8rda+ vdd, vcore... power question

[TANKORR]

Hello All,

I was interested in finding out which rails the 8rda+ uses for the following - please provide source info...

vdd
vcore
vdimm
vagp

Thanks loads

[Áedán]

VCORE comes from the +5V rail. I suspect that the others do too (I vaguly remember seeing +5V around the VDD and VAGP regulators, but it was a while ago that I looked)

[TANKORR]

Originally Posted by Áedán

V[size=1]CORE comes from the +5V rail. I suspect that the others do too (I vaguly remember seeing +5V around the VDD and VAGP regulators, but it was a while ago that I looked)

Thanks - sorry for the late reply.

BTW Per Epox:

Vcore comes from +12v.
The rest are from +5v or +3.3v.

[Áedán]

It sounds like whoever you spoke to at EPoX got a little confused and thought you were using a different board. The 8RDA+ definately uses the 5V line for supplying VCORE, rather than 12V.

[TANKORR]

So much for Tech support I guess.... good help is hard to find.

[Áedán]

Having killed the 5v rectifier in one PSU, I can attest to the CPU core being fed from 5V!

[TANKORR]

lol ur HARDCORE

Someone of your backgroud might be able to assist me in another area. I have been trying to push my MB a bit and have run up against a problem - I lose the video signal to my screen and need to do a reboot when I hit the FSB of 215 with a multiplier of 10 (9.5x215 is stable) or when I keep it at 9.5 and go above 215.

The test that fails is 3D mark 2001 Se (330) - aquamark3 - sandra dont seen to mind so I am wondering why 3D does. Mu gut tells me that my 8rda+ rev 1.1 (a2) NB just can't take the strain with that FSB - which is kinda weird since if I push the process harder say 11.5x212 and it has no problems even runing 5.2.2.2 any ideas?

BTW I tried bumping the vcore up to 1.925 and 1.95 with vdimm at 2.8 and vdd at 1.84 (not adjustable)

[Áedán]

To be honest, I'm not convinced that either the VDD supply or the decoupling is that great, which is why the board has problems with higher FSBs. Well, that's my opinion though.

The NB is the meeting point between the FSB, the RAM, the southbridge and AGP, so it's under the most stress. 3D Mark would push FSB, RAM and AGP all at once, whereas Sandra would only stress one or two at a time.

It might be worth seeing what happens if you drop the AGP speed down to something low just for testing purposes.
I have yet to make a VDD mod to my board, as I want to uprate the MOSFETs providing the supply before doing so. No point killing the board because I didn't replace them!

[Lord Vetinari]

TANKORR.
Pretty certain you will find that is the Vdd limiting you- Got mine modded to 1.82V and it will do 207FSB max. By max I mean it will run all 3D apps/games- can go higher than that and be prime95/windows stable but 3D apps always crap out in a similar fashion to yours.
When I did the Vdd mod on my Mb, I checked a variety of voltages and starting at 185FSB (my unmodded max) I found I got about 11FSB increase in max FSB per 0.1V increase in Vdd.
Unfortunately I'm not happy about taking the Vdd near 1.9V to 2.OV even though I have a HS on the little chip next to the Vdd modded one with a fan blowing on it- it gets very hot. There were reports of peeps blowing this chip up with Vdd levels of 2.0V.

I suspect the only real solution to this is to get an 8RDA+ V2.2 Mb but if you are in the US you are stuck (not available).

[TANKORR]

Thanks for the replies!

@ Aedan - thanks form confirming my gut feeling - I will test at a lower AGP to see if that helps. The interesting thing is that AquaMark has no problems running and it should be stressing the NB in the same way..?


@ Lord V (what is a member of the House of Lords doing here? I would think you might be busy with the pre-war intelligence issue...)

Now back to the problem at hand.

I guess I should not feel to bad - pre-mod my MB would do a stable FSB of 195 and with the mod 212. The only reason I am trying to go a bit higher is to push my ram (XMS PC3500) since it is not working as hard as it can with the FSB under 217. (at least my ram is not a bottleneck)

As for people blowing the MB I think there is a thread about post-mod 8RDA+ MBs. If I recall correctly the people running under 2.0 have been running w/o problems in some cases almost one year.

When I did the mod I made sure to follow the advice about using sinks on all the chips and a combo of active and passive cooling on the NB & SB.


One last thing the rev 2.2 I think is available from one US supplier... but my next upgrade will be 64bit DDR2 unless Intel come out with a better chip for us OC'n types.

[malphadour]

Originally Posted by TANKORR

lol ur HARDCORE

Someone of your backgroud might be able to assist me in another area. I have been trying to push my MB a bit and have run up against a problem - I lose the video signal to my screen and need to do a reboot when I hit the FSB of 215 with a multiplier of 10 (9.5x215 is stable) or when I keep it at 9.5 and go above 215.

The test that fails is 3D mark 2001 Se (330) - aquamark3 - sandra dont seen to mind so I am wondering why 3D does. Mu gut tells me that my 8rda+ rev 1.1 (a2) NB just can't take the strain with that FSB - which is kinda weird since if I push the process harder say 11.5x212 and it has no problems even runing 5.2.2.2 any ideas?

BTW I tried bumping the vcore up to 1.925 and 1.95 with vdimm at 2.8 and vdd at 1.84 (not adjustable)

------------------------------------------------------------------

I have the identical problem, as do several others on here. It seems that some of the boards are inferior and just cant hack it. I find I can go higher if I don't use the dual channel ram slot. Does not make sense, and looks like time for a different board.

[Áedán]

Not using dual channel and getting higher performance makes sense. Dual channel requires much tighter coordination and increases the current consumption of the memory controller.

The FSB can't really take advantage of dual channel anyhow, so most benchmarks are not capable of showing the where dual channel helps.