cant get both cpus to work!

Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

cant get both cpus to work!

Post by Thrakath »

Hi!
I got big problems running two celerons 500mhz on the bp6 i purchased on ebay. flashed the newest bios with 1.30b Highpoint and it still recognizes in bios and in windows just one of these medocino cpus. tried both cpus in single mode in the socket closer to the keyboard connectors and they work. didnt test the second scoekt yet...is it possbile that i have a mosfet problem? where can i get new ones in germany?
or is it just to a wrong jumper setting oder bios setting? maybe i just didnt see "enable dual cpu" or something like that...

Greets from Germany
Tobi
KliK
Posts: 350
Joined: Fri Mar 11, 2005 3:56 pm
Location: Zagreb, Croatia
Contact:

Post by KliK »

both CPUs shopuld work on your BP6...might be that the CPU socket 0 doesn't work...so it might me that you have flaud board!!

both CPUs are allways on...please inspect your BP6 for any scrathes or mechanical problems...also contact the guy who sold you the faulty exipment?!
davd_bob
Confused
Posts: 1043
Joined: Fri Feb 13, 2004 2:30 am
Location: Houston, TX

Post by davd_bob »

Thrakath,
First let me say welcome to the board.

Next...take a look at your capacitors bofore proceding. They should be standing straight, no bulging and no brown goo on top or bottom.

If you are running 500s you should boot at 66FSB and see if the system shows them on the POST screen. If both cpus are in and both work and both sockets work, they will show up on the post screen.
(after counting the ram it will show 2cpus).

There are no jumpers or bois settings anywhere that "enable" dual cpu or anything like that.

Also if both are working you can look in the bios. When you enter just choose the easytune II selection where the FSB and voltage can be changed. If only one cpu is working you will see only one voltage setting you can change. With 500s dont bother trying to overclock until you have a stable system at 66FSB.
There are *almost* no bad BP6s. There are mostly bad caps.

No BP6s remaining
Athlon 2800
Sempron 2000
ViaCPU laptop with Vista.(Works great after bumping ram to 2Gig)
P-III 850@100
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

Bios shows just one CPU, also if 2 CPUs are in Sockets. Another problem i can see in Bios is the Core Voltage of the CPUs....its around 0.80V :-(. It must be a sensor or reading problem...just can´t imagine that these 2.00V CPUs would boot in Windows with this voltage but the one cpu does.
The capacitors are arced...maybe thats the reason for the low voltage reading and the disappearance of the second cpu.

Got two damaged Asus A7V Board with 6,3V and 1500µF. Should i replace the bp6s cap. with these ones?

Thx for advice
Tobi
KliK
Posts: 350
Joined: Fri Mar 11, 2005 3:56 pm
Location: Zagreb, Croatia
Contact:

Post by KliK »

Thrakath,
you are better of with the new ones! go to the shop and get 1500µF capacitors...can be the bigger ones withz 10V, no diiference! :D :D they don't cost much, sthg like 10cents a peice...

replace all the bulging ones, and all that haven't got round belly (either inside or outside) & twist to the side... :D :D

then check the BIOS and write us again... :lol:
jaybird
Posts: 301
Joined: Tue Jul 30, 2002 9:21 pm

Post by jaybird »

Word of warning, do not get just any caps, need to be 8mm in dia. or you may have a problem with the CPU heat sink fitting over theZIF socket!

I am using the largest heat sinks that will fit on the BP6, GlobalWIN FEP-32's
and I have to gently bend many of the caps off to the side for my FEP-32's to fit on the Celeron's.

Regards,

jaybird
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

thanks for the hint...but i use the intel stock coolers. the are loud but good! i wrote the ebay seller first, before i will replace the caps!

thx Tobi
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

progress made...replaced all arced 1500µF Caps (7) and 1000µF (4) that were arced...by soldering around on the board i discovered a broken coil next to the 4 1000µF Caps...i replaced it with another one from another board..it had 7 turns...the new one 6...whats the function of this coil? the board still runs but still with just one cpu!

suggestions?

Greetz Tobi
davd_bob
Confused
Posts: 1043
Joined: Fri Feb 13, 2004 2:30 am
Location: Houston, TX

Post by davd_bob »

Thrakath,
Check out these links.

This one for the coil.
http://www.bp6.com/board/viewtopic.php? ... =wind+core

This one for the capacitors.
http://www.bp6.com/board/viewtopic.php? ... highlight=
There are *almost* no bad BP6s. There are mostly bad caps.

No BP6s remaining
Athlon 2800
Sempron 2000
ViaCPU laptop with Vista.(Works great after bumping ram to 2Gig)
P-III 850@100
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

In the coil link all pictures are gone...what happens if i use another coil? i soldered a 6 turn coil from an slot1 board in. but the second cpu still isnt recognized. I replaced the 4 caps near the exchanged coil too...why or when is it equal to use use 10V caps instead of 6.3? Is there a separation of the power circuits for the two sockets?
I soldered the Caps on 2-3cm long wires hope this is no problem...much easier to handle...

What can i do now? I´ll check all soldering points again when i come back from work!

Thx Tobi
KliK
Posts: 350
Joined: Fri Mar 11, 2005 3:56 pm
Location: Zagreb, Croatia
Contact:

Post by KliK »

post some pics...just to see how it worked out...

the voltage on the caps doesn't mather...the 10V are a little harder to charge, but that is inconsiderable...the 15V might work, but 20 or 30V hardly!!!
all that mathers is the capacity of the caps that should be equal to the ones you have had earlier on the board...

also, did you change the EC10 cap?! it often gives problems on BP6...it's between the sockets, on the edge of the board!! but read the articles about that on homepage of BP6.com!!
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

Last night i replaced the ec 100µF Cap with a 6.3V 1500µF one as described on one of the websites posted...no result...one of the vcores is still locked to 0.80V. The other one is good at 2.00V. I played a bit around with my voltmeter an measured the caps between the sockets and the side which is up whent its biuld into a tower. the three of the six at the first socket are showing correct voltage of 2.00V. the other ones the 0.8V. it is possible to short circuit them? to connect the via a wire to supply both cpus with right voltage? Or are the 0.8V maybe a problem of the soft menu? i tried to raise the core volltage for both cpus to 2.3V (pretty much i know ;-)) but they stayed at 2.0V and 0.8V. this shouldnt be like this if both circuits or just the one with 2.0V work correct?!

Further suggestions?
Tobi

PS: Unable to take the flash of my cam makes anything unseeable, else its to dark...
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

Duoble post sry...connection is slow...
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

check the blue thingy (heat ?) in the middle of the socket.
I have one that is broken in two, and the board only works single cpu....
might be something in or around the second socket that has kaputted...
bp6's 3 x dual @ 533
. . . . 1 x dual @ 466
. . . . 1 sngl @ 400
[( 2 x dual xeon 2.4ghz )]
[( 2 x dual xeon 2.66ghz )]
[( 1 x 2.4C ghz )]
[( 1 x 2.4B ghz )]
[( 1 x dual AMD 1800MP )]
[( 1 x P4 1600 )]
[( 1 x 500 ppga )]
3 x piii 866
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

i cant find any other damaged caps, regulators or coils...but the pcb is burnt (black smoke) where the broken coil was. i cant solder the new one correct in because one of the holes cant be desoldered...so i cant measure any voltage at this coil...but at the other one that is ( so i think) is responsible there is also no voltage detectable. any idea about short circuiting between the correct voltage giving caps and those ones with 0.8V?

Tobi
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

no more hints?
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Thrakath wrote:i cant find any other damaged caps, regulators or coils...but the pcb is burnt (black smoke) where the broken coil was. i cant solder the new one correct in because one of the holes cant be desoldered...so i cant measure any voltage at this coil...but at the other one that is ( so i think) is responsible there is also no voltage detectable. any idea about short circuiting between the correct voltage giving caps and those ones with 0.8V?

Tobi
A few notes....
1. Are you sure that the 0.8V voltage you are seeing isn't your Vtt voltage? Where are you reading your voltages?

2. Do not short circuit between two voltage points. You will probably end up overloading components on your board which will render them useless, or you might simply destroy your board.

3. When you say you can't find any other damaged caps, is that by look only? Chances are if your board contained faulty caps then even those that "look" fine are not healthy either. I just changed caps on a BP6 that didn't have s single bulge or leak - but it was unstable. The result is that I now have a stable board. I would replace the ones that look fine as well, especially the ones >= 1000µF.

4. The coil problem is worrisome. Are you sure that the burnt part of the board hasn't destroyed anything else? A coil is usually used as an inductance (the opposite of capacitance). The idea is to try to resist changes in current that flows through the coil. The thick wired coils found on the BP6 carry a high current so it is very important that you try to replace the coil with something that is similar in diameter to the one that you are replacing. You should be fine with 6 windings instead of 7 though. If this component is somehow disabled on your board (it sounds like it from your post) then I would consider that serious trouble. This isn't a component that can be removed from the board.

As for removing solder in a hole, there are a few tips and tricks on how to accomplish that. First, I would put more solder in there and then I would use this:
Image to remove the solder. It is a "wire braid" (I have no idea what it is called in english) which you put between the solder and the soldering iron. When the soldering iron warms the solder and the braid then the solder crawls into the braid. I use this extensively when working on sensitive stuff like motherboards. You can buy it in all stores that carry solder irons and accessories.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

First thanx for your effort! Im still hoping to get the board to run! Im from Germany...same time zone as sweden i think :-). Today after work i will measure those voltage regulators and change the rest of >1000µF Caps.


@1. The 0.8V i can see as one of the core voltages in bios. in moment there is no hdd attached. just of fear ;-)
But i can also measure 0.8V at the 3 caps at the socket that is closer to the ram. the vtt voltage is the voltage i can meausre at the ec10 cap? the ec10 cap is close around 1.50V.

I even measured the Voltage at both coils...both have 0V :-(. Is this correct?
Perhaps my new coil cant get any voltage...Where is the origin and the target of her Voltage maybe i can solder her there...
CLose to those coils are 4 Caps at each. Which Voltage should they have?
At the good coil one had 5V i remember.
How can i measure the capacity of a cap? Maybe i wont have to replace all...

also again thx for advice
Tobi
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Thrakath wrote:First thanx for your effort! Im still hoping to get the board to run! Im from Germany...same time zone as sweden i think :-). Today after work i will measure those voltage regulators and change the rest of >1000µF Caps.
No problemo. Yep, we share the same time zone.
Thrakath wrote:@1. The 0.8V i can see as one of the core voltages in bios. in moment there is no hdd attached. just of fear ;-)
But i can also measure 0.8V at the 3 caps at the socket that is closer to the ram. the vtt voltage is the voltage i can meausre at the ec10 cap? the ec10 cap is close around 1.50V.
OK. It seems like your Vtt voltage is correct then. This means that the voltage regulators for your CPU's are messed up somehow. It could be related to your coil issue.

Thrakath wrote:I even measured the Voltage at both coils...both have 0V :-(. Is this correct?
I don't know actually but it sounds really strange. There should be a voltage there for sure.
Thrakath wrote:Perhaps my new coil cant get any voltage...Where is the origin and the target of her Voltage maybe i can solder her there...
CLose to those coils are 4 Caps at each. Which Voltage should they have?
At the good coil one had 5V i remember.
I am sorry to say that I have no idea about the voltages. My BP6 is sitting in a box as well and I would rather not have to rip it to shreds to find out :) Time is something I don't have in any significant quantities....
Thrakath wrote:How can i measure the capacity of a cap? Maybe i wont have to replace all...
There are special tools used to measure capacity. I have even seen some voltmeters that have such functionality built-in. However, do not try to measure capacity on a capacitor sitting on the motherboard since your reading will most probably fail. Other capacitors and components connected to the cap you are measuring will affect the value received. You can only measure a cap once it is off the board.

I remember a post around here somewhere where an individual had measured the caps he removed from a BP6. Even the ones that looked OK were pretty much dead and useless.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

Thx again!
U know the description of these voltage regulators? i am in my science institute and wanted to get some new caps after work. but now i could get the new regulators in one turn.

Thx Tobi
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Thrakath wrote:Thx again!
U know the description of these voltage regulators? i am in my science institute and wanted to get some new caps after work. but now i could get the new regulators in one turn.

Thx Tobi
There are a few different ones so you should really try to figure out which one is broken (if any of them really is broken) before you start going through the tough work of replacing them.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

progress: just did some soldering this evening. no results..then measured voltages at the mosfets that should after yoshihiros site be respnsible for vcore1 that is 0.8V on my board. the one closer to the 4 caps shows 10V. the other one closer to the socket 0.8V. But now: fpr vcore2 that is correct in bios 2.0V the closer to the caps one shows 0.16V and the other one 2V. there is also a combination where three of those mosfets (all exclusive the 10V one) show around 0.18V. Maybe the 10V one is damaged. i found out that they are 30V/64A. I have some 30V/80A from another board...can i replace them?
I had also some strange behaviour: by measuring the mosfets i short circuited the board. it wont restart until i detached an reattached the atxpower. after that both cpus were recocknized with 2.00V WOW! so i switched off and attached hdd and NIC to start my server. Just placed the board in right position..startet up: back to 0.8V...MAybe all this is just a bios error? got the newest one with the hipoit 1.30b controller. Is there one more stable?perhaps flash procedure?
Or just an interrupting soldering point?

New ideas?

Tobi
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Thrakath wrote:progress: just did some soldering this evening. no results..then measured voltages at the mosfets that should after yoshihiros site be respnsible for vcore1 that is 0.8V on my board. the one closer to the 4 caps shows 10V. the other one closer to the socket 0.8V. But now: fpr vcore2 that is correct in bios 2.0V the closer to the caps one shows 0.16V and the other one 2V. there is also a combination where three of those mosfets (all exclusive the 10V one) show around 0.18V. Maybe the 10V one is damaged. i found out that they are 30V/64A. I have some 30V/80A from another board...can i replace them?
I had also some strange behaviour: by measuring the mosfets i short circuited the board. it wont restart until i detached an reattached the atxpower. after that both cpus were recocknized with 2.00V WOW! so i switched off and attached hdd and NIC to start my server. Just placed the board in right position..startet up: back to 0.8V...MAybe all this is just a bios error? got the newest one with the hipoit 1.30b controller. Is there one more stable?perhaps flash procedure?
Or just an interrupting soldering point?
Tobi
Are you seeing the 0.8V in BIOS? Is the board up and running with two CPU's regardless of the 0.8V? If that is the case then you might be experiencing problems with the chip that reads these values. I don't think the BIOS version is responsible or else we would all be seeing these errors.

As for the mosfets, they are definitely not 30V/64A. There are no 30V voltages on the board and a 64A component would need to be much larger and need a cooler comparable with a CPU cooler to keep it in check. As for your replacement question, I have no idea. You have to check data books to see if they are pin compatable, but even if they are pin compatable I would find it highly unlikely that they were a drop-in replacement. A voltage regulator uses external components to define output voltages that they regulate which makes a simple switch between two dissimilar components a risky thing if you don't know what you are doing.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Thrakath
Posts: 25
Joined: Thu Feb 09, 2006 4:52 pm
Contact:

Post by Thrakath »

the bios is showing the 0.8V and you can measure them at voltage regulator and the mosfets around the socket.

to the mosfets...i googled fpr the specs and found these ones are 30V/64A (L3103s). http://de.farnell.com/jsp/endecaSearch/ ... 0846&N=401
If they are not interchangeable with (30V/80A) i have to search around on my defect board collection. maybe im lucky. they are definetly on k7s5a. but i dont have one...

ill try to flash another bios these days. the last bios flash was incomplete. i was lucky that the board still runs...what is the history behind these parameter story in the bios section? followed these and had a 75% flash...

Thx Tobi
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Thrakath wrote:to the mosfets...i googled fpr the specs and found these ones are 30V/64A (L3103s). http://de.farnell.com/jsp/endecaSearch/ ... 0846&N=401
If they are not interchangeable with (30V/80A) i have to search around on my defect board collection. maybe im lucky. they are definetly on k7s5a. but i dont have one...
Hmmm. I think you are misinterpreting the data over there a bit.First, the peak continuous current output is 56A and not 64A. Second, the peak possible voltage is 30V, but not at 56A. There is a formula that says that power equals voltage times (direct) current. If 30V/56A was continuous then we would be talking about a tremendous 1680W of output power!!!!! Of course this would most probably melt your motherboard :)

If you check the data again you will see that the continuous power output that this transistor can generate is 83W (Pd value). So assume a continous current output of 56A, plug that into the formula above and you will find that the voltage the transistor can handle at that specification is 1.48V. Since this component is feeding the CPU's you can realise that they are working in the 1.8-2.2Volts range (I think that is the low and high end spec of the voltage adjustment you can make in BIOS).

Of course, at full power the transistor would need to be cooled. There are two of these feeding each CPU so they probably are set up to share the load which helps keep them cool. I should also mention that the Farnell site offers to find compatable components but none are listed.
Thrakath wrote:ill try to flash another bios these days. the last bios flash was incomplete. i was lucky that the board still runs...what is the history behind these parameter story in the bios section? followed these and had a 75% flash...
I am not sure what parameter story you are talking about. If I remember correctly then I think the BIOS flash doesn't show the "squares" that show flash progress going all the way. I think you may have a successful BIOS upgrade in any case, but it might be worth checking out again.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Post Reply