AMD Processors
Decrease font size
Increase font size
Topic Title: AMD Athlon 1000 MHz
Topic Summary:
Created On: 12/18/2004 12:14 PM
Status: Read Only
Linear : Threading : Single : Branch
<< 1 2 3 4 5 6 7 Previous Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 01/04/2005 06:01 PM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Congrats, Les, you found your IDE controller (VIA VT1611A)!


Unfortunately it's not your PLL chip. Any other oval aluminum cannisters, or other chips near that one?

This page is helpful for finding PLL chips:
http://cpufsb.de/FSB.HTM' ">http://cpufsb.de/FSB.HTM

Gonna see if I can find a photo of your mobo....

John

EDIT: OK here is a picture. Is this the mobo?

See that aluminum cannister at the very top of the board ("top" is the right side in this picture), above the right edge of the CPU socket? There's an oblong chip between the cannister and the CPU socket. I think that's it.

-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
 01/05/2005 02:10 AM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Ah it would be! The location you describe and I can see in the photo is right beneath the power supply. I did wonder if I had the right one.

Still the power supply has been out before but not tonight, I have a table tennis match to play this evening. I will keep ypu posted.

If I was right about the Nvidia card not liking the 37 MHz I am a bit stumped using this approach to get 1400 MHZ with an S3C unless CPUCool allows different options. Usually these frequencies are straight divisors of the basic crystal frequency aren't they? You said yourself 37 MHz was "getting a bit lofty".

-------------------------
Regards,
Les.
 01/05/2005 10:34 AM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Les,

You are right about there being straight divisors between the CPU and PCI frequencies. In this case, a new divisor of 4 is implemented right at 133 FSB, allowing the PCI bus to run at 33MHz. Hence the Manual lists "133/33Mhz/-0.5%, 133/33Mhz/±0.25%, 133/33Mhz/±0.5%."

But with the unlikelihood of actually being able to use 133 on this board, we're probably taking a different approach with the S3C anyway, making it run like an S3B by reducing the FSB to 100 and increasing the multiplier to 14.

John


-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
 01/05/2005 02:09 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Hi John,

100/3 = 33
108/3 = 36
111/3 = 37

is what I am getting at as I cannot see a way to give the S3C a 14x multiplier!

I think 107 x 12.5 = 1337.5 is probably the best I can hope for with the S3C but that is not bad.

I think the S3B has something in AMD's design/manufacturing process to make it read the SC3 and others's 5x multiplier as 14x as you yourself put me on to! Since the S3C's max is 12.5x I think that is it unlikely that it reads a 5x as 14x.

I am off to play table tennis now but it is hotting up as I won an AMD Athlon 1400 on eBay for £3.70 + p&p this evening.

What sort of 1400 it is, is not known and whether it is working or not is not known either but too good to pass up at that price for tinkering purposes I trust you would agree!

-------------------------
Regards,
Les.
 01/05/2005 04:48 PM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Hi Les,

At this point we have no idea if the 107 FSB is a hard limit because of the PCI/AGP speeds, or if your Duron 1000 just won't fun faster than that because it needs more voltage or better cooling. (I don't recall any discussion of CPU temp in this growing thread; what temperature does the CPU level out at with the 107FSB?)

We just don't know if a 14x multiplier is possible with the SC3. I'm hoping it made economic sense for AMD to make all 1400s the same internally, including remapping of the 5x multiplier, and just make them into Bs or Cs with the bridges.

We may know soon enough, when your chip arrives. Meanwhile, keep those balls in the air!

John

-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
 01/05/2005 07:58 PM
User is offline View Users Profile Print this message

Author Icon
-Milt-
Senior Member

Posts: 1722
Joined: 03/15/2004

Hi John, Les,

The 107, or 108, or 110 MHz is definitely not a 'hard limit' because it would boot at 107, but then hung on 108, going into WinXP.
So someting doesn't want to go faster than 107, at least not on the stock voltage.

It could be any one of the CPU, (most likely, even if Les only has PC100 memory), the memory, or the KT133 chipset on the mobo.

Try bumping the Vcore voltage to the max, and see if it'll load into WinXP, and run normally at 108 MHz.
If it does, then we know it was the CPU that's our 'weak sister'

But always keep one eyeball on the CPU temperature, whenever you increase the FSB or the Vcore... but especially the Vcore... it's the Vcore that will raise the power consumption. And power = heat, and heat is our enemy.
It's the temperature that shortens the life of any electronic component, not the voltage. (within reason, of course)
And, as the core speed of any CPU goes up, it becomes less and less 'temperature tolerant'.
My own personal "This is getting too hot to be expecting reliability temperature" is in the 53° ~ 55° Celsius range.
The general 'rule-of-thumb' is that for every 10°C that you can lower the temp, you will DOUBLE the life expectancy.
Here's a link to the bestest, simplest to understand explanation, that I've ever read, graphically showing you just WHY this increase in voltage can help with overclocking ANY component inside a computer...
http://www.rhcf.com/sis-bin/ul....cgi...29;t=000004;p=0' ">http://www.rhcf.com/sis-bin/ul...opic;f=29;t=000004;p=0

If the added Vcore can't get us from 107 to 108 MHz, with reliability, then it's not the CPU that's at fault. It should.
I highly doubt that it's the memory, but what have you got for memory, Les?

That leaves us with the chipset, which is already known to be the 'weak link' on boards equipped with the KT133 chipset.
Possibly, just adding a passive cooler (passive = 'no fan', or HS, to the larger VIA Northbridge to help cool it will help to get the FSB higher.
You can easily check the temps of your chipsets by using the 'finger test'... if it's 'hot' to your touch, then let's figure out how to get some kind of cooler attached to it.

The big square VIA chip, inbetween the CPU and AGP slot is your 'Northbridge'.
The smaller VIA chip to the right of the PCI slots is the 'Southbridge'.
Together, they make up your 'Chipset', or set of mobo control chips


-------------------------
Here is my new PCChips M848A' ">http://img.photobucket.com/alb...2_T-2-3-3-6-2_M848.jpg v2.1, with a 1700+ @ 2411 MHz, and my trusty old PCChips M810L' ">http://members.lycos.co.uk/mmm...00MHz_Sandra_tests.jpg v7.1A, with a Barton 2500+ @ over 3500+. Some of my fancy cars are on this pag
 01/06/2005 03:04 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

How do I, "keep an eyeball on the CPU temperature"? I see initial temperatures in the POST but after that do I have to rely on the touch test?

Anyway this evening I have taken another step forward. What I tried tonight was to set 108/36 MHz with +0.05 V and like this it did boot into Windows XP and I ran IE OK.

So with the warm flush of success again I went back to the BIOS and nudged the frequency up to 109/36 MHz leaving Vcore at +0.05 V.

When I Saved and Exited the BIOS the Nvida GForce 32 MB announcement appeared in the top right of the screen for a second or two (definitley for longer than at 110/37) but this was followed by a continuous beep and a blank, black screen. OK so it is reset the CMOS time again but now I can see that even without John optimism that the S3C might read the 5x multiplier as 14x that I might get 108x12.5=1350 MHz with a degree of stability. It gets better all the time. I just hope my "bargain" processor is a good 'un!

The 1GHz one that is in there now is an Athlon and the machine was bought during September 2001 so does this make it an early production run one. The kill joys among us are telling me that early production runs produced some really good processors and these were bridged for the highest speeds the also rans in the batch were bridged for lower speeds. The nub being that if mine is an early one it may not be capable of going much faster!

-------------------------
Regards,
Les.
 01/06/2005 04:51 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

OK I have been back to the AMD Athlon machine and used a systematic approach to see if I could determine any limiting factors.

While in there I noted on the oval can John directed me to, 14.3F0LB and next to it on a rectangular chip with legs down either side, W230H B121 1052TTA. I could just see enough without taking the power supply out.

Now to the over-clocking tests:

In the POST I can see CPU temp degC / degF and System temp degC / degF

To begin with, with the machine virtually cold:

CPU temp 23 C/ 73 F
Sys temp 19 C/ 66 F

I set 109/36 MHz & +0.1 V

The boot got as far as completing the memory check, put up the message.
Detecting IDE drives and then it hung.

So I rebooted and saw:

CPU temp 27 C/ 80 F
Sys temp 20 C/ 68 F

This time it hung after completeing the memory check and before displaying the Detecting IDE drives message.

I reset and during the start of the boot saw:

CPU temp 32 C/ 89 F
Sys temp 23 C/ 73 F

I set 108/36 MHz & +0.025 V

The machine booted to the point where the user login icons should be and then started to reboot of its own accord. I let it do this twice before I reset and saw:

CPU temp 34 C/ 93 F
Sys temp 25 C/ 77 F

So I set 108/36 MHz & +0.05 V (that had worked earlier in the evening).

This time the Safe mode Last known good/ Boot normally menu came up and froze
so I reset again (note these are CMOS jumper resets) and saw:

CPU temp 35 C/ 95 F
Sys temp 26 C/ 78 F

On the boot I was asked to load the BIOS Defaults, which I did resetting 100/33 MHz & nominal Vcore. This time at the Safe mode, Last known good/ Boot normally menu I selected Last known good and it booted up OK.

My observations are that as it warmed up it was more prone to lock up even with over-clocking setting that worked when it was cooler. In my last post at 108/36 MHz +0.05 V from more or less cold it worked. Later when I tried the above test it was warmer and unable to work with this pair of settings!

I have not tried it for a long time but I now think 107/36 MHz is probably as high as I can go.

I replaced the CPU heatsink and fan some time ago and the heatsink is as big as it can be as it just misses some capacitors between the Memory and the processor by the width of a sheet of paper!

Maybe there will be more scope with the 1400 MHz processor.

Regards,

Les.

A P.S. edit:

After 30 minutes plus with 100/33 & nominal Vcore

CPU temp 36 C/ 96 F
Sys temp 27 C/ 80 F

Then set to 107/36 with nominal Vcore the machine booted fine and was left for 15 minutes before re-booting.
During re-boot:

CPU temp 37 C/ 98 F
Sys temp 28 C/ 82 F

After a further 5 minutes on re-boot:

CPU temp 37 C/ 98 F
Sys temp 28 C/ 82 F

That is relatively stable.

Note:
All of tonight's temperatures are with the side panel off the tower.
With nominal setting and the case closed I have seen CPU temps of 40 C/ 104 F in the past.

That's all for tonight I am sure you will be pleased to hear!

-------------------------
Regards,
Les.
 01/06/2005 07:27 PM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Hi Les,

Nice experimentation! You are learning a lot, and I hope enjoying yourself as well.

There is probably somewhere in your BIOS setup, not the POST screen, where you can monitor temperature and wait for it to stabilize. But you will use that primarily to help set up a Windows-based monitoring tool, which is what you really need.

Two good ones are Motherboard Monitor 5 and SpeedFan. I use both, depending on the user and the need. Speedfan is easier to set up, and I recommend you try that first. You can download either from links in the Utilities folder in My Bookmarks, in my sig below. At first you won't know which temperature is your CPU and which is the system. Well, actually if there are only 2 you can pretty well assume the higher one is the CPU. But there may be more, and in any case it's a good idea to verify them against the ones shown in the BIOS. With a Windows-based monotor, you can watch temperatures when the CPU is idle (same as when in BIOS) or under load such as a game.

Your temperatures look very good. You are right that lockups are more likely at the higher temperatures. Temperature is a limiting facor in overclocking, although one that can be dealt with to some degree through improved cooling. There is a temperature double-whammy with overclocking: Both the overclocking itself, and increasing Vcore, elevate the CPU temperature. To make matters worse (the second "whammy" ), the CPU is more sensitive to temperature when overclocked. It will shut down at a lower temperature when overclocked. So temperature control becomes very important. Your temperatures seem quite low to be encountering this phenomenon with a mild overclock, but all CPUs are different. I usually tell people if their CPU temperature is 45ºC or less, go ahead with the next step of overclocking. And as Milt says, it could be the chipset that is getting too warm, and we aren't even measuring its temperature.

The 1400 will run hotter, but there are things you can do, even with limited space around the mobo. We'll cross that bridge if we come to it. One thing you might try right now is substituting Arctic Silver 5 for whatever thermal interface material is now between your CPU and heat sink, for a likely 3-5ºC temperature drop.

Your overclocking experimentation will become easier if we can get you set up with CPUCool, which allows FSB changes while running in Windows. If you lock up, all that is needed is a restart, but no CMOS reset. Saves lots of time and tends to lower the fear factor. Since you have now identified your PLL as an ICW W230H , you are getting closer to being able to use CPUCool. The version of CPUCool I have on this machine lists the W230( 3) A and W230( 3) B, but not a W230H. But he way CPUCool lists the names of PLLs is not always exactly the way they appear on the chip. We can look into this further.

That's great progress today, Les, and never mind the killjoys! Your Athlon 1000 is what it is and they are all different. It''s not like you are out shopping for the one with the best overclocking prospects. We'll help you get the most out of it.

John

-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
 01/07/2005 07:56 AM
User is offline View Users Profile Print this message

Author Icon
-Milt-
Senior Member

Posts: 1722
Joined: 03/15/2004

Hi guys,

Excellent posts by both of you. Nice to see some teamwork, and a distinct lack of nay-sayers!

Since your CPU temp is so nice and cool (even the 37 C/ 98 F), I'd say it's perfectly safe to set your Vcore voltage boost in CMOS to the highest allowed ( +0.10v) and just leave it there.
Get the temp monitoring programs for Windows as John suggests, so you can 'keep an eyeball on the CPU temp'
My preference is SpeedFan, but try them both and see what you like.

But your next job is to go get CPUCooL from...
http://www.pcworld.com/downloa...down...ileidx,1,00.asp' ">http://www.pcworld.com/downloa...22842,fileidx,1,00.asp

When you are first installing it, set it so that it loads with every Windows start-up.

Reboot (during this loading of CPUCooL, I'd suggest settings of +0.10 volts and an FSB of 106.0 MHz... that way you'll know you are in 'safe' territory)

First thing it asks when it comes up for the first time is "Which mobo & Model"
You need to select the motherboard brand, version etc., before you can use it to change the FSB, which is what we want to do, first and foremost

Your "Mainboard" is a Jetway, but your 663AS Pro v1.2 is not listed, so just click "I don't know"

Now, click "Functions"> "Change front side bus/change CPU speed"

Under PLL manufacturer, select "cypress/icw"
Under PLL type, select "W230( 3)B"

Start clicking on (or click and hold), the tiny little 'up' arrow, to the right of "Tray frequency 1" until it shows FSB : 111.0 MHz PCI : 37.0 MHz

Go to "Tray frequency 2" and take it up to FSB : 110.0 MHz PCI : 36.6 MHz

Go to "Tray frequency 3" and take it up to FSB : 109.0 MHz PCI : 36.3 MHz

Go to 'Tray frequency 4 and take it up to FSB : 108.0 MHz PCI : 36.0 MHz

Click 'OK', and EXIT CPUCooL, and then re-open it.
This will temporarily 'lock' the above frequencies into CPUCooL's memory.

Assuming that you told CPUCooL to start with every windows start, there should now be a new icon beside the clock in your 'System Tray', and it'll be there every time you reboot.

Double-click on that Icon to open the Main Display, anytime you want to see it.
Right-click on that same icon, and a pop-up menu appears... click the lowest "FSB: MHz" number (the bottom two numbers usually display identical numbers... probably "FSB: 109 MHz in your case... click the bottom one)

What's CPUCooL showing your processor speed at now?
Use either of WCPUID or CrystalCPUID to verify the speed, if you don't want to believe what CPUCooL is showing you in the "Main Display" window
(If you go into Display Options> This Display (Main Display)> Graphic display and check the box at lower left, it will graphically show the speeds, as you increase them)
While you're in there, check whatever else you'd like to see graphically displayed, like your CPU temp.

WCPUID is available free here www.h-oda.com/
CrystalCPUID is available free here http://crystalmark.info/?lang=en' ">http://crystalmark.info/?lang=en

Keep going up that FSB "string" in the little pop-up window that you have created until the comp 'locks' the moment you click on it, or you find that it's unstable in Windows, after you do.
When that happens, just shut off your computer, and then reboot it, and Bingo! You’re right back to your original boot setting from BIOS

Play with it a bit and get back to us with your results.
I expect that you will have very little problem with getting your Athlon 1000 system to run at 10.0x 110 = 1100 MHz, by using CPUCooL
But, I've been wrong before, and your KT133 chipset is a 'fly-in-the-ointment' in your case.
I've never worked with one, just read a lot about them.

And the nice thing is, that with CPUCooL, you can NOT damage your confuser, or your CPU, IF you stick to the guidelines given above.

-------------------------
Here is my new PCChips M848A' ">http://img.photobucket.com/alb...2_T-2-3-3-6-2_M848.jpg v2.1, with a 1700+ @ 2411 MHz, and my trusty old PCChips M810L' ">http://members.lycos.co.uk/mmm...00MHz_Sandra_tests.jpg v7.1A, with a Barton 2500+ @ over 3500+. Some of my fancy cars are on this pag
 01/07/2005 05:23 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Less progress today, my daughter brough my 10 month old Grandson to visit!

I have, however, downloaded and installed CPCool from the PC World link you gave me and have it running. When I go to Functions and select PLL Manufacturer and cypress/icw a box pops up saying:

False PLL

You have found a PLL who does not have an SMBUS.
The PLL manufacturer did not implement this feature.
It will also not be possible to change the frequency in the future version of any program - the hardware does not support it...

Then there is a link to http://www.podien.de/FSB.HTM' ">http://www.podien.de/FSB.HTM

When you click OK you can select W230(3)B but I am unsure about this now having read the pop-up message.

What do you make of this?

Until I get your answer I have simply matched the BIOS settings 107/36 by putting
107 35.6 in all four of CPCool's frequency trays.

I have also noticed in red text at the bottom of the pane with the frequency trays this mesage:

Write Only PLL There is no reading of the PLL available only setting the new frequency.

I guess this probably means I can do what I need to despite the pop-up box. Is this correct?

By the way the machine is running sweetly at 1.07 GHz and so far is still on nominal Vcore! The temperature after over half an hour on, is 23 deg. sitting there in the system tray - great eh!

-------------------------
Regards,
Les.
 01/07/2005 07:34 PM
User is offline View Users Profile Print this message

Author Icon
-Milt-
Senior Member

Posts: 1722
Joined: 03/15/2004

Hi Les,

That's the same warning message that I got when I selected the W230(3)B PLL, using my PCChips M848A mobo, except mine didn't add the "It will also not be possible to change the frequency in the future version of any program - the hardware does not support it" message.

Part of the beauty of using CPUCooL is that if you totally screw up, and select a way-too-high FSB, or even the wrong PLL, and then hit "Set Frequency", the worst that's ever happened (to my knowledge, and in my own personal experience) is that your machine will 'Freeze'
You'll have to reboot.
Big deal!

So give it a go... it may freeze on you, it may not, or it may not do anything at all (if "the hardware does not support it")

Or, it may work like a charm, and allow you to reach new highs!
You'll never know until you try it...
Good luck!

P.S. Because your CPU temp is so low (23° Celsius at idle), I suspect that you have the 'wrong' temperature on display.
Which monitoring program are you using?
SpeedFan I'm familiar with, but we'll need John's help to configure MBM5

-------------------------
Here is my new PCChips M848A' ">http://img.photobucket.com/alb...2_T-2-3-3-6-2_M848.jpg v2.1, with a 1700+ @ 2411 MHz, and my trusty old PCChips M810L' ">http://members.lycos.co.uk/mmm...00MHz_Sandra_tests.jpg v7.1A, with a Barton 2500+ @ over 3500+. Some of my fancy cars are on this pag
 01/08/2005 05:53 AM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Hi Milt,

I will give it a go today when the Saturday chores are done using the steps you gave me ealier.

I having been checking back through the threads of this growing topic and found the following two point that I can elaborate on:

i) "I highly doubt that it's the memory, but what have you got for memory, Les?"

I had posted this earlier:
"Another thought I have had is about the capability of my original memory it is probably only 100 MHz. The 512MB I added myself was from Crucial and is described as Part No. CT64M64SD75512MB 168-pin DIMM PC133.
The orginal memory the machine came with is simply described as 256 MB SDRAM".

It is still the original 256 MB DIMM that is the unknown as the 512 MB Crucial one I added is PC133. Is there a way to tell, if I fish the 256 one out and take a look at it?

ii) "That leaves us with the chipset, which is already known to be the 'weak link' on boards equipped with the KT133 chipset.
Possibly, just adding a passive cooler (passive = 'no fan', or HS, to the larger VIA Northbridge to help cool it will help to get the FSB higher".

Taking a look at the photograph John posted of my MoBo the VIA chip centre photo is, I take it, the North-Bridge from what you said in your post about temperature control. I can comfirm that on my paticular MoBo this chip has a heat sink on it already. A green "anodised" looking thing. The smaller VIA chip over near the CMOS battery has no heatsink.

Having one already on the North-bridge is a good omen though!



-------------------------
Regards,
Les.
 01/08/2005 04:17 PM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Les,

There may be a label on the original RAM, or some numbers a real memory guru could interpret. But we should be more interested in how it performs than what it is rated. Try your overclocking experiments with one stick or the other.

Having that heat sink on the northbridge may be a good thing, but some manufacturers attach them with double-sided tape that insulates them from the chips. So you may and up modding it anyway, but at least you have a heat sink to use.

John

-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
 01/08/2005 05:42 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Hi John & Milt,

Thanks yet again for the information especially the warning about double sided sticky tape John!

For the record the only program I have downloaded and installed so far is CPCooL.
In my System Tray the little CPCooL icon changes to the temperature reading I have referred to, a second or so after I send it's open window to the Tray when I have finished reviewing or making settings. So this temperature may not be what I thought it was!

As the saying goes Rome was not built in a day as I think the news is not so encouraging this evening (well it's evening here in the UK). Everything I have tried with CPcooL just freezes the machine. I can just hear one dissenting voice saying told you so already.

Anyway for those with the stamina here is what I have tried tonight:

I printed Milt's step by step instructions so that I had them in my hand and carried them out to the letter as follows: (Please spot a mistake)!

On the first boot of the evening I entered the BIOS, set Vcore to +0.1 Volts and the clock frequencies to 106/35Mhz/Off then Saved & Exited.
In XP using CPCooL I set the four Tray frequencies as Milt specified namely:
Tray frequency 1 : 111 37.0
Tray frequency 2 : 110 36.6
Tray frequency 3 : 109 36.3
Tray frequency 4 : 108 36.0
Clicked OK and then EXIT

This removed the icon from my System Tray, which did surprise me but I have a CPCooL icon on the desktop and it re-started just fine from this one. I confirmed my setting were still set, then sent it to the TRAY and saw the little CPCooL icon briefly before it changed to the temperature reading I have mentioned.

I right clicked on the temperature reading and saw the pop-up Milt said I would with the FSB's I had just set, with the exception Milt mentioned, of the lowest two being the same value, both at the Tray 3 value of 109. The 108 setting did not show.

At this point I had the buzz of success as everything had gone to plan up to this stage, so I clicked the lowest 109 (the fourth one down the menu) and the machine froze instantly leaving the mouse pointer stuck to a blue box with the frequency I had selected in it on my Desk Top!

I pressed my Reset push-button and the screen went blank but no re-boot started so I powered the mains supply off and on again at the wall socket and re-started the machine. This time it booted OK and I saw 1060(106x10.0) in the POST in-line with what Milt explained CpCooL would allow me to do.

When the machine had re-booted I went into CPCooL and changed these two settings to be:
Tray frequency 3 : 107 36.0
Tray frequency 4 : 107 36.0

With these settings I right clicked on the "temperature" in my System Tray once again:

This time I saw:

FSB : 111 MHz
FSB : 110 MHz
FSB : 107 MHz
FSB : 107 MHz
CPCooL Exit
Eject .......... etc.

To my disappointment clicking on the 107 at the bottom of the list made the machine freeze again! The press the Caps Lock key test confirmed what I already new, the machine was locked up solid once again.

I rebooted and in the BIOS set Vcore to +0.05 Volts. I don't know why I guess I just had runs out of positive ideas to try. Back in Windows I went for the 107 second from the bottom of the list just for a change but everything froze solid once again.

Finally I re-booted and in the BIOS set the Default Vcore. Back in Windows I clicked on the lowest 107 and this time (maybe by fluke) it did not freeze instantly it let me move the mouse pointer as far as the centre of the screen before locking up solid!

Is it the limitation of my VIA chipset?
Could it be to do with the False PLL message I saw when I selected cypress/ipw?
I know the machine runs fine on the default Vcore at 1070 MHz when set in the BIOS but would not let CPCooL change to this value from 1060 MHz.

I thought about emailing CPCooL support but fought shy of that as I haven't registered it yet. Mind you if it won't work with my hardware I am not going to register it!

Is it a lost cause chaps?

Perhaps I should look to John optimism that the S3C I am waiting for will read the earlier 5x multiplier link as 14x because, as he has pointed out a couple of times now, this would let me continue with the 100/33 FSB/PCI settings.

I am in need of a shot of encouragement chaps.

Regards,

Les.

A final P.S. edit for tonight.

An anonymous forum snooper from "whom" I chose my user name has emailed me with a useful link to a program that gives in depth information about your hardware, well for 14 a day trial period anyway. The link is: http://www.hwinfo.com' ">http://www.hwinfo.com
From this website I have just downloaded, installed and run, HWINFO32 and this utility has given me the following details about my installed memory:

Row 0 - 512 MB PC133 SDRAM
Row 1 - 256 MB PC133 SDRAM

So that lays that one to rest!

Goodnight from the UK!

-------------------------
Regards,
Les.
 01/09/2005 08:23 AM
User is offline View Users Profile Print this message

Author Icon
-Milt-
Senior Member

Posts: 1722
Joined: 03/15/2004

Les,

I haven't got any brilliant answers for you this morning, Les, sorry.

The only thing I can think to suggest is to try the 'other' W230 PLL and see if you get any better results with CPUCooL using it.

-------------------------
Here is my new PCChips M848A' ">http://img.photobucket.com/alb...2_T-2-3-3-6-2_M848.jpg v2.1, with a 1700+ @ 2411 MHz, and my trusty old PCChips M810L' ">http://members.lycos.co.uk/mmm...00MHz_Sandra_tests.jpg v7.1A, with a Barton 2500+ @ over 3500+. Some of my fancy cars are on this pag
 01/09/2005 09:13 AM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Hi Milt,

No worries, I can be quite relaxed about a tinkering machine while my P4 is running sweetly for me to do my other business with.

Anyway I am not ready to give up yet so this afternoon I have sent the following email to support@podien.de to see is this brings forth any enlightenment:

Hi Podien,

I downloaded a trial version of PCUCooL from PCWorld.com, pcucool9.exe, installed it and ran it.

I have an AMD Athlon 1000 on a Jetway 663AS Pro rev 1.2 Motherboard with a PLL chip with the following on it icw W230H B121 1052TTA.

When I run PCCooL having selected, "I don't know" because for Jetway there is no 663AS Pro MoBo listed and go to Functions and select PLL Manufacturer and cypress/icw a box pops up saying:

False PLL

You have found a PLL who does not have an SMBUS.
The PLL manufacturer did not implement this feature.
It will also not be possible to change the frequency in the future version of any program - the hardware does not support it..

Not fully understanding the implications of this message I clicked OK and chose W230( 3)B as there was no W230H listed, following advice from others more experienced in these matters that me.

With 107/36 MHz "hard set" in my BIOS the machine is stable in Windows XP Pro on default Vcore.

So I rebooted, entered the BIOS and set Vcore to +0.1 Volts and the clock frequencies to 106/35Mhz/Off then Saved & Exited.
In XP using CPCooL I set the four Tray frequencies to be:
Tray frequency 1 : 111 37.0
Tray frequency 2 : 110 36.6
Tray frequency 3 : 109 36.3
Tray frequency 4 : 108 36.0

and saved them.

Next I right clicked on the "temperature reading" (23 deg) in my System Tray and in the pop-up menu saw the frequencies I had just set, with the exception that the lowest two being both 109 MHz. The 108 setting did not show!

At this point I had the buzz of success as everything had gone to plan so far, so I clicked the lowest 109 (the fourth one down the menu) and the machine froze instantly leaving the mouse pointer stuck to a blue box with the frequency I had selected in it on my Desk Top!

Subsequently I have re-booted a number of times and tried lower frequencies in the Trays, frequencies I know work when "hard set" in the BIOS but each time even selecting these frequencies the machine freezes.

I noticed in red text at the bottom of the pane with the frequency trays this message:
Write Only PLL There is no reading of the PLL available only setting the new frequency.
Seeing this is what encouraged me to try selecting frequencies from the right click pop-up menu in the system tray.

Does your False PLL message preclude me from using PCCooL with my Jetway/icw hardware?

Any help and/or advice would be very much appreciated.

Regards,

Les.


-------------------------
Regards,
Les.
 01/09/2005 09:33 AM
User is offline View Users Profile Print this message

Author Icon
-Milt-
Senior Member

Posts: 1722
Joined: 03/15/2004

Excellent!

Wolfram Podien is known to be a caring individual, who almost always responds to a well written query like yours.
I sure hope that he has an answer for you.

Did you try the 'other' W230 PLL yet? It just might work, but don't get your hopes up too high.

The newest version (well, the Vers 7.3.2 from Oct 10, 2004) only lists the W230( 3)B, and not the 'other' one, that's in the earlier version that I recommended to you.
This 'later' version can't 'remember' your settings when you exit... just Wolfram's way of encouraging you to 'register' your copy... then it gets it's memory back!

-------------------------
Here is my new PCChips M848A' ">http://img.photobucket.com/alb...2_T-2-3-3-6-2_M848.jpg v2.1, with a 1700+ @ 2411 MHz, and my trusty old PCChips M810L' ">http://members.lycos.co.uk/mmm...00MHz_Sandra_tests.jpg v7.1A, with a Barton 2500+ @ over 3500+. Some of my fancy cars are on this pag
 01/09/2005 12:21 PM
User is offline View Users Profile Print this message

Author Icon
snooper
Member

Posts: 77
Joined: 12/18/2004

Hi Milt,

While waiting to see if Wolfram Podien will reply I tried the other W230( 3)A option but it is much the same as the B.

What I did try, not that it would be much use pratically speaking, was to put the low end frequencies in the Trays with the following results obtained twice for each PLL version:

......................W230( 3)A......W230( 3)B
Tray1 : 107 ........ froze ............. froze
Tray2 : 106 ..... 1059.5 .......... 1059.5
Tray3 : 104 ..... 1039.9 .......... 1039.9
Tray4 : 102 ..... 1020.1 .......... 1020.1

The "hard set" BIOS values were 107/36 MHz and +0.1 Volts Vcore and CPUCooL showed 1070.1MHz in it's panel before I changed anything. I can change between values of 106 and below but as soon as I try 107 or above it locks.

I think this absolves CPUCooL from not working with my hardware though, as I can change frequency at the very low end. I am coming to the conclusion that a 7% increase is all I am going to get out of this particular processor and that only when it is hard set in the BIOS!

I used the set frequency option instead of the right click approach to set 100 MHz and 102 MHz successfully but using this option it locked up again at 107 MHz.

Just to see what would happen I also used the set frequency option to try 133/33 but this locked up solid as expected.

While I was doing all this the highest temperature CPUCooL reported in my system tray was 23 deg but looking in the POST I saw CPU 38C/ 100F and System 27C/ 80F. I have accepted the 7% increase for now with 107/36 set in the BIOS at default Vcore. As all appears to be stable at this voltage I can see no point in driving it any harder.

I have also sent the following email to AMD TSC Support at tsc.support@caexmtap.amd.com :

Hi Ryan,

It has been a while so I trust you are well. I now have a further question.

The information you gave me below has finally encouraged me to look at
upgrading my machine.

The information I have found on the web about AMD Athlons leads me to
believe that the maximum multiplier you can get with the links on the
processor is 12.5x. If this is correct I can understand how an A1400AMS3C
runs at 1.4 GHz as with an FSB of 133.3 MHz x 10.5 = 1400 MHz.
The A1400AMS3B on a 100 MHz FSB with it's double data rate facility needs a
14x multipier to achieve 1.4 GHz. Is this achieved by reassigning a lower,
say 5x multiplier to be 14x. I know AMD K6 processors read the 2x multiplier
on some PCChips mother boards as 6x and wondered if this same techinque
applied to the Athlons. If it does can you tell me which lower multiplier
has become 14x and if it applies to both the S3B and the S3C?

Your help and advice would be very much appreciated.

Regards,

Les

> Hello Les.,
>
> Thank you for contacting AMD's Technical Service Center. Yes, the 1.4Ghz
> Athlon was available with both a 200Mhz FSB (A1400AMT3B), and a 266Mhz
> FSB (A1400AMT3C), as well the 1Ghz part was available with both the
> 200Mhz (100 x 10), and 266Mhz 133 x 7.5) FSB's.
>
>
> Hope this helps. If you have any other questions, please feel free to
> contact me.
>
> Regards,
> Ryan Gardner
> CPU Specialist
> AMD TSC




-------------------------
Regards,
Les.
 01/09/2005 05:24 PM
User is offline View Users Profile Print this message

Author Icon
JHawk56
Elite

Posts: 1084
Joined: 07/27/2004

Hi guys,

I've searched far and wide on the Web and can find very little about the 14X multiplier. I did see info. on BIOS settings for one brand of board (can't remember which) that remapped one of the lower multipliers to 14X, but this time it was the 10.5X. Maybe the 5X I saw earlier was a typo.

The Jetway board cannot set the multiplier in the BIOS or my jumpers, IIRC, so all this hardly matters. All I can think to do is find the bridge configuration of the B model (200MHz), and replicate it on the C model, if indeed that is what arrives.

John

-------------------------
My ECS/PCCHIPS/Amptron PCs<br>ASRock K7VM2
Statistics
112018 users are registered to the AMD Processors forum.
There are currently 0 users logged in.

FuseTalk Hosting Executive Plan v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.



Contact AMD Terms and Conditions ©2007 Advanced Micro Devices, Inc. Privacy Trademark information