8RDA+ high Vcore shutdown fix!

[mw0ca]

I hope! Property with a measuring instrument the passage examined.

[shortcircuit]

I can confirm that this works on the 8RDA3+.

The resistor that pin 16 connects to has a different name (R119). I left it on and cut the trace instead. Previously I couldn't go much above 1.9v / 2.6ghz, but now I'm looping prime95 at 2.05v/2.8ghz with my xp3000...

[chopper]

This is a great thread thanks guys.
I thought I'd add to the debate with a problem I'm now having with my 8RDA+ 1.1 & Xp2600(333) combo.

I've recently done the vmod on this board and have also put a iceburq on the NB and Zalman on the SB. I haven't done anything with the Mosfets yet as I need some advice on that - is it ok to use the extra sinks that came with the iceburq.... and what do I stick them with??

At the same time I did the NB&SB upgrade (the vmod has been done for sometime) I upgraded to an Antec True Control 550w PSU.

Now,all of a sudden, I can't get my vcore about 1.7v without getting random shutdowns! I can only assume that this has been casued by previosuly having a bit of a dodgy PSU (unbranded 420w) and when running the vcore at 2v it wasn't actually anywhere near that or I've done something seriously wrong with the NB & SB upgrades??

I now can't prime at 180x12.5 and before on a completly standard setup I was getting 200x12.5.

Oh what fun. Any help greatly appreciated.

Chopper

[Matt McFaul]

nice to see such well respected names in here. I too am having a little problem with a 8rga. Its seems to be some sort of voltage regulation in that the vcore while set anywhere above 2.05v will always show up as 2.06v. I have pulled my hair out and switched from a vantec 520 to the new antec 550 true control just so I would have front case control......stability improved however Im still at 2.06 max vcore?
Am I missing something with the cpu? its t-b 2800+. The max I can get out of her stabily is 2,808mhz regardless of fsb verses multiplyer changes. Have I just hit the limits of this chip? Sure feels that way.


http://www.docmodgod.com
http://service.futuremark.com/compare?2k1=6534407

[Spartacus]

Originally posted by chopper
This is a great thread thanks guys.
Chopper

Sure!

Hi Chopper!

I was talking with you over on AOA the other day (I post as "Photon" over there) and I know you were about ready to do this mod. Just curious if you did it yet.

Good luck!

[chopper]

Yeah I've done it. Piece of cake although I've reseated everything and all is much better. Also re-lapped my SLK800 which seems to have helped as well.

[causticVapor]

Well, after replacing the 2400+ with a DLT3C 1700+, I was rather impressed - the chip did 2.2GHz at 1.5v. For some reason, however, above 1.65v the motherboard's overcurrent protection circuitry would shut down the board on this chip. The 2400+ did not exhibit any of these problems - I could run it up to 2.2v and the board would not shut itself off.

All I know is that the heatsinks I put on the voltage regulator were not hot even at 1.95v with the 2400+. They were also cool with the 1700+, initially.

After cutting pin 2 on the vreg IC, however, the MOSFET heatsinks were hot even with the chip running at 1.5vcore. Active cooling helps and Ive replaced the 80mm beta sl with a 120mm sunon on the MCX462+, overhanging the vreg area. That 1700+ is a sweet chip, it'll do 2.4GHz at 1.85v even if it's running at 47-50C load.

Yet another person's testament of this mod - do NOT do if you do not have the voltage regulator heatsunken and with active cooling!

[Matt McFaul]

when in doubt......slap a delta 68.9cfm bad boy on those puppies!

[Spartacus]

Originally posted by causticVapor
After cutting pin 2 on the vreg IC, however, the MOSFET heatsinks were hot even with the chip running at 1.5vcore.

Huh? Pin 2 ?

Hopefully that was a typo. Pin 2 is part of the voltage correction feedback circuit. The circuit will not function correctly with pin 2 cut.

That's strange about everything running hot at low Vcore too. I don't think cutting pin 16 would affect the overall power levels (heat levels) very much. Cutting pin 2 sure would though.

[Matt McFaul]

I think he just inverted the numbers.....I understood it though.....now if someone could find a way to increase the voltage acceptance on the chip itself...lol. I cant put more then 2.1v on my 2800+. Regardless of timings as soon as I try to go 2.125 or above I get errors up the wazoo!

[causticVapor]

Yep, I just inverted the numbers. Pin 16 was exactly what I cut... and yes, the MOSFETs run rather warm even at 1.5v (the heatsink was cool to the touch before that, even at 2.24v)

Strange phenomenon, isn't it?

[causticVapor]

Originally posted by Matt McFaul
when in doubt......slap a delta 68.9cfm bad boy on those puppies!

Hell yeah!!!

[spaceman]

w00t, add me to the list of resistor-remover-type persons. I don't have any kind of real electrical background, just got brave after reading here in the forums. After I did the vdd mod 3 times, using different fixed resistors, settling on 1.9v, (that's with a 470 ohm resistor) and did the vgpu mod on my 9700 Pro, I finally decided to try this one. I read through things a FEW times, and I was sure this was needed. In case anyone actually reads this far back into the thread, I didn't have vcore probs till AFTER the vdd mod. Anyhoo, my resistor must have been soldered to the mobo better than some, cause it was a little hard to remove. But I can use at least 2.0 vcore again, yippee.

[Pita^Norf]

I did the mod last week, and found that after the mod my rig was still cutting power at highish Vcore.

This pointed the finger to my PSU being a party to the shut downs and after putting a more powerful PSU in my problems have been cured and I think the disabling of the OVP has also helped with my OC as well.

[Matt McFaul]

has anyone done this yet on a 8rga? I did and Im wondering if you have noticed any long term effects. I didnt even test it prior to doing the mod cause I was moving up (or so I thought) from a modded 8rda+.
Inncodentally I had a chance to chat with an electronics engineer designer in a think tank at kurz instruments today. Kurz instruments makes all sorts of devices for measurement at nuclear power plants and stuff.
I asked him about my burning 8rga coil and discolored pcb and he said that all of that stuff may be rated for the voltage that the board can supply but may not be rated for the wattage that Im drawing through it. Im like oh geez....so they make a board that can supply 2.3 volts to the cpu...then you have to ovp mod it anyway...then you cant draw more then so much wattage. Damned if you do and damned if ya dont. Intel is screaming my name now. If you can push a 3.06 to 4.2 with a mere 1.85volts then why am I going through this headache? Is this like an evolutionary process that every oc'er goes through or what? Have to learn on amd then relax with intel?

[Spartacus]

Originally posted by Matt McFaul
Intel is screaming my name now. If you can push a 3.06 to 4.2 with a mere 1.85volts then why am I going through this headache? Is this like an evolutionary process that every oc'er goes through or what? Have to learn on amd then relax with intel?

My 8RDA+ met it's demise when I bumped up VDD from 1.85v to 1.88v. I was pissed but not surprised, the EPOX boards just don't seem to have much margin built into their power supply circuitry.

I didn't even think of moving to Intel when this happened. It was just a bad experience with a particular mobo. Also, in my mind, the Athlon XP is a superior processor to the Pee4.

I moved to the Abit NF7-S and I've been very happy with it. It has none of the power problems the 8RDA+ did. It makes use of the extra 4-pin power connector, power is very strong and stable (also better OC voltages than 8RDA+). My NF7-S clocks to 225FSB out of the box too. No mods or hocus-pocus (like with the 8RDA+) to make it run fast. Now my 2400+ CPU/RAM are my weak links to more speed, the NF7-S has more in it.

I would suggest the NF7-S if you're not getting anywhere with the EPOX boards.

No reason to change platforms when Athlon64 is close anyway. We will all be starting over on a new platform this fall!

[uwackme]

Anyone determine of the same thing is happening to other NForce2 boards...NF7-S, A7N8X ?

[Spartacus]

My NF7-S v2.0 has been extremely stable using all of the same components I used in my 8RDA+ system (Vcore power problems detailed in this thread).

No power problems of any kind on the NF7-S.

[Matt McFaul]

I have been runing my 8rga(non plus p.o.s.) vdd@ 2v for about a month now....and Ive been folding and playing ut2k3 with it and had no probs....i did however have the mobo cpu coil burn up pretty good. I chatted with a epox tech on another forums and he said that just because 2v(vcore) is available in the bios doesnt mean its capible of it....he compared it to having the settings for cas1 available. I held my opinion to myself at that point because that forum has been known to ban people for criticizing epox

[ns_ripper]

I suffer the same shutdown problem when i set my Vcore to 2V

Yesterday I removed that resistor but it didn't work for me. When I leave my bios, it just reaches the post procedure and shuts down. I also noticed a real gentle alarm going off. You realy have to be in a near silent room to hear it.

When that happens, I have to shut off my psu, put it back on and boot again.

weird §§§§ I suppose. Anyone who had the same problem and knows a fix for it?

thnx in advance.

edit:

are you guy's shure it's R578 we need to remove?

[Pita^Norf]

Originally posted by ns_ripper
I suffer the same shutdown problem when i set my Vcore to 2V

Yesterday I removed that resistor but it didn't work for me. When I leave my bios, it just reaches the post procedure and shuts down. I also noticed a real gentle alarm going off. You realy have to be in a near silent room to hear it.

When that happens, I have to shut off my psu, put it back on and boot again.

weird §§§§ I suppose. Anyone who had the same problem and knows a fix for it?

thnx in advance.

edit:

are you guy's shure it's R578 we need to remove?

I had a similar problem and I tracked it down to a weak PSU; the 420W ran out of watts with all the devices and OC I was running.

Got a 550W now and problems are fixed.

[ns_ripper]

I got a 460W enermax with a vapo running on it :s

today my system refused to work. When I normaly got my desktop I saw a black screen with a cursor. I solderd the resistor back on and the problem was gone :s

weird §§§§. I think I'm going to leave it like this. ( /me got scared)

[Holst]

its better to cut the leg ot lift it rather than just pull the resistor off.

[ctgilles]

Originally posted by Holst
its better to cut the leg ot lift it rather than just pull the resistor off.

that's something I don't get. Desoldering the resistor would normaly be the same as cutting the trace or lifting the pin on the IC. Since you also open the cirquit by doing that.

[ns_ripper]

sorry, that last reply of ctgilles is mine. He forgot to log out @ my pc.