Biostar M7VIG Pro

[Pentium4]

A guy I know gave me a box full of motherboards that were bad or going bad. This one just so happened to have an Athlon XP 2200+ installed in it. Only thing I could spot wrong with it was a potentially fried DDR RAM slot, and an OST on the VRM high that got so hot that the sleeve shrunk. Interesting how they used 16V caps for 5V in for the VRM. Probably just for the lower ESR. Just sitting at the BIOS, the exterior of those VRM high caps got to 63C. That's pretty warm...Besides the VRM, all the other caps on the board are GP. Was it just not necessary on older boards? They're all OST RLG. Would regular motherboard grade caps be too low for this old thing? I think this board is pretty neat. It does not have a fan controller, and the CPU fan is way too loud @ 12V. I made a custom little fan plug in for it that just gives 7.5V through a molex connector. The CPU still stays cool (especially with new thermal paste) but it's almost silent, at least compared to the 4,250RPM beforehand.

[Sparkey55]

I have its older sibling the M6VLQ Socket 370 PIII Taulatin supported. It to was loaded with OST RLS caps. Tiger Direct specials.

[momaka]

Originally Posted by Pentium4

Only thing I could spot wrong with it was a potentially fried DDR RAM slot, and an OST on the VRM high that got so hot that the sleeve shrunk. Interesting how they used 16V caps for 5V in for the VRM. Probably just for the lower ESR. Just sitting at the BIOS, the exterior of those VRM high caps got to 63C. That's pretty warm...

63C? Ouch.
Yes, they likely used the 16V caps for lower ESR. My JetWay N2PAP-Lite was like that too, except it used GSC ME caps (none of them were bad at the time of replacing, though).

By the way, this is the first BioStar motherboard I see with more than 5 caps for the VRM low side. Those empty spots look ashamed. Fill 'er up .

My vote goes for Panasonic FR of FM. Or even Chemicon KZE / Rubycon ZL will work. Save your valuable HM. HN, and HZs for something else that needs them. Those old Athlons run at 1.6V core or thereabouts, and they don't consume a whole lot of power, so it takes for quite a bit of ripple to appear on the VRM low side before things go bad. That's how they got away with using nearly general purpose crappy caps back in the days.

Originally Posted by Pentium4

Besides the VRM, all the other caps on the board are GP. Was it just not necessary on older boards? They're all OST RLG. Would regular motherboard grade caps be too low for this old thing?

Except the CPU VRM (and occasionally Northbridge on some motherboards), most stuff back in the day used linear regulators like 1084 or discrete circuit with MOSFETs. Therefore, general purpose caps worked fine for just about everything on those.

That said, there would be no harm from using very low ESR caps on these boards. In fact, even polies should be fine. I bet that burned RAM slot was a result of either somebody trying to remove the RAM while the PC was plugged in, or just those crappy OSTs.

Originally Posted by Pentium4

I think this board is pretty neat. It does not have a fan controller

Yeah, looks like a decent board indeed. After all, this is the last generation of boards that used through-hole CPU sockets and leaded solder. Should be indestructible with a recap.

Wouldn't it be cool to set up a retro gaming cafe with these old boards one day? Like, maybe 10-20 years from now? I bet they would still work like a champ.

[Pentium4]

Originally Posted by momaka

63C? Ouch.
Yes, they likely used the 16V caps for lower ESR. My JetWay N2PAP-Lite was like that too, except it used GSC ME caps (none of them were bad at the time of replacing, though).

Thanks for bumping this thread, I had forgotten to update it. I ended up recapping it I think later that day. I ended up using polymer caps for the VRM high since oddly, those caps ran much hotter than the VRM low. Maybe because of those FETs? Ouch, GSC ME? I just replaced some of those also (Well, GSC LE) on the VRM high of a VIA P4 board, and even though none of them bulged, they ALL tested bad. They were all 1500µF 16V 10x25mm. They all tested between 900-1050µF and all were above 0.5Ω ESR. What did yours read?

By the way, this is the first BioStar motherboard I see with more than 5 caps for the VRM low side. Those empty spots look ashamed. Fill 'er up

I tried! For some reason, I had trouble getting the negative holes to clear. So I left what I put in there, since their ESR is lower than the original. By the way, every single OST cap tested good. The VRM high were closest to being out of spec.

and they don't consume a whole lot of power

Really? This thing is hungry. Just sitting at the "NO BOOT DEVICE DETECTED" screen, it was pulling 72W DC. This was just with one DDR RAM module, no peripherals, and even on board graphics. I replaced it with a 1.53GHz Athlon XP since it's not much slower and runs much cooler.

Here's something interesting to note from a little experiment I did. With the exact same scenario (Just sitting at the BIOS) with the same PSU and ambient temperature, before and after the recap, the CPU idled 4C cooler afterwards.

Except the CPU VRM (and occasionally Northbridge on some motherboards), most stuff back in the day used linear regulators like 1084 or discrete circuit with MOSFETs. Therefore, general purpose caps worked fine for just about everything on those.

That said, there would be no harm from using very low ESR caps on these boards. In fact, even polies should be fine. I bet that burned RAM slot was a result of either somebody trying to remove the RAM while the PC was plugged in, or just those crappy OSTs

Very good to know, thanks for the info As you can see, I replaced all the GP caps with nichicon HM.

Yeah, looks like a decent board indeed. After all, this is the last generation of boards that used through-hole CPU sockets and leaded solder. Should be indestructible with a recap.

Wouldn't it be cool to set up a retro gaming cafe with these old boards one day? Like, maybe 10-20 years from now? I bet they would still work like a champ.

Gotta love the lead based solder. Also, that CPU fan is a Delta ball bearing. That would be pretty neat It will absolutely still be running then. I haven't decided what to with it yet, but I'd want to use it for something like an ISF. Something that doesn't need to do much but needs to be reliable. I have a pair of Barracuda IV's. Get this....One of them has 82,000 running hours on it, with no bad sectors And part of its life it was powered by a PSU with failed caps.

[Pentium4]

By the way, I used our simple little voltage trick to get the CPU fan to run much cooler, but still be adequate

[momaka]

^ 7V trick? I have that on a few "misbehaving" loud PCs as well .

Originally Posted by Pentium4

I ended up recapping it I think later that day. I ended up using polymer caps for the VRM high since oddly, those caps ran much hotter than the VRM low. Maybe because of those FETs?

Yes, it's possible. Both my Jetway N2PAP-Lite and ECS P4VXASD2+ are like that.

Originally Posted by Pentium4

Ouch, GSC ME? ... What did yours read?

ESR *appeared* in spec, but the capacity on many of them was high and lingering on the border of +20%. I think one or two were above that. That tells me they were likely becoming leaky and ready to go soon (and GO they did ). And when caps go leaky, the ESR will appear fine on an ESR meter.
That's how early Nichicon HN and HM from the defective batches seem to have failed.

Originally Posted by Pentium4

By the way, every single OST cap tested good. The VRM high were closest to being out of spec.

Interesting.
I guess this further proves that OST hate the heat.

Originally Posted by Pentium4

Really? This thing is hungry. Just sitting at the "NO BOOT DEVICE DETECTED" screen, it was pulling 72W DC. This was just with one DDR RAM module, no peripherals, and even on board graphics.

Yes, those older AMDs were never able to scale down properly. This is typical for older CPUs on large nm technology, where idle power is almost the same as under full CPU load.
Also, if I am not mistaken, on old boards like these, CPU HALT instructions aren't loaded until the OS kernel is. So what you are seeing may as well be the maximum that board will pull. And if not, the power draw under max CPU load won't be much higher than that.

I'm not sure which 2200+ you had, but here are a few:
http://www.cpu-world.com/CPUs/K7/AMD...2200DUV3C.html
http://www.cpu-world.com/CPUs/K7/AMD...2200DKV3C.html
You can see that the minimum and maximum power draw are about the same.

Originally Posted by Pentium4

I replaced it with a 1.53GHz Athlon XP since it's not much slower and runs much cooler.

Hmmm.. overclocking time then?
I have a 1.4 GHz Duron Applebred that also runs very cool (due to 1.5V core). But it only has a lame 64 kb of L2 cache .
Man, when I got that back in late 2003 and was able to play Half-Life 2, I was totally like this:

Originally Posted by Pentium4

Here's something interesting to note from a little experiment I did. With the exact same scenario (Just sitting at the BIOS) with the same PSU and ambient temperature, before and after the recap, the CPU idled 4C cooler afterwards.

Interesting. So seems like another theory is proven here - that good caps can lower your temperature a bit.

Originally Posted by Pentium4

I have a pair of Barracuda IV's. Get this....One of them has 82,000 running hours on it, with no bad sectors And part of its life it was powered by a PSU with failed caps.

That is impressive! Barracuda IVs are amazing drives.
I have 3 of them (20 GB, 40 GB, and 60 GB). They all have a few bad sectors, but they have "settled" for a while now.

[Pentium4]

Originally Posted by momaka

^ 7V trick? I have that on a few "misbehaving" loud PCs as well .

Gah, I said "cooler" and I meant "quieter" This trick is a lifesaver. I can't stand loud PC's anymore, which is one reason I like Topower PSU's so much, they're always incredibly silent.

ESR *appeared* in spec, but the capacity on many of them was high and lingering on the border of +20%. I think one or two were above that. That tells me they were likely becoming leaky and ready to go soon (and GO they did ). And when caps go leaky, the ESR will appear fine on an ESR meter.
That's how early Nichicon HN and HM from the defective batches seem to have failed.

Ah, typical. Doesn't help that they're very old. On the board I recapped, I even replaced all the little caps. All the <470µF caps tested perfectly.

Interesting.
I guess this further proves that OST hate the heat.

It also proves that they're half decent. I guarantee you Teapo SC and SEK in the same positions would not all have been in spec.

Yes, those older AMDs were never able to scale down properly. This is typical for older CPUs on large nm technology, where idle power is almost the same as under full CPU load.
Also, if I am not mistaken, on old boards like these, CPU HALT instructions aren't loaded until the OS kernel is. So what you are seeing may as well be the maximum that board will pull. And if not, the power draw under max CPU load won't be much higher than that.

Yeah, it was kinda funny because I started a memtest on it, and the power usage only went up a couple watts. I thought these had no power saving features whatsoever?


I have this one http://www.cpu-world.com/CPUs/K7/AMD...2200DUV3C.html

Hmmm.. overclocking time then?

Maybe! The new/better caps makes this more possible.

I have a 1.4 GHz Duron Applebred that also runs very cool (due to 1.5V core). But it only has a lame 64 kb of L2 cache
Man, when I got that back in late 2003 and was able to play Half-Life 2, I was totally like this:

Yeah, poor things. That crippled them so much. I remember them being very slow. But heck, in 2003 that would be a nice proc! At that time I was using a Celeron 533/128K/66.

Interesting. So seems like another theory is proven here - that good caps can lower your temperature a bit.

I'm guessing because of the lower ESR. The OST caps were good but the ESR was starting to rise.

That is impressive! Barracuda IVs are amazing drives.
I have 3 of them (20 GB, 40 GB, and 60 GB). They all have a few bad sectors, but they have "settled" for a while now.

Yeah they are! They're both 40GB. The other one has 36,000 running hours on it and it has some bad sectors. Do you know how many hours are on your drives?

[dragos2009]

i have a system like that with athlon xp 1700+ 1gb of ram and the really great nvidia geforce 6600 belive me or not with that config i was able to play a lot games like call of duty mw3

[momaka]

Originally Posted by Pentium4

Gah, I said "cooler" and I meant "quieter" This trick is a lifesaver. I can't stand loud PC's anymore, which is one reason I like Topower PSU's so much, they're always incredibly silent.

Lol, you'll hate my computers then. Most of them are quite loud. I think the loudest I have is a 933 MHz Pentium 3 PC with two 20 GB ball bearing HDDs and a 250W Deer PSU that has it's 80 mm PowerLogic fan on full blast. With the 7V trick on the CPU fan (a 60 mm 5000 RPM screamer), it's a little better, but still terrible by any standards. The problem is one of the HDDs is kind of old now - 50k hours if I am not mistaken. So those ball bearings on the spindle in it are quite loose and loud now.

I think the second loudest PC I have is my HP NetServer with its 10k RPM Maxtor Atlas HDD. I believe that is a ball bearing design too. The CPU fan already has the 7V trick and I removed the rear 92 mm fan, which was some DC Koala (or something like that) brand that is rated for several watts and is extremely loud on 12V. PSU is made by Astec, and the 80mm fan in it is quite loud too, even though it does have a thermal controller.

Originally Posted by Pentium4

Ah, typical. Doesn't help that they're very old. On the board I recapped, I even replaced all the little caps. All the <470µF caps tested perfectly.

Yet, put those OSTs in a compact Shuttle computer, and nearly all of them will spew their guts - even the 220 uF ones. Keep them cool, and they are fine. Now I wonder why Shuttle still insists on using them. Planned obsolescence?

Originally Posted by Pentium4

Yeah, it was kinda funny because I started a memtest on it, and the power usage only went up a couple watts. I thought these had no power saving features whatsoever?

HALT instruction is supposed to lower the power usage... but these processors just don't scale down. IIRC, the K7/Athlon XP architecture doesn't support S3/S5 state either. So when you put the computer in Standby (Sleep) or Hibernate, the PSU will stay ON and the system fans will continue to run. However, the HDDs do turn turned OFF and power to the CPU is reduced down to a few Watts maybe (if even that).

Originally Posted by Pentium4

Yeah, poor things. That crippled them so much. I remember them being very slow.

The small cache only affected game performance. I could do the L2 bridge mod, though, and possibly unlock it to 128 KB or 256 KB. But I never tried it. Need to find me / fix me a socket 462 mobo (have 2 spares, but one is finicky and the other one was Bestec-ed ).

Originally Posted by Pentium4

Do you know how many hours are on your drives?

Only on the 40 GB - 15848 hours. The 20 GB might have more. The 60 GB had few IIRC. I need to update my HDD SMART logs.

Originally Posted by dragos2009

i have a system like that with athlon xp 1700+ 1gb of ram and the really great nvidia geforce 6600 belive me or not with that config i was able to play a lot games like call of duty mw3

That's amazing!
I actually did a YouTube search, and people with similar setups were indeed able to run MW3. Granted most did it on 800x600 or lower and all settings on low... but I still find that impressive.

When Half-Life 2 first came out, my PC had *only* 256 MB of RAM, so I could only run the game at 640x480 and med-low details. When I upgraded to 640 MB, it felt like I had unlimited RAM . Then it ran at 800x600 and nearly maxxed .
That's with an Asus Radeon 9200 SE (stupid edition?) 64 MB video card.
Fun days .