VOGONS


First post, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

A thread recently talking about some of the best BX440 boards brought up the Asus P3B-F. I never had this board before and really wanted to get a Slot 1 board that supported the later P3s. It took a lot of fooling around to get it to not crash from DMA in games as soon as a sound sample was played. Not 100% sure how I resolved this issue. I also had a lot of general instability until I upgraded to the final bios revision on asus' website.

In any case, every time I start up the PC it complains about the hardware monitor and I should check it out. So I did and it always says the negative 5 volt rail is at negative 6 volts. Is this OK? Everything seems to be running OK now. Is there a probe point I can test somewhere to see if it really is negative 6?

To be fair, I bought this relatively cheap PSU a while ago: http://www.newegg.com/Product/Product.aspx?It … N82E16817170016. It had -5V and I heard some older ISA stuff wanted it and this was the only new PSU I could find that had this ability. I know it's probably not a true 480 watt PSU, but I was OK with that since I'm only using a GUS, Voodoo 3 AGP, and a NIC. The CPU is a P3 800mhz/100 FSB and I have 512MB RAM. So I don't think I'm overloading it. I tried putting a P2 400 in there to see if the voltage level changed and it didn't.

Reply 1 of 15, by Jepael

User metadata
Rank Oldbie
Rank
Oldbie
Maraakate wrote:

In any case, every time I start up the PC it complains about the hardware monitor and I should check it out. So I did and it always says the negative 5 volt rail is at negative 6 volts. Is this OK? Everything seems to be running OK now. Is there a probe point I can test somewhere to see if it really is negative 6?

Without load the voltage can be higher (or otherwise out of specs). No estimate how much higher or lower is still normal.
The measurement circuitry on the motherboard does not load it much, and it may also be inaccurate. Most likely the measurement circuitry it is built to handle some margin.

The white wire on the ATX connector, measure with a multimeter at the connector at the ATX connector, poking the multimeter (in voltage mode) between the wires to touch the metal at the end of white wire. Be careful not to short anything.

If you want, put a 1 kilo-ohm resistor as a load somewhere (isa bus, multimeter) from -5V to GND and measure if it is near -5V.

Reply 2 of 15, by brostenen

User metadata
Rank l33t++
Rank
l33t++

Measure the white line, and/or try a different PSU. If you want -5v straight, it will be hard as some inacuracy will be in present in all electronics.

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk
My YouTube: https://www.youtube.com/user/brostenen

001100 010010 011110 100001 101101 110011

Reply 3 of 15, by Jepael

User metadata
Rank Oldbie
Rank
Oldbie

Yes, the multimeter is not going to show exactly -5.000000 volts.

Usually -5V rail is made from -12V with a simple 7905 type regulator normally having +/- 5% or better tolerance, or -4.75 to -5.25 volts, but only when they have a minimum load in the order of 5mA connected, otherwise their output will go out of regulation.

ATX negative voltage tolerance specs are +/- 10%, so -4.5 to -5.5 volts.

Reply 4 of 15, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

I know it won't be dead on -5.0v. I just figure if it was reading somewhere around 4.5 to 5.5 I should be OK. I don't even know if the GUS card I'm using even uses -5V.

To check the pin on the ATX connector i just red probe that point and then probe black to ground?

Reply 7 of 15, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

Well considering nothing is using it I guess it's not worth the effort to put a load on the PSU for it. My original concern was that the -5V could be used by something and could possibly be affecting stability or hurting some component. Then I did some research and found that only a handful of ISA cards even used it (and some early FDC). Because of that most PSUs dropped -5V after a certain generation.

However, I am still interested in measuring it just for kicks to see how accurate the measurement IC is.

Reply 8 of 15, by shamino

User metadata
Rank l33t
Rank
l33t
Maraakate wrote:

A thread recently talking about some of the best BX440 boards brought up the Asus P3B-F. I never had this board before and really wanted to get a Slot 1 board that supported the later P3s. It took a lot of fooling around to get it to not crash from DMA in games as soon as a sound sample was played. Not 100% sure how I resolved this issue. I also had a lot of general instability until I upgraded to the final bios revision on asus' website.

I'm not sure if this is what you did, but you need to go in the BIOS and designate the ISA card's IRQ and DMA channels as reserved for ISA. That way Plug-n-Play won't try to use them for anything else.
I had the same problem recently when I ran WinAmp with an ISA Sound Blaster. Confused me for a while, I guess it's been too long since I've used ISA cards.

To check the pin on the ATX connector i just red probe that point and then probe black to ground?

Yup - if red is on the white wire (or a connected point) and black is on the chassis of the PSU (safe and easy ground), then the meter should read between -4.5v to -5.5v.

Reply 9 of 15, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

I know about locking the IRQ and DMA's away from PnP but that made no difference.

I toggled PnP OS -> yes/no quite a few times and resetting the ESCD data and removing various components including the RAM (had more than one stick) and at some point it started working and I slowly put back all the components and it still worked so I'm not sure what the deal was.

After I ran the BIOS update I've never seen the issue again.

I had a few thoughts on what it could possibly be:

* BIOS update mentions a fix for data corruption related to using legacy USB devices in combo with PS/2 devices (i.e. I had a USB Keyboard and a PS/2 Mouse).
* Capacitors may have been slightly weak and after being on for quite sometime they reformed and now it works OK.
* ICU issues was pointed out by a friend of mine and he sent me the DOS drivers for this, I installed these but the issue returned at one point.

It's been a two days now and I have used the PC quite a bit and it hasn't returned since so I suppose the issue is resolved.

Another issue I had during this project, which was not related... The P3 800 I bought had a very crappy aluminum heatsink which I think may be either OEM or aftermarket. It used plastic clips to hold it in place and they were somewhat loose. I removed the clips to put new thermal grease on it and clean the heatsink since it had obviously been used in a house with cigarette smoking. The plastic clips were brittle and broke so I had to run to a hardware store and fab up a bracket. Even after all this the CPU was overheating after some time. I had some extra P3 slot 1's that were 500mhz so I swapped those heatsinks over with a fan and the new bracket and it works very well. I get about 32C idle and 44C is the highest I've ever seen it. Even after running DOS games for an hour or two then immediately resetting and checking the BIOS. I know the CPU wasn't the issue because I tried using known good slot 1 processors like my P2 400 to be sure it wasn't some issue with the motherboard revision and instability with high frequency P3 slot 1s.

Reply 10 of 15, by shamino

User metadata
Rank l33t
Rank
l33t

There are a ton of slot-1 CPUs that made the rounds on eBay with really wimpy heatsinks that seem to be not much more than folded tinfoil. I don't know what they came from but they must have relied on some serious forced airflow in their original application. Decommissioned rackmount servers I would guess. Several years ago I bought a bulk lot of those and they actually had some different sizes of those "heatsinks", but they were all equally worthless.
I ended up buying another lot of CPUs just to steal the HSF assemblies from.

Reply 11 of 15, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

I have real server slot 1's. They actually have ECC on them which real server boards will identify it as such, they are quite different and the passive heatsinks for them are very large and quite heavy compared to the consumer brands. I'm not sure if pictures are available of such types of processors. I got lucky in getting this board in a trade, but unfortunately it is limited to a P3 550 as the max limit. I would love to have another true server board that allowed later P3 slot 1s but they are expensive on ebay. The board I have, QDI Brilliant IV has lasted extremely well over the years and is not picky about stuff put into it and is very easy to setup BIOS options in regards to legacy ISA configuration. It's quite honestly one of (if not the best) slot 1 motherboards I have ever used. For someone looking into a an excellent P2 era build you should find one as the dual processors with a P2 400 allow it to work very well in Windows XP since it can take advantage of these dual processors.

I can take a picture of the real server versions if someone here is interested. In any case, the real passive heat sinks do their job extremely well. Whatever was on this P3 800 was basically like you said, folded tinfoil. The fins were large and thin, like a little thicker than tin foil and I could easily bend these fins. It was amazing the processor survived at all, especially with the amount of tar residue from cigarettes.

Getting off-topic a little more, I recall years ago someone asked me to take a look at their parents computer because it was getting mysterious lock ups and the like. When I got there the place reeked of stale cigarette smoke and I asked the kids mom if she smoked a lot around the computer. I opened the case and the residue was thick that it actually made a nasty mix of dust and tar and the fan was unable to move any more in addition to plugging up the fins. After I cleaned it all up it amazingly all still worked again.

Reply 12 of 15, by MajorDan

User metadata
Rank Newbie
Rank
Newbie

This issue is actually very easy to resolve. I have several of these ASUS P3B-F mainboards and discovered a couple of them were displaying this hardware monitor error message during boot up which is very annoying as it indefinitely pauses the process of booting into your operating system until you press the F1 key to bypass this message. The fix is to go into CMOS setup by pressing the delete key during boot up and then once the setup screen appears navigate over to the Power tab and press enter then navigate downward to Hardware Monitor and press enter. Next scroll down to the -5V and then press the enter key to highlight it and choose [Ignore]. Finally press the F10 key to save your changes and then reboot the computer. You will never see this error message again.

Reply 13 of 15, by Maraakate

User metadata
Rank Oldbie
Rank
Oldbie

Hi MajorDan,

Yes, I ended up doing that in the end. However, it does have a new problem where the PS/2 ports are not working reliably any more. I have to use USB now. Not sure why, but plugging any keyboard into it will make it think a keypress is being continuously generated. It's a real shame, because otherwise the board has worked fairly well over past couple of years.

Reply 14 of 15, by meljor

User metadata
Rank Oldbie
Rank
Oldbie

check this: P3B-F PS/2 ports stopped working

asus tx97-e, 233mmx, voodoo1, s3 virge ,sb16
asus p5a, k6-3+ @ 550mhz, voodoo2 12mb sli, gf2 gts, awe32
asus p3b-f, p3-700, voodoo3 3500TV agp, awe64
asus tusl2-c, p3-S 1,4ghz, voodoo5 5500, live!
asus a7n8x DL, barton cpu, 6800ultra, Voodoo3 pci, audigy1