Dell Precision M6700 Owners Thread

Think you might have shorted out your board somehow ... :-\
I did this by mistake when working on my NGFF card install, I think it's not too difficult if one of the traces gets scratched.

I was going to say that M6700 motherboards are cheap these days, but then I actually looked and found that they're actually hard to find on eBay and the ones that are there are rather expensive. :-\
Around two years ago I was able to pick up two for around $80.

 

Try only the motherboard, cpu, power button and see if the cpu gets hot (by hand), if it is, put a stick of ram and the display.
If nothing works your motherboard is probably fried (or the cpu), a motherboard is about 90$ on aliexpress, as far as I know there is no special version to take since every motherboard have the two soldered video output.
About a year ago they were at around 50~60$

 

Thanks for the reply, this really sucks for me atm since I'm traveling so it's limited what I can do or have access to. Ok I'll try that, and rip apart the mobo more to have a look. The CPU doesn't show any burn marks or signs of damage though. I'm also going to clean the board with alcohol and try different configurations. I didn't lose any data since this thing is so modular which is a major reason I love these old precisions tanks, my OS was only on the 125gb msata everything else is plug and play.

I was thinking that there must be some life in it since at least the little caps lock light goes on when on battery or plugged in, but there's little to no info from dell, google etc I'm definitely getting the barest bones M6700 on ebay I can find since I've got all the memory, hopefully the cpu and everything else. I was looking at a modern gaming laptop at walmart today, which they didn't have in stock predictably, and they are, well you know over 1000$. I could get another m6700 for $500 and sell the components I don't carry over to make up for it.

I'm not thinking of going the mobo replacement route, too complicated and time consuming. Hopefully another 980m could come up around 200 -250 later on but make due with the 970m up there for a while. Gaming was the major use for me with some cad here and there. My backup m4600 will do but man it sucks just got a couple of games I was really getting into, rdr2 and doomet. Blurry slideshow gaming lol. I loved that thing, it was my frankenstein that I put together, did everything I wanted this sucks man ((

 

As an update I had just the cpu and the mobo, took everything off or unplugged it, including the io board, it was kinda hard to just get the little power switch and cable that's attached to the top bezel, nothing. Took out the cpu, no indication of anything wrong anywhere. No top of the mobo burns etc I could see. Cleaned w alcohol in several places put it all back together, nothing. Not even the caps lock light on now. Although I did put the 980m back in bc it's easier to store it there really.

Dead. On the other hand the extended takedown ie including the IO board and take out the CPU kinda convinced me to try to replace the mobo, it's worth a try as the lowest usable m6700 on ebay right now is around 400 bucks. I just wish there was a way to know if it's the cpu or not, I think the burning smell did come from that area but the cpu has no indication of a burnout or damage it's cherry.

 

So there is are two issues I've had with my m6700, it's not gigantic, but it is annoying.
The gpu clock speeds will remain low unless I re-plug in the charger sometimes.
Throttlestop is needed for the cpu to actually turbo properly.

From what I recall this happens with other chargers.
The system is fully functional and has no other issues

I have 4x 1866 ddr3, i7 3840qm and an 8gb dell nvidia 980m.

I'm wondering if anyone else has faced this issue.

 

I undervolt my 980m to resolve any issues honestly.
The stock voltage table is terrible and may be something to do with why these chips run unnecesarily hot, you should actually be able to do a fairly substantial undervolt even with a badly binned card.

 

There is no way to know if the cpu is dead or not, do you have access to a multimeter/ohmeter ? we could check basic thing such as shunt resistor and fuse to see if they are working or not, might be an easy fix but I doubt it.

Never had such issues with my M6700 with it's cpu at 4.2GHz and my P4000 or even a 880M.

You don't really need to undervolt them, the mandatory thing is to use very high quality thermal pads on the mosfet (other component are not as critical), to check if the contact with the heatsink and the mosfet thermal pads is good, I remember that I needed to replace/exchange some thermal pad to get my GTX970M not dying by burned mosfet, "stock" thermal pads from kepler card are not the right size to fit Maxwell and Pascal card, iirc they are too thick on the coil so the card bend and mosfet are not in contact at all, even with a lower power GTX970M and 4 mosfet (to lighten the load on them), it burned pretty quickly.

YOU NEED TO CHECK IF YOU HAVE A GOOD THERMAL PAD CONTACT WITH THE CARD (mosfet are the most important, then it's the memory) AND HEATSINK BEFORE MOUNTING IT ON THE MOTHERBOARD, you can use thermal paste to check if there is good contact on "hidden" zone, if it spread well it mean that there is a good contact.

For me the best thermal pad is the Gelid GP-EXTREME (the ultra version is worse)

 

I've got a real basic dollar store multimeter, where would I get what the correct readings should be on the 3940xm, is there a tutorial on it? Just to see if it is in working condition would be good, if not I can get a 3740qm cpu for 70$ and just have to get a new graphics card that would be sweet at this point. Rather have the cpu go then the mobo for sure. It looked pretty mint, never taken out. Be nice before I pull the trigger on a barebones replacement whole system.

 

I had odd throttling behavior in my M6700 which would be solved by a re-plug. I ended up determining that it was the "Delta"-branded 240W power supplies that I was using. I had three of them, but when I swapped them out with a different type, this problem went away. (Note, Dell actually uses different manufacturers to produce these power supplies. I had a genuine "Dell" 240W power supply which also had the "Delta Electronics" logo on it sort of near the top of the label.)

As for throttling, I've never had to use Throttlestop. However, I do sometimes notice that it does not turbo instantly; it likes to wait until the CPU is under sustained load for several seconds before deciding to clock up above around 2.8 GHz. This seems to be better if you set the thermal mode to "Ultra Performance" instead of "Standard". (To set the thermal mode, you need the Dell Feature Enhancement Pack installed, and then it is in Windows advanced power options. It gets whacked every time there is a major Windows 10 upgrade and you have to uninstall/reinstall DFEP.)

 

As an update I tore down, took out the 980m again and started up, the power adapter light stays on now but the power button etc lights don't come on. The GPU fan runs for a few seconds then stops then sometimes get a buzzing/whine from what I think is the area where the gpu was, all in a loop. I put the gpu heatsink back in and covered all the surfaces where their was gpu contact, with electrical tape. I don't have another gpu to put in there. CPU fan doesn't come on. Sometimes the caps lock light comes on though.