packman

Has anyone here who uses Win2K or WinXP recently invested in Drive Image 7? If so, how reliable is it, particularly on backup to/restore from CD-RWs and another partition? Am vaguely thinking of getting it myself. One might assume that a backup application like that would be 100% reliable. Well, I used to have Drive Image 5 and it worked properly only about 50% of the time. On restore, it would delete the original partition (as it should) but then fail just before completion. Despite error messages, I never did discover the exact problem. I think DI5 was just inherently buggy. Perhaps, with DI7, the bugs have been ironed out now, though? Anyone?

Really, Sapiens74? I was reading a magazine review of Drive Image 7 today and it said precisely the opposite. Indeed, it's claimed that, inside Windows, you can back up the boot partition. The review included a successful result from imaging (into another partition) a primary partition, under Win2K.

Ugh! Ross, are you seriously saying you're not using any form of thermal paste or thermal pad between the base of the heatsink and the CPU? Surely not! You cannot expect the heatsink to efficiently conduct away the heat from the CPU unless it's helped with a film of heat-conducting paste or similar. There'll be minor air gaps between the heatsink and CPU. If there's no paste or pad there, you run the risk of your CPU overheating. That said, the quiescent temperature you've quoted, and the higher temperature, are all very reasonable. How are you measuring those temperatures? Those given in your System BIOS are probably the most accurate. Don't rely on utilities like SiSoft Sandra; the figures can be wildly out. The maximum temperature tolerated by a P4 is in excess of 75C but around that region the CPU might well shut down. It's best to maintain the temperature below 60C (obviously, the lower the better), as the useful life of a P4 is said to be directly proportional to the temperature at which it's operated. My 1.8GHz P4 is cooled by a Zalman low-noise ducted cooler (EMPLOYING THERMAL PASTE) and even in the height of the recent summer, the CPU temperature never rose beyond 53C.

4 & 20, What sort of backups do you do? Contents of bootable partition to CDs, or to a separate physical drive, or to another partition on the same drive? Which OS do you use? Has DI6 or DI7 ever failed to complete a backup, or to complete a restore? In other words, has DI6 or DI7 ever been put to the ultimate test by you? In DI5, the backup process to CDs used a CD writing utility built into the application. This was completely outside the control of the user. I think much earlier versions of DI didn't have that and instead you had to separately envoke a third-party CD writing program. So, what's the situation with DI7? Does DI7 assume that the user already has a third-party writer, such as Nero, or does it use its own built-in writer?

I've suddenly hit a problem of my Win2K machine refusing to boot up when powered on. I just get a l-o-n-g beep, and sometimes a short one at the end of that long one. For a properly-working machine, it should be just one short beep. It doesn't do it all the time. When powered on from cool, it refuses but when warmed up, it boots okay. I suspect it's a power supply problem of some sort. The BIOS shows all voltages as good, though (well within the 10% tolerance and the CPU core voltage is spot on) and I've metered the +5v and +12v and they're spot on. I've also checked out the cooling fans and they're fine. Could be a problem with the 5v standby, though, or the -12v, or even the derived 3.3v. Can't easily get to those, to meter them, though, when it refuses to boot. I've not recently changed anything, hardware-wise. So, can someone remind me of the beep codes associated with P4 machines? Are the beep codes standard, and can someone list them for me? Am I correct in thinking that a long beep implies a PSU problem?

Ali, this is an old posting. The problem was caused simply by the memory module having partially popped out of its slot. Pushing it back in cured it.

With the greatest of respect, Ali, your suggestions are a diversion. The fans run on a 12v supply, not the 3.3v supply used for memory modules and suchlike. As for looking in the BIOS at the various supply voltages and fan speeds, well yes, under different circumstances or ones where that troublesome fan is clearly running, that would be the thing to do. But, if the CPU fan has stopped or is running very slowly, you certainly wouldn't want to power the PC on at all, because that will certainly stress the CPU, if not destroy it within seconds! Furthermore, looking at the 12v in the BIOS can be misleading when a fan is failing. I know this from personal experience. The 12v supply can momentarily dip, giving the impression that it's that that's at fault, whereas it's simply the fault in the fan dragging the voltage down. However, the 12v supply is normally well regulated and isn't effected in this way unless the fault puts virtually a dead short across it. If that were the case, though, the PC would be malfunctoning all over the place. Almost all the supply voltages in a PC have to be within +/-10% of the nominal value. Study the detailed data sheets of the various components. Outside that tolerance, proper operation cannot be guaranteed. Where nominal 12v fans are concerned, they'll normally continue to operate on a voltage from around 15v down to about 7v, though the rotational speed will change with the voltage. This is exploited in low-noise systems (my own, for example), where the fan's speed is reduced with the use of a series resistor or a variable resistor. Watch out for old BIOSs which are unable to monitor slow rotational speeds properly (less than about 2000 rpm).

This is very unlikely to be a PSU problem, as an 80mm fan (for example) will draw less than 100mA from the PSU. Also, if your +12v was well down, other devices in your PC wouldn't work either, including possibly your hard drive. I know from personal experience that fans are difficult to de-bug in situ. And when it's the CPU fan, it's doubly-difficult, as the CPU will shut down as soon as it begins to overheat - at least Pentiums do! Like anything else, fans don't last for ever, especially those running at high speed and which, being over the top of the CPU, get hot. My advice would be to replace the fan immediately with one of the same specification (voltage, speed, diameter), making sure the new one is a ball-bearing type. Sometimes, manufacturers use sleeve-bearing types but they're not so good in hot airstreams and won't last anything like as long as ball-bearing types. You might want to replace the entire cooler, ie heatsink included, in which case a visit to somewhere like www.quietpc.com would be useful. If it's an 80mm fan that's used and you have difficulty in getting a suitable replacement, I could give you the model nos. of two brands that are worth considering. The normal running speed for adequate cooling needs to be known, though.

Can't see what your problem is. Win2K Pro's Desktop is almost infinitely customisable and it's all there, already in Win2K. Go to Display and open up the Appearance tab. In there, you can select icon sizes, font sizes, parts of windows sizes, colours, etc., etc. Perhaps your idea of customising is different to mine, though.

I think the problem's solved. At least, I'm keeping my fingers crossed about it. Checked out the main DC connector block to the m/board this morning. Also, I temporarily removed the one memory module I've got. No sign of any oxidation of its contacts or of any undue dust around. Wiped some Kontakt66 over the contacts and pressed the module back in. Re-connected all the external cables, pressed the power-on button and ...............it booted! No long beep. Used the PC for about 2 hrs, then closed Windows down. Returned to the machine a few minutes ago and powered on again. .....................no long bleep! It booted okay. I can't be absolutely certain that re-seating the memory module's done the trick. Only the next day or so will tell. But it looks as though that was indeed the problem. In anticipation of its confirmation, thanks everyone for your help. Incidentally, the label on the BIOS chip says "Phoenix BIOS".

Thanks for that link, ThatSteveGuy. Bit difficult to tell these days if it's Award or Phoenix. It says both on the actual bootup sequence and in the m/board manual. Didn't Phoenix take over Award some time ago? Anyway, have looked at that link and I think I have to assume that mine's Award, in which case it points to a memory problem. Looks like I'm going to have to disconnect all the external cables, take the machine into the other room to have a good look at the m/board and its components, in particular the memory module. It could, of course, be the 3.3v supply to the memory module that's at fault. A loss of the 3.3v would affect the AGP graphics card as well, since I know that my graphics card specifically uses 3.3v.

Well, I don't get the long beep repeating itself. And I don't get three short ones after, either. Next time I power on from cold and get this problem, I'll listen a lot more carefully and, assuming the beast finally boots up, I'll report back here. In the meantime, thanks for your suggestions. Keep 'em coming.

Another thing - when this problem occurs, you can hear that the hard disk isn't booting, though the longer I delay turning off, I hear the hard disk doing some short bursts of activity. I don't think it's anything to do with the graphics card, unless there's a dodgy video connector problem somewhere, as the hard disk doesn't even attempt to boot. I'd imagine that if the graphics card failed (it's been perfect for the three years I've had it), the hard disk would just continue to boot.

I forgot to mention that the BIOS is Phoenix-Award. My m/board manual doesn't give any info on the beep codes. Can't imagine that it's the memory card giving trouble. Nothing's been touched in that area for quite some time and the machine's run okay for a year or so now, since I first built it. When it refuses to boot I get absolutely zilch on the screen, not even the memory check sequence. Can't even get into the BIOS. But if I keep powering off and then on again, after 4 or 5 tries, it'll finally boot. Obviously, an intermittent fault. I've paid high attention to the venting and cooling. Also, I've incorporated dust filters and consequently the interior keeps quite clean.

Make sure, in your System BIOS, that resources are assigned by the BIOS. It'll be under 'Resources Controlled By' or something similar and normally you should set it to Auto. Definitely DON'T set it so that Windows takes control of the IRQs. I'd definitely install Service Pack 3 as well. SP4 is available but is regarded as quite buggy and it might be best to hang on for a few months before installing that (if ever). Anyway, under an ACPI-compliant system, a number of IRQs may well end up shared, if there are insufficient of them to go round. And that's usually the case. Sharing is not wrong and normally works OK. I've certainly had my soundcard shared with my AGP graphics card before, with no problem. At various times, on different machines, I've had up to four devices on one IRQ. (I use Win2K). You won't alter the net sharing but it might help if you moved your soundcard to a different PCI slot. Don't use slots 1 and 2 next to the AGP. Put the soundcard in slot 3, for instance, or a slot even further down (if you've got that many). You'll still end up with some sharing but it'll be with different devices, and perhaps will be enough to avoid the problem you're experiencing. Are you sure, anyway, that your problem is actually down to IRQs? It might be caused by something else, instead.

Don't forget any operating system service pack that you might need to install. For XP, SP1 is, I believe, recommended, but not SP1a (buggy). It's best to apply the service pack straight after installing Windows. Then install the chipset drivers. Go with the latest ones, unless there's some special reason why not to. Then your graphics driver. Then the rest of your software. Well done. You can rename and re-format individual partitions, using Disk Management, which you'll find somewhere in Computer Administration (or a similar name) in the Control Panel. Disk Management is a very useful section, I find. As I and others have pointed out earlier, it's best to use Windows Setup to do the HD formatting. Beyond that, I would also recommend that, in Setup, you format to just two partitions initially - C: plus an Extended. Then use Disk Management, once Windows is there, to split the Extended and create your other partitions. Also, at this point, in Disk Management, it's best to re-assign your CD drives to driveletters that are well away from C, D or E. Otherwise, you can end up with the driveletters not optimumly assigned. Still, you can correct it again, in Disk Management.

Yup, I agree with all of that. Normally, when I use this method to re-format, I deliberately leave about 300MB unassigned, ie unformatted. I do that because I read in one of my applications (Drive Image, was it?) that it's advisable to do that to enable some hard disk admin to automatically take place when necessary. By today's standards, 300MB is very little wastage. Anyway, good luck with it, Sway.

No, why do that if instead you want NTFS? It's best to format to NTFS in the straightforward way. If instead you format to FAT32 and then use the Convert command, you introduce overheads which will actually slow down your file searching on the hard drive. Forget about Convert. And the only reason you'll want to put the Win98 disc in the drive is to give WinXP Setup a sniff of it. The Setup will tell you when to put the Win98 disk in. It's just a quick check by the Setup that it's a valid update. None of Win98's files get loaded. I think possibly in your mind you've got the order of things wrong. Boot with the XP installation CD and you'll probably find that, in the Setup sequence for installing WinXP, there's a HD formatting utility. Other contributors here with experience of WinXP will probably confirm that. You'll come to the utility naturally, as you go through the Setup for WinXP. If it's there, you'll find that a lot easier to use than fdisk. Once in the partitioning utility, make an active bootable C: partition in NTFS, then leave the rest as an extended NTFS partition. Once Windows is all installed and running, you can further split up the extended partition in Windows (in the Disk Management section). It's a doddle to do it there. Basically, just forget about FAT32 and the Convert command.

I accept that you've got what appears to be a genuine problem but why, when you fdisked and re-did everything, did you re-format to FAT32 again and then use Convert? Surely, there's no need to re-format to FAT32 again? Using Convert will slow down your system, anyway. I'm not an XP person myself but can't you just run the WinXP installation Setup and do the formatting, straight to NTFS, in that? Doesn't WinXP's Setup allow full and easy control of formatting? Win2K's certainly does. You'd run the XP installation disc and, at an early stage, it would ask you to momentarily insert your Win98 disc, just to check validity of the upgrade and then to allow it to proceed.

I had these Windows Messenger Service pop-ups occurring. They were incredibly annoying. Assuming you're using Win2K or WinXP, go to Administrative Tools in the Control Panel. Then go to Services. Click and highlight 'Messenger'. Right-click and select Properties. Set 'Startup type' to Disabled. Press the Stop button. Exit Services and Admin Tools. I found that you had to DISABLE Startup type, rather than just setting it to Manual.

I've been experimenting with changing a few cards around in my main P4 machine. Result? Well, not unexpectedly, the net difference it's made is very little indeed. I've still two examples of where two services or devices share an IRQ. For instance, IRQ16 is used by both my AGP graphics card and the soundcard. Prior to this and prior to adding the video capture card, the soundcard had its own IRQ, as did the graphics card. On the face of it, it looks a bit risky to have my graphics card and soundcard sharing the same IRQ. However, I had that situation, and more, with the other machine and that machine was pretty stable. So, the answer seems to be that, if you ain't got enough IRQs for the devices and services installed, there ain't much you can do about it, on an ACPI-compliant machine. The only real value in swapping cards around in the PCI slots, it seems to me, is in ensuring there's adequate clearance between cards, for cooling and for avoiding undue electrical pickup between the cards. Swapping 'em around won't necessarily throw up an extra IRQ from somewhere, though it might combine devices that are less likely to conflict. Although the use of IRQ9 in Win2K for multiple sharing is often seen, PCs are sufficiently diverse that, really, no two machines are the same in the way that they allocate resources and hence IRQs. I think the method should be that you look at what the BIOS sorts out for you and, if you're not satisfied with the result, you should experiment a little. In a more ideal world, we should have at least 30 separate IRQs at our disposal.

I don't know whether this helps but, on my old Pentium 3 machine (which uses Win2K as well), four services/devices share IRQ9: MS ACPI-Compliant System Creative Platinum soundcard AGP graphics card USB Host Controller The P3 machine is quite stable. It's not for ever bombing out or freezing or crashing or whatever. The BIOS and the operating system seem to cope okay with allocating resources to these devices/services. The P3 machine, being older, allows for just 16 IRQs altogether. Hence the sharing of IRQ9. With my current P4 machine, there's much the same AGP and PCI devices being used but I've added also a PCI video capture card. The BIOS and chipset, in this instance, allow for 24 IRQs, but four of those alone are taken up by USB 2.0 channels, so I still end up with IRQ16 and IRQ17 being shared.

One of the problems that the convenience of USB technology has brought us is that for every USB port (or pair of ports) you have, there's usually a corresponding IRQ (I'm assuming here that the machine has been set up as an ACPI device, not a Standard PC [HAL configuring]). Normally, after the standard services and peripherals have been added to the machine, you're left with just two IRQs to play with. Thus, IRQs get used up very quickly. And because USB ports aren't themselves permanently uninstallable, you can't release their IRQs for use by other devices. You can disable a USB port but you can't uninstall it, because it'll simply get redetected the next time you boot [PnP]. (I've just been trying this on my own machine). Disabling a port or a PCI card will show in Device Manager as a red cross. In Computer Management/Hardware/IRQs, the device and its IRQ will disappear from the listing - but will still exist! Thus, no reallocation of IRQs will occur and you might still observe sharing. From experience with my own machine, re-positioning cards in the PCI bus can produce some swapping around of IRQs but it might not actually allow for any extra ones to be used. I'll keep you posted on that, as I'm still experimenting with mine at the moment. Do remember that, if you uninstall or re-position a piece of hardware, you need to tell your BIOS about it, by entering the BIOS and setting Reset Configuration Data to Enabled and then booting into Windows again. You usually find that setting under PCI/PnP Configurations. All in all, the PCI bus and the automatic allocation of IRQs is a bit of a pig's ear. It makes a nonsense of the provision of six PCI slots, plus one AGP slot as, in practice, you find you can only use some of those PCI slots if you're prepared to accept IRQ sharing. Whilst on many systems, IRQ sharing might not be a problem, it'll depend (as I think the previous contributor stated) a good deal on what kinds of devices they are and whether at times they'll be used simultaneously. The manufacturer of my motherboard, as with others, doesn't recommend changing the machine from an ACPI device to a Standard PC and assigning IRQs manually, as subsequently allocating resources can apparently be very problematic. It seems you have to reinstall the operating system, anyway (though don't quote me on that).

There's nothing wrong with that, mate. With an ACPI-enabled computer, Win2K (or rather the BIOS) does often allocate several devices to just one IRQ. Resources are then handled dynamically. On my previous Win2K machine, I too had several devices on IRQ9 and never had a problem with it. On my current machine, it being another Win2K machine, IRQs extend up to IRQ23 and I've got two devices on IRQ16 and two devices on IRQ19. If you're really unlucky, you'll sometimes find that a particular device won't work in a shared environment (soundcards can sometimes be touchy). A lot depends on how often shared devices are actually in use.

What I mean is Is the installation of a USB 2.0 driver for Win2K radically different to that for WinXP or WinME? Here's the problem. I use a Win2K PC based around a QDI motherboard and when I built and then set the machine up I installed QDI's USB 2.0 driver for Win2K. Thereafter, I've been using my USB scanner quite successfully on the machine. However, I recently bought a USB flash drive but it won't function properly on my machine. It'll read and write but the flash icon in the Systray and other entries don't disappear when the flash is unplugged. Also, Windows hangs when I subsequently try to shut Windows down. I've another Win2K machine which has a USB 1.1 interface on it instead and the flash drive works perfectly in that. So, why doesn't the flash drive work properly on my main machine? Any ideas? Is there a special problem with USB 2.0 and Win2K, in the context of USB mass storage devices? The article at: http://support.microsoft.com/default.aspx?scid=kb;en-us;819332 would seem to suggest that there IS something special in the way that USB 2.0 drivers should be installed on Win2K machines. There, it seems to be saying that both the USB 2.0 update in SP4 and a third-party USB 2.0 driver need to be installed on any given machine, for that interface to work properly. This seems unnecessarily complex and precarious. In fact, the impression given is that the USB 2.0 driver is two-part and is perhaps chipset dependent. Anyone an expert in this domain?