Jump to content

14" Dell Latitude E6430 - Performance Upgrades and System Mods


Recommended Posts

@timohour, @TheClassicalCat,

I've applied high signal on pin 5, then I wrote a modified flash with read access and write access set to 0xFF.

I restarted and now I can read the flash without applying high signal.

But I can't write again! It says Error 280: Failed to disable write protect for the BIOS space!

I can still read, but can't write even with applying high signal on pin 5.

Any idea what to do?

EDIT:

ok I can write just the DESCRIPTOR which I modified in hexedit

working! I've got a 2x port 1.

Thanks.

  • Thumbs Up 1
Link to comment
Share on other sites

@timohour, @TheClassicalCat,

I've applied high signal on pin 5, then I wrote a modified flash with read access and write access set to 0xFF.

I restarted and now I can read the flash without applying high signal.

But I can't write again! It says Error 280: Failed to disable write protect for the BIOS space!

I can still read, but can't write even with applying high signal on pin 5.

Any idea what to do?

EDIT:

ok I can write just the DESCRIPTOR which I modified in hexedit

working! I've got a 2x port 1.

Thanks.

Great Work there! It seems like the iGPU model have the 92HD90B2X5 chip on the other side of the motherboard.

Added it here for future reference.

ljmjDEAm.jpg

Is it possible to make the pinmod without taking apart the laptop?

Thanks for your insight

Link to comment
Share on other sites

Great Work there! It seems like the iGPU model have the 92HD90B2X5 chip on the other side of the motherboard.

Added it here for future reference.

ljmjDEAm.jpg

Is it possible to make the pinmod without taking apart the laptop?

Thanks for your insight

Probably from another 3.3v source, Its very difficult to short 5 and 9 because the chassis is in the way. But pin 5 has a unattached pad about a cm near it away from the chasis but is easier, so you can supply 3.3v there. I just used a 3.3v from the big resistor on the other side.

I wrecked the SATA/PCI-E Module bay to solder a SATA PORT (from the disk drive) to the ODD's PCI-E lane,

but I can't get 2x (port 3) enabled to test it, Maybe I need to supply "MOD_SATA_PCIE#_DET" or "MOD_MD" something?

Edit: Guess not had to shutdown not restart the laptop :P moving on to soldering a riser.

Don't know if it were mentioned but the module bay is a PCI-E/SATA combined port, its a normal SATA connection on one side and on the same connector above it is PCI-E.

If we can get a proper module bay connector (SATA like L shape connector but with 11 smaller pins above) we wouldn't have to tear it.

Link to comment
Share on other sites

Probably from another 3.3v source, Its very difficult to short 5 and 9 because the chassis is in the way. But pin 5 has a unattached pad about a cm near it away from the chasis but is easier, so you can supply 3.3v there. I just used a 3.3v from the big resistor on the other side.

I wrecked the SATA/PCI-E Module bay to solder a SATA PORT (from the disk drive) to the ODD's PCI-E lane,

but I can't get 2x (port 3) enabled to test it, Maybe I need to supply "MOD_SATA_PCIE#_DET" or "MOD_MD" something?

Edit: Guess not had to shutdown not restart the laptop :P moving on to soldering a riser.

If you did flash your descriptor as I mentioned it will be permanently writable. You have to do it though after you apply the pinmod and before you reboot.

Don't know if it were mentioned but the module bay is a PCI-E/SATA combined port, its a normal SATA connection on one side and on the same connector above it is PCI-E.

If we can get a proper module bay connector (SATA like L shape connector but with 11 smaller pins above) we wouldn't have to tear it.

You didn't have to wreck your ODD PORT. As I noted on this post a mini pcie card fits in there (as long as thickness and pin distance is concerned) and you have to cut some of the PM3 to make it fit into the ODD slot and thinsulate those pins that are not used! Anyway, I hope that it works that way too. Waiting for your results!!

If you are interested since you unlocked your descriptor you can know unlock your ME firmware so it will allow BCLK OC. Check this thread out. Unlocking your ME firmware will allow you to OC your CPU up to 5%. I OCed an my i7-3630QM from 3.4GHz to 3.56GHz by setting BCLK @ 104,9. I will be posting more about this mod too.

Link to comment
Share on other sites

I can't get module bay PCI-E to work with express card using PE4L + RISER and show in GPU-Z.

Using the two GND lines and RX lane capacitors as "polarity" I think I identified which is which (GND, RX-, RX+, then two GNDs, TX-, TX+, GND, REFCLK-don'tcare, REFCLK-don'tcare, GND) from the other side unless they go in the connector not in this order, the schematic just shows it as a long port, so maybe the +, - are switched.

Then I soldered a SATA port I took from the disk drive (it has a SATA to flat connector inside) to the module bay, and using a sata cable to connect to the riser (RX+ to lane2 RX+, RX- to lane2 RX-2, TX to TX... and one GND which I soldered to one of the other lanes since they are all tied.

Soldering seems ok with multimeter and impedance also seems fine.

grounding MOD_SATA_PCIE#_DET pullup looks like just enables/disable USB3.0 somewhere, how is that related to SATA/PCI-E.

Maybe someone should try 1x from module bay first.

Link to comment
Share on other sites

I can't get module bay PCI-E to work with express card using PE4L + RISER and show in GPU-Z.

..

Maybe someone should try 1x from module bay first.

If you have a x2 capable PE4H or PE4C then wouldn't x2 1.0 would be easier using port3-EC+port4-module_bay. Your EC port would provide sync clock signals and need to wire only the 4 TX/RX wires + GND shields from port4. If were just testing port4 as a x1 link then need to wire through the CLK signals as well.

Link to comment
Share on other sites

I can't get module bay PCI-E to work with express card using PE4L + RISER and show in GPU-Z.

Using the two GND lines and RX lane capacitors as "polarity" I think I identified which is which (GND, RX-, RX+, then two GNDs, TX-, TX+, GND, REFCLK-don'tcare, REFCLK-don'tcare, GND) from the other side unless they go in the connector not in this order, the schematic just shows it as a long port, so maybe the +, - are switched.

Then I soldered a SATA port I took from the disk drive (it has a SATA to flat connector inside) to the module bay, and using a sata cable to connect to the riser (RX+ to lane2 RX+, RX- to lane2 RX-2, TX to TX... and one GND which I soldered to one of the other lanes since they are all tied.

Soldering seems ok with multimeter and impedance also seems fine.

Maybe someone should try 1x from module bay first.

It is not that easy to test x1 on the module bay. You have to solder at least REFCLKs WAKE# and PERST# along with RXs TXs.

I assume from what you write that you connected pins 18-23 from port JSATA2 on schematic. If I am right on this one, is there any possibility that you have connected the wrong pins or in the wrong sequence? From my observation I think that those pins are near with those of the SATA power, correct? Maybe they are the other way around?

Is it possible that youcan get a picture of the area so we can have an idea how it is beneath the sata connector?

Thanks

grounding MOD_SATA_PCIE#_DET pullup looks like just enables/disable USB3.0 somewhere, how is that related to SATA/PCI-E.

I don't think that we need any of the two MOD_ * They both seem to be related with the USB 3.0 function of the E-module II.

Link to comment
Share on other sites

If you have a x2 capable PE4H or PE4C then wouldn't x2 1.0 would be easier using port3-EC+port4-module_bay. Your EC port would provide sync clock signals and need to wire only the 4 TX/RX wires + GND shields from port4. If were just testing port4 as a x1 link then need to wire through the CLK signals as well.

That's what I did, port 3-EC to PE4L and a RISER to the GPU, and soldered another lane of the riser to port4 module bay using a SATA cable (GND, RX, TX). [[bTW there are also PCI-E clock signals on the power part, again upside down of the normal power plug]]

I don't think that we need any of the two MOD_ * They both seem to be related with the USB 3.0 function of the E-module II.

Yes, but isn't that port combo a PCI-E + SATA, where does the USB come in.

Its possible I've got the pins wrong especially the pairs themselves but I was using a multimeter, the fact the RX lines has capacitors on them, and double GND between TX and RX which isn't between TX and CLK (and pin count).

I'll get a picture but I destroyed the port for easy access. Might not been the best soldering but its just a test.

And just to check, PCI-E connection is RX to RX (+ to +, - to -) and TX to TX right? not cross?

Link to comment
Share on other sites

post-19480-14495000187946_thumb.jpg

rightmost two pins of the SATA(?) power are GND, assuming pins 12, and 13. adding that pin 31 is connected to a 10k resistor (R1183), assuming the one in the picture which I measured to be 9.8K, and the resistor on its left also measured 9.9K going to MOD_MD pin 11 near the two GND I mentioned of SATA(?) power. So I'm guessing the count goes from upper part left to right 1 to 13 and then to lower part left to right 14 to 31.

So I assume I got the pins right.

Here's one of the connectors: post-19480-14495000188219_thumb.jpg

as you can see its has pins on top and pins on the bottom, its divided (top/bottom) like in the groups in the schematic JSATA2 with the same pin count. The right plug was a SATA L shaped female I destroyed. It's top was a normal SATA, it just has extra 11 PCI-E pins on the bottom that touch the plastic (don't connect) if its a normal SATA male connector since its they are one sided pins.

Why is there even PCI-E on this port? I can't find any dell equipment that uses the module bay for PCI-E.

Maybe a disassembled displayport/HDMI connector (without the casing) might have the correct spacing and fit (might add some buffer as it seems a bit thinner), giving us a high speed shielded cable right to the port?

Display port

post-19480-14495000191268_thumb.jpg

HDMI:

post-19480-14495000190528_thumb.jpg

Link to comment
Share on other sites

If you are interested since you unlocked your descriptor you can know unlock your ME firmware so it will allow BCLK OC. Check this thread out. Unlocking your ME firmware will allow you to OC your CPU up to 5%. I OCed an my i7-3630QM from 3.4GHz to 3.56GHz by setting BCLK @ 104,9. I will be posting more about this mod too.

Can you post the ME firmware, his guide isn't really clear.

Link to comment
Share on other sites

Yes, but isn't that port combo a PCI-E + SATA, where does the USB come in. Why is there even PCI-E on this port? I can't find any dell equipment that uses the module bay for PCI-E.
Dell USB 3.0 E-Modular Bay. Schematics show that port 4 goes to the modular bay. Since it is a shared product for Sandy Bridge (Latitude E6x20 laptops) and Ivy Bridge (Latitude E6x30) I assumed that the Modular Bay uses some kind of PCIe-->USB 3.0 card (There are NEC USB 3.0 drivers on the E6420 drivers page, Sandy Bridge did not offer native USB 3.0). Here is a shop that guarantees that it works with E6x20 and E6x30, which means that pcie port 4 is probably wired.
It is possible I've got the pins wrong especially the pairs themselves but I was using a multimeter, the fact the RX lines has capacitors on them, and double GND between TX and RX which isn't between TX and CLK (and pin count). I'll get a picture but I destroyed the port for easy access. Might not been the best soldering but its just a test. And just to check, PCI-E connection is RX to RX (+ to +, - to -) and TX to TX right? not cross?
Correct.
[ATTACH=CONFIG]15063[/ATTACH] rightmost two pins of the SATA(?) power are GND, assuming pins 12, and 13. adding that pin 31 is connected to a 10k resistor (R1183), assuming the one in the picture which I measured to be 9.8K, and the resistor on its left also measured 9.9K going to MOD_MD pin 11 near the two GND I mentioned of SATA(?) power. So I'm guessing the count goes from upper part left to right 1 to 13 and then to lower part left to right 14 to 31. So I assume I got the pins right.
Seems like it. Everything you say is as I see it on the picture. I will test on my E6430 tomorrow and get back. Just make sure that on your riser RX+- goes to pcie lane 1 RX+- (pins A21-22), TX+- goes to pcie lane1 TX+-(pins B19-20) as on the picture in the spoiler. I connected them the other way the first time I tried. If you tried and it still doesn't work, there is maybe something changed since the revision of the schematics we have.

1668D21E4A793AA1229E75

Have you confirmed that port3 is set @ x2 mode? If you own Setup 1.x or any other program that can check, make sure that port 1 and port 3 are set @ x2 mode (port 2 WiFi would be disabled).

Link to comment
Share on other sites

Dell USB 3.0 E-Modular Bay. Schematics show that port 4 goes to the modular bay. Since it is a shared product for Sandy Bridge (Latitude E6x20 laptops) and Ivy Bridge (Latitude E6x30) I assumed that the Modular Bay uses some kind of PCIe-->USB 3.0 card (There are NEC USB 3.0 drivers on the E6420 drivers page, Sandy Bridge did not offer native USB 3.0). Here is a shop that guarantees that it works with E6x20 and E6x30, which means that pcie port 4 is probably wired. Correct. Seems like it. Everything you say is as I see it on the picture. I will test on my E6430 tomorrow and get back. Just make sure that on your riser RX+- goes to pcie lane 1 RX+- (pins A21-22), TX+- goes to pcie lane1 TX+-(pins B19-20) as on the picture in the spoiler. I connected them the other way the first time I tried. If you tried and it still doesn't work, there is maybe something changed since the revision of the schematics we have.

1668D21E4A793AA1229E75

Have you confirmed that port3 is set @ x2 mode? If you own Setup 1.x or any other program that can check, make sure that port 1 and port 3 are set @ x2 mode (port 2 WiFi would be disabled).

Yes that's how I connected it, and it's the same riser image I used. It seems the orientation for this schematic is as if looking ontop of a desktop motherboard female slot (the risers has VCC 12v (B1-B3, A2-A3) tied together so I assume its correct).

I have setup 1.x which confirmed its x2 on port 3 (I did it x2 on 1-4, 5-8).

maybe flashing back the restricted descriptor since it has strange effects on audio and stuff.

Link to comment
Share on other sites

maybe flashing back the restricted descriptor since it has strange effects on audio and stuff.

My laptop works fine after descriptor unlock. No problems at all.

I will be testing later today with my E6430's port 4...

Link to comment
Share on other sites

There was a misunderstanding... I meant that my laptop works fine after the descriptor unlock. Not about enabling port 4.

heh, I meant how are you gonna test port 4. soldering? which bplus/GDC adapter? I guess will see anyway.

Link to comment
Share on other sites

heh, I meant how are you gonna test port 4. soldering? which bplus/GDC adapter? I guess will see anyway.

It seems like I am the who misunderstood :tranquillity:

I am going to cut my minipcie to sata adapter where I show below... it seems like it aligns perfectly.

EDIT: ok, I don't know how I managed to do that but this is a flipped image of the original. Will fix it soon...

zY5mcOx.png

Only problem is I couldn't make the PE3A to work and I can't connect my GPU to my EC slot :(

Link to comment
Share on other sites

Hello

I am really wanting to do this with my dell notebook the E6420. I have kingston 2133k ram so I believe this should work. Is there any particular difference between thaiphoon burner and SPDTool? The link for SPDTool's download is 404, so is there another alternative free version? Here is an image of my ram around at the bottom of page. http://forum.techinferno.com/general-notebook-discussions/9950-getting-new-sodimms-kingston-hyperx-2133-however-3.html#

Link to comment
Share on other sites

Hello

I am really wanting to do this with my dell notebook the E6420. I have kingston 2133k ram so I believe this should work. Is there any particular difference between thaiphoon burner and SPDTool? The link for SPDTool's download is 404, so is there another alternative free version? Here is an image of my ram around at the bottom of page. http://forum.techinferno.com/general-notebook-discussions/9950-getting-new-sodimms-kingston-hyperx-2133-however-3.html#

I can upload spdtool for you but it won't work for newer laptops. It will only work for GM45/PM45 and older laptops. I used an older laptop from work to make the flash. Thaiphoon Burner free version will only let you read memory SPD, not write. You need to purchase in order to enable writing. But according to @Khenglish here your RAM will work @ 2133 with an older bios (on his E6520 which is almost the same as the E6420). Have you thought to try downgrade your bios to an earlier revision? It will probably save you a lot of time.

Adding XMP profiles to your RAM (using SPDtool or ThaiphoonBurner) will only help if you can enable XMP on your E6420 (you probably can). Check here how to use the UEFI variables and extract your IFR. If you do, then you can try and set your RAM to run @ 1866 using UEFI Variable. If it doesn't work, you can add XMP profiles to your RAM and use UEFI variables to enable them. If you need any help feel free to ask.

EDIT:Updated my results with dual channel. An amazing ~20% improve from 1333-->1600 and ~ 15% from 1600-->1866. More

...

  • Thumbs Up 1
Link to comment
Share on other sites

I can upload spdtool for you but it won't work for newer laptops. It will only work for GM45/PM45 and older laptops. I used an older laptop from work to make the flash. Thaiphoon Burner free version will only let you read memory SPD, not write. You need to purchase in order to enable writing. But according to @Khenglish here your RAM will work @ 2133 with an older bios (on his E6520 which is almost the same as the E6420). Have you thought to try downgrade your bios to an earlier revision? It will probably save you a lot of time.

Adding XMP profiles to your RAM (using SPDtool or ThaiphoonBurner) will only help if you can enable XMP on your E6420 (you probably can). Check here how to use the UEFI variables and extract your IFR. If you do, then you can try and set your RAM to run @1866 using UEFI Variable. If it doesn't work, you can add XMP profiles to your RAM and use UEFI variables to enable them. If you need any help feel free to ask.

EDIT:Updated my results with dual channel. An amazing ~20% improve from 1333-->1600 and ~ 15% from 1600-->1866. More

...

So will STPTool work just enough to enable xmp? You said it doesnt work on older laptops, but maybe it can just do that? Alright, I will try the older bios thing first before I get to thaiphoon burner. Do I just keep downgrading till one of them works at 1866 or 2133? Im really wanting to do 2400 at something like 11-12-13-31 and I can definitively do that with my new kingston ram that apparently can even do 2666mhz, though I dont think I will do more than 2400 or 2133 for stability reasons. The link you gave seems to orientate the UEFI variables for the purpose of increasing cpu multipliers and not ram, or was there something I missed?

Link to comment
Share on other sites

Update: Well I went through all the bios's and found that A17 was the cut off point to which 1867 would no long be supported and the next one, A19 & A21(the last one) would force it at 1600. Interestingly none of them actually did 2133, but I assume in the path we are going that that will change.

A17(current one on) Dell Latitude E6420 ATG System BIOS Driver Details | Dell US

A19 www.dell.com/support/home/us/en/19/Drivers/DriversDetails?driverId=V99MK

There isnt an A18 for whatever reason. I want to say that its the "1.5.33 version of the Intel PXE OROM." according to the bios "enhancement" description.

Link to comment
Share on other sites

So will STPTool work just enough to enable xmp? You said it doesnt work on older laptops, but maybe it can just do that? Alright, I will try the older bios thing first before I get to thaiphoon burner. Do I just keep downgrading till one of them works at 1866 or 2133?

@Khenglish wrote here that earlier bios's would allow up to 2133 on the E6520. You should ask him which version was working. You can also set 1866 on your current BIOS version.

Khenglish mentioned a modification to uefi parameters to get the ram running 1866Mhz I think that seems like the only method. They will run 1866 and 2133 without buying Thaiphoon burner if you can get the bios to set one of those like Khenglish mentions in the other thread.
Im really wanting to do 2400 at something like 11-12-13-31 and I can definitively do that with my new kingston ram that apparently can even do 2666mhz, though I dont think I will do more than 2400 or 2133 for stability reasons. The link you gave seems to orientate the UEFI variables for the purpose of increasing cpu multipliers and not ram, or was there something I missed?

It gives you the idea of how to use UEFI variables. If you understood this then you have to extract your IFR find which variable is responsible for XMP and use UEFI variables to enable XMP Profiles.

SPDTool and Thaiphoon Burner are both software which change your RAM SPD. You can use them (spdtool on older laptops) to add XMP Profiles or change the JEDEC SPD values on your RAM. They have nothing to do with enabling/disabling XMP on your laptop.

<hr>

EDIΤ :

Extracted your BIOS IFR here E6420-A17_IFR.txt. Seems like your E6420 has an option for XMP profiles

Setting: DIMM profile, Variable: 0xC4[1] {05 09 C4 00 01 2F 01 30 01}0x4877F
Option: Default DIMM profile, Value: 0x0 {09 09 31 01 00 00 13 00 00}0x48788
Option: XMP profile 1, Value: 0x1 {09 09 32 01 01 00 10 00 00}0x48791
Option: XMP profile 2, Value: 0x2 {09 09 33 01 02 00 10 00 00}0x4879A
Option: Custom profile, Value: 0x3 {09 09 34 01 03 00 10 00 00}0x487A3
End of Options {10 02}

I don't know if it works though. It should.

  • Thumbs Up 1
Link to comment
Share on other sites

5ZkafiG.png

Lets enable overclocking on all 6 and 7 series laptops is a thread started by Khenglish where you can learn the basics on how to do ~5% BCLK overclock using XTU. This will work on any i5/i7 CPU dual core or quad.

To do this BCLK OC you have to unlock your flash descriptor, use Flash Programming Tool (fpt.exe) to take a dump of your ME Firmware, use Flash Image Tool to modify your ME Firmware and then use again fpt.exe to flash your updated ME Firmware dump. After this you would be able to OC using XTU.

We have discussed here how to unlock your flash descriptor.

 

Flashing modified ME FW

DISCLAIMER: There are reports that people had problem upgrading/downgrading their BIOS after applying  the ME FW OC mod. Make sure that you keep a backup of your original bios dump in order to revert your ME FW to the original. It is also recomended that you mod your own ME FW instead of using the supplied one here. If  you don't know how to do it ask for help.

 

E6430UCME

1. Download Series 7 fpt.exe from here. Although there is a Windows version flashing with it can be harmful so you should better run it from DOS.

2. Backup the original BIOS dump

 

 
fpt.exe -d filename.bin

 

3.Flash the modified me fw with fpt

 

 
fpt -me -f E6430UCME.bin

 

4.Shutdown (not restart) and then install XTU or reinstall if you have already installed.

p0un5esm.png HtLsz8wt.png

 

EDIT: Updated results with my i7-3720QM running @ 3.77GHz

UeDq4djm.pngrOFs5Mst.png

Edited by timohour
  • Thumbs Up 3
Link to comment
Share on other sites

For those of you who do try timohour's ME FW, make sure you back yours up first in case it does not work properly. For HP systems there is something laptop specific in the MW FW that prevents flashing someone else's ME FW. Dell likely doesn't have this stupidity, but you should back up just in case.

Also beware that updating your integrated graphic driver will kill BCLK overclocking. After updating your driver you must then reinstall XTU to get it back.

When doing BCLK overclocking you should enable HPET in command line, or else you will see stuttering in games. If you have Win8 or 8.1 then you can't overclock because there is no way for Win8 to track time properly if BCLK is changed after startup. I don't know about Win10.

In an administrative command line enter:

bcdedit /set useplatformclock true

Then reboot.

  • Thumbs Up 4
Link to comment
Share on other sites

Oh wow, thank you so much. I just have one question, how do I actually implement this? Im not sure how you even extracted this because the site for the donovan universal extractor is doing maintenance for however long. donovan6000's Blog: Universal IFR Extractor

Also the link to the UEFITool 0.18.8l tool is broken.

Since this is for the A17 bios, is that best since its the latest one doing 1866? Again im not sure how to implement this.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.