated speeds also work?

 

Considering buying 7540, would appreciate advice on a few points while I'm reading 135 pages of this topic

1. Is there any luck with 3200Mhz modules on i7 9850H or only 2666Mhz work reliably? Same question as one post above essentially.
   
2. Is there any recommended "smaller" charger? I have found this topic at 51NB, but the whole construction looks a bit dodgy.
3. Is there a "summary" of known quirks for 7540? Say like a WIKI page or so.

Thanks!

 

Hi,
Just making sure, what is the consensus regarding the Precision 7740's ability to boot from 2TB SSDs? My old 1TB SSD (Samsung 970 PRO) is just about gone, and I've been considering using a 2TB as the boot drive. Specifically one of the Samsung 970 models (EVO,EVO PLUS, or PRO)
I ask because upon looking this up just incase (I remember that the 7740 manual says that 3 of the M.2 slots support 2TB sticks), I came upon a forum thread saying that a few 2TB sticks were causing BSODs to 7740s and 7750s, but 1TB sticks worked just fine.

 

I am using a 2TB ADATA SX8200P without any Problems. IIRC 4TB Sabrent was confirmed to work as well.

 

Typing this from a 2TB boot drive, thanks!

 

Zero problems. I have four Samsung 970 EVO Plus drives, each 2 TB, in mine, all four slots can be recognized in the UEFI boot sequence setup. Use GPT partition table.

 

I have been away for quite a while, and I'm back with some updates and some questions. People probably don't remember this far back, but when I first acquired my 7740 last year, I was having tons of problems with thermal stability and crashes. Thankfully, that's all resolved. Here's the TL;DR summary. I'll post the questions separately since they're on-topic for the thread but not related to my thermals.

• After trying all sorts of software updates, Dell's firmware diagnostics, conservative power settings, using Intel XTU to reduce Turbo Boost, etc., I finally decided the problem had to be hardware and not my settings.
• Dell's diagnostics were finding no problems, but HID recommended OCCT stability test. That immediately found errors in RAM and elsewhere, under load, and even with all power and clock settings at defaults.

• After some discussions with HID Evolution, I decided to invoke the onsite service contract I had purchased from Dell. This was during the early pandemic lockdown. Dell wanted me to depot the machine, but I insisted on an onsite call because it was my production computer and avoiding long downtime for depot repairs was the very reason I'd paid extra for onsite.

• Dell's tech was dispatched from Unisys and met me at my company's data center, which was safer for both of us because better HVAC and more social distancing space than my home office.

• Dell replaced the system board under warranty -- no questions, no problems. Their tech was amenable to using the Arctic Silver 5 heatsink gel that I supplied. I selected this based on recommendations from a close friend who happens to be a Dell-certified service tech working in the IT department of a major university. She's repasted hundreds of CPUs and GPUs, so if it's good enough for her, it's good enough for me. The tech had no objection to any compound that is non-conductive (which applies here, despite the word "Silver" in the name).

• The new motherboard passed the OCCT test, and we resolved the ticket -- but soon after, the problems returned almost exactly as before.

• Dell dispatched the same Unisys tech, and he called ahead to discuss the situation with me. He said it was entirely possible the new motherboard was defective -- it's not common, but it does happen -- but he also said he wanted to replace the heatsink as well this time.

• Onsite, the tech found that the stock heatsink had a small (about 1cm x 1.5 cm piece of plastic stuck to it, in a place that wasn't obvious and which he hadn't noticed before. He recognized it, though, and said it was part of the bulk shipping container that Dell gets at their assembly plant, when these heatsinks arrive from the upstream supplier. The tech said he'd seen these when he toured the plant during his training.

• New motherboard, new heatsink -- and the problems changed but were still causing crashes. This time, OCCT found that it was RAM, and by removing part of my DIMMs in different combinations, I was able to determine that the board was good but the aftermarket DIMMs were faster than this machine could reliably handle under load.

• Since the RAM came from HID Evolution, it wasn't covered under the Dell warranty -- but HID made good on their own warranty and sent me a new set of DIMMs, this time Dell stock, as an advance-shipment swap.

• Conclusion, per the Dell/Unisys service tech: That stray bit of packing material had caused the heatsink to perform very poorly, and my original motherboard, replacement motherboard, and that RAM stick were probably all cooked by the heat buildup. Replacing the heatsink and motherboard together fixed it, but the RAM compatibility was a separate issue.

After all of that, the machine was rock-solid thermally and passed every OCCT test, and it ran quite stable until one of my fans started to fail (the usual bearing noise this time). At my request, Dell shipped out a new heatsink assembly, and this time I did my own repair. (To their credit, Dell offered to dispatch their tech, and said they were still open to doing so even if they had pre-shipped the part to me.) So I did my first CPU and GPU repaste in a laptop -- it was a little nerve-wracking, but it seems to have gone well, and I have had no further stability issues. I also replaced an SSD recently, but in fairness this was one that I had migrated over from my previous computer, so I don't blame Dell or HID for that.

I give very high props to both HID Evolution and Dell for how all of this was handled. Specifically:

• As usual, HID worked with me and wasn't stingy with their time during diagnosis. They had good recommendations for testing software -- I was so impressed with the free version of OCCT that I cheerfully paid for the pro version.
• When it came to the RAM, HID didn't insist on an exact swap. They met my request for OEM parts, even though the OEM parts were by that time more expensive. This has made it easier for me to get tech support from Dell, and I'm willing to trade off some cutting-edge performance in exchange for fewer headaches keeping my work computer running.
• Dell support was terrific through the whole thing, once I got past them wanting me to depot the machine. I had two new motherboards plus a new heatsink, all no-charge under warranty, during the first incident. Then when the fan started failing earlier this year, they made absolutely no quibbles about shipping me a new one for self-installation. They offered me the choice of depot, onsite tech, or DIY, and listened to and respected my preference while confirming that this does NOT void the service contract. The only real sticking point on their service contract is that they don't support non-OEM hardware, which is entirely reasonable and understandable.

I should also mention that the full service manual for this system is online as a free download from Dell, and I highly recommend having that on hand before working inside the case.

 

I wonder if power draw could be the culprit as some do draw more than others and it came up as an issue in a review of an 8TB model on YT that caused it to not work in some machines.