RTM875t-606( Info gathering and OCing)

For a better Understanding of How and Why Check Moral_Hazard's PLL Mod Thread

He has already explained what is done below

Thanks goto Moral_Hazard, Nando4, and Darnok44, not to me!

Update

This can now be done!

Step 1. Download SetFSB->RUN it->CLICK Diagnosis->Chose PLL diagnosis in the Clock Generator field->Click GetFSB

Check Byte 09h bit 6

If bit 6 = 1 , You'll need to do a hardware Mod to get SetFSB to work.
If bit 6 = 0 , You can Skip the Hardware Mod.

Check Byte 07h

If Byte07h = 1E, you have the RTM875t-606
If Byte07h = <>1E, you DON'T have the RTM875t-606( And your reading the wronge Thread!)

STOP!!!!!!!!!!!!!!!!!!!!!!!

Before you go about taking your Notebook apart, you need to check if the Hardware MOD will Work for you............................

Click 00h set bit 2 to 1->Click Update->Click Apply

If your system keeps Running like it does not care, then it's Safe to do the Hardware Mod.

If your system FAILS.............. STOP, you need to try and setup the PIN17/18 Pin Config (I'll add a table for this later if we have anyone that needs it, post a reply to this topic.)

Step 2

The Hardware Mod:

You need to take apart your NoteBook, I won't Help you!



I won't care if you KILL your NoteBook, Burn your House Down, or knock up your Dog, with anything you read here. It's not on me, it's yours, you did it, deal with it.

Here are some Pic of How Small these resisters are..............

If you have the TSSOP package you need to Ground PIN 4.

How you do this is up to you, that's it, your done, go ahead and put your Notebook back togather.

Step 4. Run SetFSB->Click Diagnosis->Chose RTM875t-587->Click GetFSB

Click 00h change bit 2 to 1->Hit Update->Hit Apply

Move the Top slider to the FSB/RAM timing you want to test->Hit SetFSB







We don't have a data sheet for this Clock Generator, and it seems to be in about half of the ICH8M Notebooks that use a Quanta Logic Board.

As that's at least 2 million Notebooks made in the last few years, it maybe usefully to find all the info we can.

I'm going to lay out the RTM875t-606 Registers and PINOUT as best I can, and hopefully with help from others we can gather the needed info to have it added to SetFSB.

It seems the PINOUT and Registers are near if not the Same as the ICS9LPRS365, but there maybe a few differences.

PINOUT: I'll list the Pins that I can confirm and any that maybe different.

4. PCI2/TME (I'm not sure this is the TME pin, it seem to be Pulled Low(Logic 0) on my LoBo. However Bit 6 of Byte09h is 1, I grounded this pin and removed the resister, and TME is still Pulled High, but I could have made a mistake.)

10. USB_48MHz/FSLA (Confirmed)
11. GND48 (Confirmed)
57. FSLB/TEST_MODE(FSLB is Confirmed, I don't know if RMT875t-606 has a Test Mode, I assume it does)
58. GNDREF(Confirmed)

That's all I've tested in Hardware so far I'll add to it as I can, heck the ICS9LPRS365 Datasheet for a full pinout.

Registers: At this time it seems they are the same as ICS9LPRS365, I can Confirm that Byte 00h-09h line up, However A0 and up maybe differant.

Byte 07h seems to be Read Only ( 1F Hex 00011110 Bin. )

Can anyone confirm that 1110 is Reaktek's Vendor ID?

OCing:

I can confirm that 133/167/200mhz FSB works on my ICH8m, I did try 266, but it did not POST, my CPU is T5550 so 167FSB, I'd like to try with a 800mhz CPU and 1066 too.

I'd put money on 1066 CPU's working with my ICH8M chipset, I'd have to Patch my Bios for the CPUID.

Another nice thing is: On my HP DV6871us the RTM is POLLED in Hardware on resume from sleep, so I should be able to add a set of Dip Switches, to the out side of my NoteBook.

How is this usefull?

I should be able to go from 100/133/167/200 FSB on Resume from Sleep. My CPU Supports SpeedStep from 6x to 11x with Voltage from 0.950v-1.162v, so on the Low Side 600Mhz, should save me Lots of Batt. when I don't need much CPU/Bus Bandwidth, However 2,200Mhz on the High side. NICE!

I still need to check the 100mhz setting and see if my System as a whole can run that Slow without RTC or other issues.

Also I'd like to edit the PStates in my DSDT and see if I can Drop the VID from 0.950v as that's far more than is needed @600-1200Mhz on my CPU.

 

I've replied here as half the answer is above, and I'd like to get your take on PCI2/TME, as I can't seem to pull it LOW?

I've gone half the FSL pin mod, that is, Pull FLS x LOW or High at POST, this works for me, for now.

As I've outlined above, I'd like to check 100mhz FSB and do a Dip Switch setup, atleast until we can get SetFSB working with RTM875t-606.

My system has T5550 167mhz FSB CPU, so my CPU Frequency Select Table reads FLSc 0 FLSb 1 FLSa 1, I pull FLSa(Pin 10) low at POST by shorting it to PIN 11 with the head of a needle, then remove it after about 2 sec.

I also pulled FLSb( Pin 57 ) Low with Pin 58, the same way for 133mhz FSB.

I got lucky and my Ram is running good @ 800mhz, it has SPD for 533 and 667, but not 800.

I'm still testing, but I'll add a howto with Photo's and links to the first post when I'm done.

ORTHOS still going....................................

 

I have seen some PLLs that use a different register for the TME bit.
So it could be possible you're looking at the wrong register in setfsb.

Maybe TME mode is disabled for you, but setfsb is trying to reprogram the wrong bits.

Have you tried using every PLL in setfsb that starts with RTM?

What exactly happens when you use setfsb? can you describe it, does the system lock up?

EDIT: Also you can try to use setfsb to change the TME bit. Change it to 0. If the bit is not RO then it should work. But it's likely the TME bit is RO.

 

I think RTM's is also B09h bit 6.

I Lost the DataSheet for the Aspire1 A1x0, I think it may have showed an ALT Pin for TME, Does anyone still have it????

Yes, I must chose RTM, otherwise Error!

SetFSB seem to Read bits 0-3 of B07h Vendor ID, and return Err if other than selected.

I've looked at the Source Code however my Japanese....................

A mix of the Code for the ICS9LPRS365,and RTM code with a dash of what ever the TPM work around was for the 2510P.

I can change my SRC_Main_SEL B00h bit 2 to PLL3 per your post, so maybe we can find a way to help other not have to PIN MOD.

It does change some Bits, but with TME ON, no effect one the FSB.

However I can change other things, like disable spread modulation......

Note the Clock speed change, so we know that the Registers are correct.

I'm 99% sure the ICS9LPRS365 and RTM875t-606 share the same Registers, I've done many tests B00h-B009h and can confirm most of them.

I gave a shot at trying the Work around you outlined for the 2510P, but it gets a bit over my head in the datasheet as what to change in PLL3 mode @0Eh and 11h, maybe Nando4 can give it a look and dumb it down a bit for me to test.

That Bit seems to be RO so I'll have to try and pull pin 4 down once more, have a look at the TRACE......

I polled that resister in the 2nd row up, where it seem to go, and it showed ground, I PULL it DOWN hard at boot with a copper wire to the DC ground of the PSU, I also Removed that resister, none of that hand any effect.

However, my MM is junk, and I really don't know how to use one that well.

Even with that resister removed, PIN4 still shows ground, however B09h bit 6 is still 1?

Once we get this TME trouble out of the way, maybe we can send a nice Email to Abo with our finding all cleaned up and ready for him to work his magic!

 

Do you still have this pdf? I would like to take another look but the link doesn't work.

 

Resource not found

No, it seems I've lost it too, maybe someone else saved it.

I tried the WayBack, but no luck.

 

So do you think TME is pin 4?


EDIT: Can you use setfsb to try to disable PCI2. Like I did in my thread?

 

I dont know what byte it is for your PLL, but are you sure the M/N Enables are enabled.

For me they are bit 0 and bit 1 of byte 21.

Consider changing those bits to 1, then try setfsb again.

 

Disable PLL2? Won't that crash my SATA?

I did change SRC_Main From PLL1 to PLL3?

I'll give Byte21 a shot in the morning, off to bed now, do you know the Logic of it?

 

change byte 21 from 00 to 03.

If you dont want your SATA to crash, change the source of it. Probably bit 1 of byte 0. Set it to 0 to make the source for the SATA clock = SRC_Main.

 

ok this is great news!
i am not exactly sure how to do the pin mod yet
i dont quite understand "If you do find a resistor connected to the pin, then all you have to do is remove the resistor. Then connect the pin to GND with another resistor (or use the same one)." is this what Darth_nVader did?
maybe i will just wait till you get this figured out and post the how 2 or guide eh?
i sure am looking forward to it

i am building a new case for my laptop for super cooling (i never go anywhere with it) i am going to put 4 fans that i found in some old desktop cases
also going to try and mod the heat sink by somehow attaching a small desktop heatsink i will post pictures when i am doen lol

 

I know that the MLF packaged clock generator has the same pinout as ICS9LPRS365, maybe the TSSOP is different and TME is a diffent pin.

 

Summary of what was done to overclock the RTM875t-605 PLL

My intepretation based on this thread and the comments in the PLL mod, 200mhz to 266mhz help please thread.

He did method 2 in that thread which is a pinmod to set the PLL into the desired frequency of operation. This pinmod involves identifying what is connected to the desired FSLc, FSLb or FSLa pin, then cutting the track or lifting a resistor to disconnect the pin from the rest of the circuit and running a patch wire to GND (0) or 3.3V (1) for the desired FSlc/FSLb/FSLa settings. As no datasheet can be found to identify what pins are FSLx , he's figured out that the FSLx pins of the Realtek PLL match that of an ICS PLL based on an Aspire One PLL document that we don't have access to anymore.

His 166->200Mhz OC works. 200->266Mhz OC doesn't. To go higher than 200Mhz would require either setfsb software overclocking that provides 1Mhz increments to find the point of stability OR try the 266Mhz OC again but with modified RAM timings and/or increased CPU voltages.

setfsb requires that the registers can be written and recognised by the PLL. If it's manufacturer pin-set to TME mode (trusted mode - no overclocking), then the PLL ignores the register settings, sometimes even freezing the system. Finding what the TME pin would mean a pinmod could disable TME mode and *hopefully* setfsb can be then made to program the PLL.

 

The datasheet is still available, for example here:
http://vassyl.net/work/acerone/aspire0nemb.pdf

One thing to keep in mind is that ZG5 uses the MLF version, while the OP has a TSSOP on his MB.

 

Wow - what a document. RTM875t-606 shown uses pin 11 to set TME mode in the 64-pin package. So just need to GND pin 11 to allow setfsb overclocking. They are suggesting that ICS9LPRS365 is pin compatible. Though while they may use the same FSlx pins to set FSB, they may use different internally registers set the FSB via software. setfsb supports RTM875T-587 and RTM876-660, a mix of numbers that looks similar to RTM875t-606 so worth trying those, as well as ICS9LPRS365 with setfsb.

 

@nando
But don't forget, it's a square 64 pin MLF package, not the rectangular TSSOP. The pinouts are different.

 

Actually I think nando is right, when you look at the picture of the PLL you see it's a rectangle and has pin 11 as TME:

 

@moral hazard
That's just a schematic, it doesn't show the actual shape and the pins are mixed to match the schematic without making it too difficult to read. I developed the ZG5 overclocking instructions so I should know I hacked the hell out of that poor PLL and my ZG5's PCB surrounding the PLL looks pretty burned now, but it's working like a charm

 

internal PD Anyone know what that means?
I think they are saying, Internally Pull DOWN. That explains why PIN 4 shows Ground, for me.

That's an Odd one, forgive me, is Pulled Down the same as Pulled LOW?




I assure you PIN 11 is GND48, in the TSSOP Package(RTM875t-606).

 

Please explain how to do this, I seem to be missing it?

 

Yes, I think your right about internal PD.

Now that I think about it, I don't think it will get us anywhere, so forget I said anything about that

 

Nando4, forgive me if I seem to argue, it's not my intent, I'm a bit confused.

Did you look at the PINOUT for the ICS9LPRS365(rectangular) TSSOP Package from the Data Sheet, it shows the PIN out for Both TSSOP and MLF Packages.

On the MLF package, yes PIN 11 is PLL2/TME.

I need to check Pin 4 with a MM, after POST and see if it's 3.3v?

 

what does pin 11 do in the TSSOP package?

 

That's got me stumped, I believe that this was ADDED to "Secure The Chain of Trust", somehow I'm missing how to properly set TME to Logic 0

 

PIN 11 = Ground pin for the 48MHz outputs.

Remember my FSB is 0 1 1 or 166Mhz, when I SHORT PIN 10(USB_48MHz/FSLA)

Code:

Fixed 48MHz USB clock output. 3.3V./ 3.3V tolerant input for CPU frequency selection. Refer to input electrical characteristics for Vil_FS and Vih_FS values.

, to PIN 11 at POST, I get 0 1 0 200Mhz FSB.

I assure you PIN 11 is GROUNDED

 

Apologies. You are quite correct. pin4 is TME on a TSSOP as shown by the example below, courtesy of Moral_Hazard:

PLL pins of interest: blue: method 1 | red: method 2/3. From ICS9LPR501SGLF PLL datasheet

 

Ok, I just did not do a Good Job of getting PIN 4 to Ground.

I can now Confirm that Pin 4 is TME, and B09h Bit 6

So as far as TME, everthting seems the same as what Moral_Hazard laied out in the PLL MOD Thread

So if you have RMT875t-606 64pin TSSOP, you should check Byte 09h bit 6, if it's set to 1, you have TME = No Software OCing.

Next check to see if you can select PLL3 as you SRC_MAIN, set Bit 2 of Byte00h to 1, if your system DOES NOT LOCK UP, Abo maybe able to make SetFSB work.

 

Another update, if I select RTM885T-926B in SetFSB, it CAN change my FSB

That's the Good News, the Bad News is, if I set it to anything other than my FSB Speed @ POST, the system only runs for about 3min., then Hard Lock or BSOD

Now when I select RTM885T-926B, and hit SetFSB, it Does NOT read my FSB Correct, it reads it as 133Mhz FSB 266Mhz Ram, it does read the PCI-E/PCI Buses Correct @ 100/33.3Mhz.

If I move the Slider to my REAL FSB, and Hit SetFSB, the System Works just fine, However if I go any Lower or Higher, system chrashes in 3min.

It seem to be a Video issue, so I think the PCI-E speed is getting FUBAR, tho it still reports 100Mhz.

At this point, I think TME will need to be disabled, and a Update to SetFSB for the RTM875t-606, then we can just Soft Clock, without the PLL Mod.

 

does changing SRC_MAIN lock up the system for you?

 

I can Change SRC_Main from PLL1 to PLL3, however, it Links the PCI-E Speed to the Top Slider, and Reads 600MHZ, but the system does not Crash, it will if I try to over/under clock it, same as PLL1.

 

Hi
I have that same RTM chip and I disable non overclocking function bz simplz cutting the trace on the motherboard behind the resistor from pin 4 I didnt ground it I have 64 squared package ic. To make setfsb work i Have to choose rtm875t+ and i have to change bite 2 in byte 0 from 0 to 1 source of srs pll
And now I have to go to work i i will be back after 12 h. sh..et.

 

Thank you, that worked, I Grounded the Resister for PIN 4, but it works for me View attachment 40013

 

BTW,

TME = Throttle Test Mode Enable

 

From my PLL datasheet:

 

Don't Make Crazy Assumptions

 

what? 10char

 

i just want to hardware pull my FSB to 200 from 166. how do i do this without holding a needle on 10+11 pins everytime i boot

 

If you have RTM875t-606 TSSOP Package, you should Ground Pin 10.

Best to do it at a Resister, so you can undo it, if needed.

Pin 10 is also 48Mhz Source Clock for USB, I'm not sure how it's used, or if it's used at all in ICH chipsets.

 

if thats true then why do i have to release the needle connection between 10/11 in order for it to post? as long as i hold the needle there the computer does nothing, but as soon as i release it, it posts with 200(800)FSB

 

Because of this:

 

You are trying to pull pin 10 low via a needle *without* using a pulldown resistor. If the notebook isn't powering on then it's power circuit is detecting your short-circuiting.

Suggest placing a pulldown 10k resistor between your pin 10 and pin 11, or lift any resistor running off pin 10 and run it's leg to your pin 11 (GND).

 

Ok so find the outside of a resistor on pin 10 and patch it to 11? can u explain the easiest way to run a patch wire? the pins are so small that i don't know how to solder it...

also it powers on... just doesn't post.

 

ok so can some one in detail explain to me how to disable my tme or turn it to 0
i heard some one cut the line and then did darth ground pin 4?
some pictures would really help

 

Ok
I have some problems with setting FSC to logic 1 to achieve 333MHz. This is the photo of pll witout any mods http://img62.imageshack.us/i/img1072x.jpg/ and this one with resoldered two resistors on new places: http://img141.imageshack.us/i/img1078l.jpg/. It's working and its booting with resistors on new places. But if I cut the trace where is show on the photo and solder wire with external 3.3V thru 2kOHM there is no boot. Another thing is that: on this pin ther's potencial about 1.95V (with dafault configuration) and even when I cut trace ther's still that same voltyge on the pin(witout added 3.3V). It's releted probably with FSLC./TEST_SEL http://img42.imageshack.us/i/61101919.png/but It looks like those 1.9V is coming out the chip, I dont understand this frankly speaking. I still have know idea how to force it to logic 1. I found on russian sites: http://www.modlabs.net/notebook_overclocking very good article about overclocking notbooks and there are example with one PLL chip what I think is silego equivalent of ISC9lprs365 and RTM875t-606 because pins FSB and FSC are on that same places. This guy did 333MHz mod simply by cutting trace and giving 3.3V thru 2kOhm on pin 5. That's what I understand from this site. I'm open on suggestion because cutting trace and giving 3.3V on pin 5 isn't working for me.

 

ok so quick question... by shorting pin 10 with 11 its setting 10 to logic 0? so, does that mean i can just cut the trace?

 

I don't think so, and you may lose USB.

I tested my system via, pulling pin 10 to Ground, and everything was fine.

I highly recommend, you pull pin 10 to Ground by removing the first resister in the trace from the pin, then go to Ground with a 10k ohm resister from the PAD of the removed resister, the PIN side.

Sorry, I know that could not be less clear.

Any Ground will do, I used the Shield for my card reader.

BUT Why not just PULL TME to Ground and SetFSB via Software?

 

333MHz FSB, hmmm.......

1333 FSB Chipset?

 

could u post a pic and explain how u are soldering.. like solder size and what kind of iron u r using.. thx ... sorry i'm new to soldering on circuit boards.

 

Here is a thread:
http://forum.notebookreview.com/showthread.php?t=393027

 

I'm have some trouble uploading PIC's to the BBS, but I'll try.
Ok, that's the resister I removed

Here, is the resister still in place, look at the trace from pin 4, it goes to the top row last on the right. You can also see Pin 10 and 11 share some odd double I/O Resister.

Any old smd rework soldering station should work

Or, I used the cheapest thing in my toolbox, a 30W iron, the trick is a fine tip.

Remove the resister with care, you don't want to distroy the PAD's under it.

I use a neede, heat one end of the resister, heat the other, go back to the first end, heat it and pull THAT end up with the needle just a bit, do this to the other end, go back and forth until it's free.

Get a small piece of coated wire, strip the end you are going to connect to the Resister Pad, exposing as little wire as you can.

Heat the wire and add some solder to it, heat the PAD, connect them, heat them both.

Add an in line 10k Ohm resister if you like, but odds are the resistance of the wire will be fine.

It's help full to Tape the wire in place and bend it where you want it to stay, before you connect it to the Pad, but after you get some Solder on the wire.

Don't forget to connect the other end to Ground.