600X. Is This a Non-Speedstep Board.. or what?

[Robin Mutoid]

Details: 600X.

Model: 2645 8EA.

BIOS: ITET55WW. (Latest).

Board: J1D8U060UWK.

Unit: 99TZGN4.

O.S. Win 2K/XP. Intel 'Speedstep' Installed. Good Battery installed.

I just fitted an Intel 'PMM65002201AC' processor to the above (shows PIII 650Mhz 'Speedstep' in BIOS). It came with a 500Mhz unit as standard.

With Intel 'Speedstep' prog. installed and set to 'Max Performance' on Batt and AC, when the machine boots I get flashing '-' cursor only. THEN when I unplug the yellow AC adaptor to run on Batt only, it boots fine @ 650Mhz. Once booted I can replug AC adaptor. It appears to behave in the same way if I initially boot on Batt only (ie. flashing '-' cursor only) until I plug IN AC adaptor when it then boots OK @ 650Mhz.

Can someone please explain what's going on here?!

Is this a non-'Speedstep' board being 'fooled' by the unplugging trick.. or is something else possibly at fault?

Thanks

Robin.

PS. I actually have two machines that seem to behave in the same way, the other is:

Model: 2645 4EA.

BIOS: ITET55WW.

Board: J1BM49251DN.

Unit: 99RL972.

[whizkid]

That's the classic symptom of putting a SpeedStep CPU in a non-SpeedStep system board. When it hangs, try generating an ACPI event (I like to press Fn+F2). That might unblock it.

Most units that came with the 500MHz CPU, and all that came with a 450MHz CPU had non-SpeedStep system boards.

[Robbyrobot]

Is this a non-'Speedstep' board being 'fooled' by the unplugging trick.. or is something else possibly at fault?

I think you were right the first time (non-Speedstep mainboard). As far as I know, only the 600x systems that originally came with a 650MHz CPU had SpeedStep support.

[Robin Mutoid]

Thanks Guys..

'Fn+F2' does the trick fine.

Would carrying out the 'PIII to 600E' processor board speedstep mod (cutting the track and fitting a 2.2K resistor) negate even having to enact 'Fn+F2' @ boot on the 600X?

Robin.

[Robin Mutoid]

Before this topic disappears forever down the list..

Does anyone have an answer to my last posted enquiry?

ie. Would effecting the 'speedstep' processor board mod used to allow a speedstep processor to run at its higher speed on the 600E board also function to allow the non speedstep 600X board to initially boot the speedstep processor at it's higher speed?

Not hasseling.. just keen to know from those who may actually know!!

Thanks

Robin.

[Robbyrobot]

Unfortunately, I have no experience or knowledge on this, but you might ask Ray (rkawakami) by PM if he hasn't seen this thread, since he has a great deal of experience with the 600 series.

[ms5133]

I'm kind of interested also.

[Robbyrobot]

I'm kind of interested also.

In that case, ask him to answer here... he's very active and generally comments on anything he reasonably can, but maybe he hasn't seen this thread yet.

[pkiff]

I don't have experience upgrading a non-SpeedStep 600X to a SpeedStep CPU, but I'll mention a couple stray ideas nevertheless.

First, I would make clear that the non-SpeedStep 600X models are somewhat unique. They do not behave the same way as non-SpeedStep 600E or 770X/Z models. Nor do they behave like SpeedStep 600X models (obviously). Also, as you have no doubt discovered, there has been considerably less work done on upgrading non-SpeedStep 600Xs than on the 600E series. And there appear to be significant differences within the 600X model series, too: some models seem to choke up permanently with a SpeedStep CPU, others run them at low-power (stepped down) speeds, while others actually seem to run SpeedStep CPUs at full SpeedStep speeds even though they supposedly don't have SpeedStep support on their motherboards. My point with all this is that it is unlikely, therefore, that anyone will be able to give you a definitive answer about the expected behaviour of your specific machine with a SpeedStep CPU installed.

A couple other stray points:

1. Have you "initialized" your BIOS? This should be done after the CPU upgrade before trying other options.

2. You say "O.S. Win 2K/XP". Are you dual booting? The Intel SpeedStep app does not work correctly under XP since XP has its own SpeedStep functions built-in. These are two VERY different OSes with respect to SpeedStep behaviour in the 600 series. If you have no preference otherwise, then I would recommend trying Win2K, since some people have managed to get non-SpeedStep 600 machines to run SpeedStep CPUs at max speeds without any hardware mods.

3. Did you get the Error 127 [corrected from 128], or are you getting it now? You don't mention this in your post. If you aren't getting that error with the PIII installed, then there may be something else going on. And if you DID get the 127 Error, then did you edit the BIOS to remove it? And are you running the PowerLeap cache now??

4. Have you tried disabling the onboard memory? And are you sure that you have PC100 memory installed in the user-accessable memory slots?

5. Some folks with symptoms not unlike yours have tried to change the "processor driver" for the PIII in XP. I haven't heard of this being necessary in Win2K. But if you are using XP, then it is worth investigating.

And finally, to answer your core question,

Would effecting the 'speedstep' processor board mod used to allow a speedstep processor to run at its higher speed on the 600E board also function to allow the non speedstep 600X board to initially boot the speedstep processor at it's higher speed?

I can't say for sure about how a non-SpeedStep 600X will respond, but in general, yes, that is what the MMC-2 hardware mod is supposed to do with the 600E. And then the typical user will also install the DeepSleep utility to make sure that the higher speed is enabled on suspend/hibernate/resume as well as on boot-up. But I don't recall seeing any actual reports of real experiences with this on a non-SpeedStep 600X.

Phil.

[rkawakami]

For the record, I have not performed any CPU mods to the 600 series besides replacing a 366Mhz board with a 400Mhz board in a 600E. Nor do I have much experience with a 600X Speedstep system. The only one I had was recently sold it to gator.

I did notice a strange thing with using that system with Windows 2000. Win2K would throttle back the front side bus speed in half when running on just the battery, in addition to Speedstep's reduction of the CPU speed. Both of those coupled together would sometimes result in the normally 650Mhz system running at around 260Mhz .

[Robin Mutoid]

Firstly, 'Thanks' for your interesting advice on the many different apparent behaviours of the various 'guises' of the 600X.. I had no idea that so many variations had been reported!

Of the 'options' both my 600X non-speedstep machines appear to be of the 'will actually run the processor at full speed despite not professing to support 'Speedstep' variety. I have a full speedstep 600X with which to draw comparison. The generation of an ACPI event (Fn+F2) as suggested by 'whizkid' successfully allows booting to high speed with minimum 'fuss'.

As to the other 'stray points':

1). Yes I 'Initialized' BIOS on fitting the processor.

2). Re. Win2K/XP; I am not dual booting but have two HDD's one loaded with each OS. I was aware that XP contains its 'own' Speedstep but loaded the Intel version as I felt that it would not conflict with XP but simply serve to give me some control over things.. do they 'clash'? I actually prefer 2K in the 600's... lots less 'BS' to load!

3). The reason I didn't mention 'Error 128' is because I wasn't getting it. I am running 'Powerleap' (as a processor speed meter) in 2K but not XP.

4). Yes I am sure I have PC100 memory installed and have not yet tried disabling the onboard memory. I will try this.

5). I have tried all the processors 'on offer' in XP to no apparent avail.

OK.. well there's a first time for everything, so I look forward to trying the '600E' processor mod on the 600X.. I guess it can't actually damage anything?

Thanks

Robin.

[pkiff]

2). [...]I was aware that XP contains its 'own' Speedstep but loaded the Intel version as I felt that it would not conflict with XP but simply serve to give me some control over things.. do they 'clash'?

My impression is that they do clash a bit in XP. I would recommend instead installing Christian Diefer's SpeedSwitchXP for use in XP instead of the Intel applet. However, that is only recommended on machines that actually support SpeedStep. For a non-SpeedStep 600X, I wouldn't recommend installing any SpeedStep applet, under Win2K or XP. You should try and use some other method of throttling your speed. I would expect the SpeedStep applets to cause conflicts and possible freezes.

3). The reason I didn't mention 'Error 128' is because I wasn't getting it. I am running 'Powerleap' (as a processor speed meter) in 2K but not XP.

My mistake. I forgot that your 600X shipped with a PIII to begin with -- so it should not generate a 127 Error [corrected from 128] the way a machine upgraded from a PII would. -- and same for the memory issue -- I forgot that you would have started with PC100 to begin with.

Phil.

[Robin Mutoid]

Quick update regarding point (5). above..

Yes, you are correct regarding the XP processor driver, allthough all three on offer:

ie. 1). 'Intel PentiumIII Processor'
2). 'Intel Processor'
3). 'Processor'

register in 'Device Manager' as 'Intel(r) Pentium(r)III processor'
the latter two allow glitch-free booting. The first is very 'sticky' causing constant freezing. I've settled for 'Processor'!

Could there however be a BIOS 'APCI' Hex edit solution to allowing a totally clean boot?

Thanks

Robin.

[pkiff]

1). 'Intel PentiumIII Processor'
2). 'Intel Processor'
3). 'Processor'
register in 'Device Manager' as 'Intel(r) Pentium(r)III processor'
the latter two allow glitch-free booting. The first is very 'sticky' causing constant freezing. I've settled for 'Processor'!

Interesting. This processor driver business is very strange. Up until seeing beatoem's post on Wim's BIOS (IBM TP600X SPEEDSTEP AND XP = MICROSOFT BLUNDER WITH CPU) and serverbook's post here in the Thinkpads.com Forum (600x (non-SpeedStep) with 850mmc-2), I hadn't realized that there even were different drivers available for a processor. In my case, I used the second one (Intel Processor -- intelppm.sys) to resolve a SpeedStep issue in my 600X (SpeedStep Issues with XP on 600X Upgraded to PIII 850MHz). But I found that this driver was not necessary on an almost identical 600X that I'd upgraded to a PIII 750MHz. So the whole thing seems very random and unpredictable now to me.

Phil.

[Robin Mutoid]

Yes.. it is interesting (or perhaps frustrating is a better word!), the apparent 'random' factor here!

I have had no problems running the speedstep board @ 850Mhz with iether 2K or XP, but it seems that the non-speedstep board can be coaxed into running at high speed with both the 'Fn+F2' and the alternative XP processor driver 'tricks'.

I wonder what discussions went on between M$oft, Intel and IBM at the time that XP was being written!?

The next move is to see if the 600E processor board 'mod' removes the necessity to 'Fn+F2' @ every boot on non-speedstep 600X's?

Cheers

Robin.

[pkiff]

The next move is to see if the 600E processor board 'mod' removes the necessity to 'Fn+F2' @ every boot on non-speedstep 600X's?

I really can't say about that one...proceed at your own risk!

Phil.