IC7-MAX3 Fixed?!?

On another forum I was talking with a guy who just bought the MAX3 and he said his mobo had NNOOO problems with the vdimm voltages. He believes the mobo's coming off the production line right now have fixed this problem (without the enduser having to remove the MC64 cap)

Is this true? Can someone confirm this for me?

they are bringing out a 1.1 version, i dont know when or if it has already been released

I just got a MAX3 from excaliberpc.com and it came new in the box without the mc64 cap. It had obviously been removed, not manufactured without it. I was told there was a ver 1.1 , and excaliber had them, so when I ordered I specified a 1.1. They emailed back and said there were no 1.1 versions according to Abit, only "reworked" 1.0. They said all their max 3's were "reworked" to "fix" the vdimm issue.

I have a gig set of Mushkin black level 2 PC3500, and it works up to 217 OK at 2.8 volts (not 2.7, though). At the same 217, selecting 2.9 volts errors test 5 out terribly. Any voltage higher than 2.8 is unusable. Vtt is still not fixed - only vdimm. Its of no use.

- Chris

I saw that Abit had fixed the volt problems and were accepting RMAs with the old boards.

Should check out the Abit forums.

Xymurgy said:

I saw that Abit had fixed the volt problems and were accepting RMAs with the old boards.

Should check out the Abit forums.

Click to expand...

It would appear from iamjcl's post above, that vtt is still an issue and according to the Abit forum, Abit is not accepting ownership of the vtt problem.

The AI7 doesnt appear to have these issues, however it uses the 865P chipset and doesnt have the extras that the IC7 MAX3 has.

I have been waiting several months to see this resolved and it looks like it wont be so I am heading toward the P4C800-E, when New Egg gets them back in stock...

I have just written an email to the editor in chief of zif davis magazine detailing out the problem and the fact that aBit refused to fix the problem with the vTT. I hope something comes out of this. I will update with any response I get from him.

axiom said:

I have just written an email to the editor in chief of zif davis magazine detailing out the problem and the fact that aBit refused to fix the problem with the vTT. I hope something comes out of this. I will update with any response I get from him.

Click to expand...

That is great! We probably should be doing more of that in cases like this!

I had to rma my first max3 for a problem other than the vdimm. the one I have right now has stable voltage for each vdimm.

aBit just closed the sticky so no one else could post to it. They sent me a PM stating that I broke the rules by posting the review from cluboc.net and that if I posted it again or any other comments that made Abit look bad, they would ban me. Boohoo, who cares.

Well I wrote to PCMag and here is my response... I think everyone who is experiencing the problem should also write to them. In the next few weeks I will be creating a petition website that people can go to who feel they were wronged by aBit and then we can take this up a further level.

---- snip----
Hi, Al:

I've forwarded your message to Abit through our channels. Let us know if
anyone contacts you in the next few days.

I should note that the AI7 has a special coprocessor that dynamically
monitors the clock rate. So it may, in fact, not be running at a
constant voltage or clock rate at all, but varying the clock rate
dynamically in order to maintain stability. How this translates into
real-world performance is unknown.

Also, overclocking is always a bit dicey. It's always possible that you
got a CPU that won't tolerate a higher voltage. Or it's possible that
some Abit boards happily run at higher voltages, while many don't.

At any rate, give the process a couple of weeks; if you don't hear from
them, drop me a line.

Best regards,

Loyd Case
Technical Director
ExtremeTech / Ziff-Davis Media
www.extremetech.com

Voice: 408.732.1347
Mobile: 408.712.5235

-----Original Message-----
From: Miller, Michael
Sent: Tuesday, December 09, 2003 2:36 PM
To: deleted
Cc: Stam, Nick; Case, Loyd; Salvator, Dave; Louderback, Jim
Subject: RE: Can you please help me?

I'm not sure what I can do, but I'm copying some of the folks from our
ExtremeTech web site, which covers more of the motherboards and
overclocking issues.

Michael

-----Original Message-----
From: deleted
Sent: Tuesday, December 09, 2003 3:08 PM
To: Miller, Michael
Subject: Can you please help me?

Dear Sir,

I am just a regular guy in desperate need for help from someone as
powerful
as yourself. Yes you do have the power to help. In the past I have
seen
many different magazine and letters to magazine that help people out
with
their problems. Today my problem is not one easily solved.

I purchased an aBit motherboard called the IC7 MAX3 about 3 months ago.
I
was one of the first people in Canada to have one because I paid the
premium price of first delivery through my reseller. Along with this I
also purchased some high quality ram made by OCZ which was rated to run
at
250 Mhz FSB. I also purchased a intel pentium 4 2.4C. All this was
great
until I started to overclock a little, after all the board was marketed
as
"Built for Overclockers, By Overclockers". This was clearly the biggest
marketing joke ever when people started to find out that this board was
"Built by stiffs for stiffs". The proposed marketing propaganda was
their
new and improved OTES cooling, though it looks good it serves no real
world
purpose. They also claimed that the board was able to do up to 3.2v for
memory, however 2.8v is the maximum anyone can run stable. The same
memory
in a ASUS P4C800 deluxe will do 280 FSB as opposed to the Max3 at 266
FSB
at 2.8v.

I hope you understand the technical babble because it is very easy to
explain this in plain english. Simply, Abit's board does not do as
advertised. And now they are saying that there is no problem and that
we
will not fix the problem. However, they released a new board called the
AI7 which apparently works perfect at 3.2v. Please help me and please
help
a whole community of aBit IC7 MAX3 users who really need to get what
they
paid for. They don't answer our calls, they delete posts from their
forums
when we put the explanations of the problem. Anything that is bad PR
for
them, they delete. WE NEED YOUR HELP SIR! HELP!!!!

Yours Truly,
Al

--- snip ----

Cheers,
Axiom

dissappoiting...

Bought mine from newegg on dec 3rd and it has no m64 capasitor on the board.
The board is very stable and kicks butt

UnrealCpu said:

Bought mine from newegg on dec 3rd and it has no m64 capasitor on the board.
The board is very stable and kicks butt

Click to expand...

I'm with you unreal. mine is stabel as hell. I haven't looked for the cap however.

I don't think its a question of it being stable or not in general, but is it stable when using either 2.9, 3.0, 3.1 or 3.2 Vdimm ?? And what memory are you using if it is ??

I have an IC7 and this MAX3 , and have (2) 2.4C chips. Each chip is virtually the exact same animal in each MB. Actually, the IC7 can go one voltage level less on vcore with the same stability, but its probably variation in VR sections between the boards. They both do the same on memory timings and max FSB with my generic ddr400 and with mushkin lvl2 black PC3500.

The Max3 will not even begin to pass memtest with either set of memory at just about any timing when the above voltages are selected. 2.8 is the highest voltage that "works".
If you have memory that has low Vtt needs, the extra Vdimm might help you - The 2 kinds of memory I have won't work with it.
And BTW, the Vdimm IS stable - its the Vtt that is way out of spec on the upper voltages.

The above is when using the same antec TC 550 PSU.
Adjusting rails on the PSU makes no diff on either board.

So, that said, I am not surprised by the sameness of the results - I bought the MAX3 to have the CSA LAN, since I needed another P4 board anyway. I have just returned it in favor of the P4C800-E Dlx - We'll see.

- Chris

iamjcl said:

I don't think its a question of it being stable or not in general, but is it stable when using either 2.9, 3.0, 3.1 or 3.2 Vdimm ?? And what memory are you using if it is ??

I have an IC7 and this MAX3 , and have (2) 2.4C chips. Each chip is virtually the exact same animal in each MB. Actually, the IC7 can go one voltage level less on vcore with the same stability, but its probably variation in VR sections between the boards. They both do the same on memory timings and max FSB with my generic ddr400 and with mushkin lvl2 black PC3500.

The Max3 will not even begin to pass memtest with either set of memory at just about any timing when the above voltages are selected. 2.8 is the highest voltage that "works".
If you have memory that has low Vtt needs, the extra Vdimm might help you - The 2 kinds of memory I have won't work with it.
And BTW, the Vdimm IS stable - its the Vtt that is way out of spec on the upper voltages.

The above is when using the same antec TC 550 PSU.
Adjusting rails on the PSU makes no diff on either board.

So, that said, I am not surprised by the sameness of the results - I bought the MAX3 to have the CSA LAN, since I needed another P4 board anyway. I have just returned it in favor of the P4C800-E Dlx - We'll see.

- Chris

Click to expand...

I'm using what is in my sig at 3.2 vdmm. The Vtt reads 1.4. All the Vdmm is stable 2.9-3.1

Ok mine is stable and im at in windoctor

3.10/3.0 in bios
vtt is 1.60 and doesnt move
i have no m64 cap on my board
bought from newegg around the 1st week of dec

This board kicks butt
i like it and it has tons of features the AI7 dont have

I just bought some OCZ PC4200 enhanced latency ram 2*256
2.5 4-4-4-6
cant wait to get to 3.7 gighertz or even 3.8 hopefully stable

I'm sure everyone is aware that HWDR. reports Vddq and Vtt reversed - that is your AGP voltage displays in the Vtt line, and the Vtt displays where the Vddq / AGP should read. So if you have your AGP set in bios up from 1.55 to 1.6, your Vtt will read 1.6 in HWDR. It is incorrect. Read Vddq to get what is really Vtt.

- This has been discussed, but just in case anybody missed it...

- Chris

hardware doctor
it read 1.33 -1.34 for vddq
vdimm is at 3.10-3.09
vtt 1.6

bios is vdimm 3.0
agp 1.6

maybe someone can figure this one out.

After aBit PR closed the vDimm/Vtt thread and deleted a whole bunch of my posts, and gave me a threat that I would be banned if I talked about it on the abit forums anymore. I wrote to PC Magazine Chief Editor, who wrote back to me and said that he has spoken to a special contact inside aBit and that they will be contacting me soon via Email.

Today the thread magically opened up and aBit PR made it look like he had no idea why it was closed in the first place. Meanwhile we was sending me PMs telling me not to talk about it anymore.

Who was it that PC Mag talked to that made aBit open the thread up again. Why did they open it up? Did PC Mag figure something out warned aBit not to mess up cause they have their eye on them.

Magazines are always looking for a good story. Is aBit really conspiring to not fix the bugs and trying to cover up. Did PC Mag catch on and ABit decided to revist the issue instead of ignoring it?

Then I PMed aBit PR and what rule I broke by posing a question of "Why have closed the vDimm thread?", and he replied that I was breaking the rule making aBit look bad. How does that make aBit look bad. It was just a question, a simple answer would've been adequate. Instead he just deleted the thread. Then I said to him.

Me: Why are you beating around the bush, why don't you just tell me that you guys can't figure out what is wrong with the board, and I will cut the damn trace and fix the board my self".

aBit PR: "I don't know what trace cut you are talking about"
I'm sorry but this is all a little too strange.

NOTE: CATCH22ATPLAY and BIGTOE presented their finds to aBit PR last week. And the mod has already been in several forums.

Me: There is already a fix that Catch22AtPlay found, will you guys cut the trace for me, and soldier the wires and still give me my warranty?" The mod is located here: http://www.3dmaxx.net/forums/showthread.php?threadid=546

30 mins later, the vDimm forum had a posting from aBit PR saying that he will send this finding to aBit Taiwan to see if it is a workable solution. Where the hell was he last week went the solution was sent to him, and was he just bull$hit his way through?

This is all too strange.

I would love to hear some comments on this.

Cheers,
Axiom

I just read the abit issue on abits forums . I now understand all about the vdimm and vtt issues. I know for a fact that my m64 cap is gone when i purchased my board from newegg. I however still have the vddq issue which is below half than my vdimm which might cause higher overclocks. I would like to know if the AI7 is fixed. I might return this board for something else unless i hear of a new 1.2 version coming out soon. Also i just ordered some OcZ pc4200 EL ram for my ic7max3 i. I hope i can take full advantage of this ram with my 3.0c

what do you think?

UnrealCPU,

Youre lucky your board is stable (if you are pushing your memory limits, anyway) because you clearly have the Vtt issue. If your Vdimm is 3.0, your Vtt should be ~1.5. It is low - that is the point I was trying to make. The fact that hwdr. confuses AGP voltage and Vtt voltage in the readout is only confusing the matter for people - just check the voltage in bios to be sure.

consider yourself very lucky if memory is stable at 3.0 Vdimm
and 1.3x Vtt. If you change memory you may not continue these reslults, I'm afraid. Good luck!

- Chris