3.2v
Printable View
3.2v
4 dimms and you must disable command per clock (aka you should have 2T enabled). this is not a RAM nor CPU issue, its a design limitation of the amd64 platformQuote:
Originally Posted by nie_realny
ok thx
I can not even put stable 200MHz 2-2-2-8 :{ on tension 3,2 V for 4x512MB, and 2x512 walks stable on 224MHz :{ gold memory
even SPD to it can break down on 4x512MB gold memory
i run only my system in 2T ang cry
4 DIMMS is 2T no matter what speed in A64 ...Quote:
Originally Posted by nie_realny
has anyone here experienced a case with TMSP where if you up the voltage you get more errors? 1A4T to be specific about the chips. stable at 200mhz and up to 3.1-2v's. go over and error away. i'm not talking 10 errors here, i mean thousands on a single pass.
Its normal if u dont cool the zone enough.
Twinmos 2700 256mb dual sided bought early 2003. Rated for 2.5,3,3,7
Chips read:
M.tec
02474
TTD7608F8E60B
Will do:
230mhz 2.2.2.5 2.9v
Probably will do more since my old NF7-S would only go that far.
Yeah, I have 4 sticks of 1A4T SP3500 and found that issue with one.Quote:
has anyone here experienced a case with TMSP where if you up the voltage you get more errors? 1A4T to be specific about the chips. stable at 200mhz and up to 3.1-2v's. go over and error away. i'm not talking 10 errors here, i mean thousands on a single pass.
Did some testing to try to find the best matched pair before burn-in, 2-2-2 @ 2.8v, 3.0v, 3.2v for max memtest stable (when I say stable, 10-20 passes, just wanted an indication of what they would do).
All of the sticks were about the same up to 3.2v, 2.8v = ~220, 3.0v = ~230.
At 3.2v, 3 sticks would do ~240, but the other wouldn't do anything but error, and error hard, even at the speed it was error free at 3.0v.
Out of the four sticks, the best performer was one that had been given some burn-in previously since it was from an earlier batch than the other three.
The other two sticks which did not error above 3.0v have been the ones I've been playing with this week, gave them a bit of burn-in and they're responding well (2x512Mb 2-2-2-5 252mhz@3.3/3.4v), doing 255 @ 3.4v but need more tweaking time to get stable. (I'm new to BH so it's a learning curve for me, may have been a setting thing, they seem to have a volt/setting range that you can't deviate from much).
I'm planning on giving the uncooperative stick a right royal voltage bashing to see if it will shake it's reluctance to work with volts, should be in the next couple of days, I'll let you know how it gets on.
My hunch is that it's 50/50 burn-in/settings for me. I have found a lot out about these sticks from working with just the 2 over these past three days, they are very finicky about having the right volts for the right speed with the right settings. TCCD seemed more forgiving since I never had to worry about having the right voltage for the speed, it was always 2.7v.
Big ups to Rabbi for supplying some awesome sticks, this holiday is proving to be great fun. :)
Does TMSP nowadays come with Brainpower PCB instead of KO-6633?
Still KO-6633..
guys, this any decent for my "xtreme Twinmos memory"? :p:
It's not UTT but hey....
Brainpower PCB
Mtec labeld Chips
50B's & AMT in the serial
They do:
238MHZ 7-3-3-2T at the moment and I've got them to 6-3-3-2T but I lack the PSU for it at the moment :)
is this any decent for these chips?
^ Depend on voltage? :confused:
Email from twinmos.
http://img315.imageshack.us/my.php?image=twinmos0bm.jpg
Damn.Quote:
Email from twinmos.
it's @ 2.9Volts
Doesn't look good if I have to RMA any of my sticks. Luckily I bought 4 sticks of TMSP with Winbond BH5 chips (two in family computer running at stock :D) so if any goes wrong and I RMA only to receive non-Winbond, I've got 2 spare :D.Quote:
Originally Posted by beta67
If you've got RAM sitting here (Kingston Value PC3200) that has the Winbond's two circles and the little small indent at the bottom right of the chip, are you pretty much guaranteed it's UTT or something? I've had a look through the pictures and my chips look exactly the same as the TMSP with the winbond chips but just with obviously Kingston markings. They're also older than when the UTTs seem to have started appearing. I got those two modules around September of last year.
alot of different companies used the winbond chips. just post a pic and someone here can verify for you.
Another good way of checking is setting the Vdimm to 2.8-2.9v and setting the timings @ 2-2-2-5 1T 200mhz. If it boots and runs memtest, it's UTT (most likely).Quote:
Originally Posted by Chuck232
WARNING: Do not do this unless you know how to Clear CMOS on your board as if it isn't UTT, it will fail to POST
hi guys..
i Bought 2x512 twinmos pc3500, this week. but there are not bh5.
de last 4 code is: 1AAT.
14AT is utt bh5 but this???
LOOK the image
http://www.infohard.mpsnet.cl/twinmos.jpg
chip code:
A2S56030BTP
515ACN1P
Taiwan - 5
any info to tihs ram?
sorry my english
:D
The Twinmos web site is still saying their SP 3500 512mb sticks still have "BH-5 dies" but they also say "*All products specification change without notice". I've never seen that extra sticker in the middle of the stick - is that gold leaf on the sticker and IC? That sticker also has a recent production date so it might a revised version. Only way to tell what you have is to test it. Post your results ASAP. :)
those are powerchip ic's most likely, just like cadt & aadt.
i've seen aa-t's couple of weeks ago and ic's were labeled psc.
Got 2 512MB SP-3500 with 44B (BH-5) chips, and fitted them with individual chip-sinks (pic is coming). I`m not so familiar with BH-5 settings (got G.Skill GBLA), so if anyone could point me in the right direction, it would help me alot! Aiming for 250 at 2-2-2-5 1t... Yeah, I know, I might have to dish out for VX or Redline to obtain that....
I have 4x 512Mb SP-3500 BH-UTT that will all do 250mhz at these settings:Quote:
Originally Posted by type_a_positive
2-2-2-5-7-12-2-2-2-2-1-bank interleave enable @ 3.3v for 2 sticks
same settings but 3.4v @ 245mhz for all four sticks.
2-2-2-7-12-2-2-1-1 should work like a charm with the same voltage :toast:Quote:
Originally Posted by szukalski
Cool, I'll try that when I get home tonight. :clap:Quote:
Originally Posted by esdee
Try to avoid those new Twinmos PC3500 Speed Premium sticks that have ACMT ending P/N:. Those sticks use rebadged Hynix D43 chips that can´t do 2-2-2-5.
So no more BH-5 / CH-5 UTT? :down:
A peep in a local forum reported these ACMT's doing 2,6V 2,5-3-3-8 260Mhz
Not bad for 2.6 volts, especially the timings are great for this chip.
I know these aren't high voltage chips but - from what I've heard - more can be expected with a slight voltage bump?
I have 2 sticks of Twinmos 256MB PC3200 UTT (Winbond BH-5 chip), it's new BH-5 chip, right? The sticker on the ram stated "1A4T" & MTEC chip.. May i know how high the FSB can i go & what's the tightest timings for my Ram? Currently i'm running at 220Mhz@3.0vdimm (2.5-3-3-8). It seems like my RAM can't go any higher than 225Mhz... Can anyone of u help me to figure out the best timing?
1.5-2-2-5 is possible, probably 240mhz or more at 3.3v...
yeah, BH-5 it is.
i reach 250
You mean i can get 240Mhz if i change the timing to 1.5-2-2-5? But one problem is my KV8 PRO doesn't provide such high vdimm. My mobo is not vtt modded. I'm just using a modded BIOS which allowed me to change the vdimm to 3.2v, but my system seems can't boot up if i use 3.2v..Quote:
Originally Posted by Revv23
and the funniest thing is:
Cas 1.5 isnt even 1.5 (lower than 2), its just a MADE UP NUMBER, but people think they're cool and stuff because they can 1.5 lol...
the magic number is 2-2-2-X. DONT FOOL YOURSELF!
1-2 quick test's on a new non-burned 1X256 utt , waiting for the second one ;) to run few benchs :hehe:
3.56~vdimm [with 3.2 fron dfi does 260~ error free :)]
http://img387.imageshack.us/img387/6712/untitled6bj.jpg
http://valid.x86-secret.com/show_oc.php?id=38318
Athens, what do you mean by this --> 3.56~vdimm [with 3.2 fron dfi does 260~ error free :)] ??Quote:
Originally Posted by Athens[2004]
You're using Twinmos 256MB UTT BH-5 also? The same model as mine?
he used 3.56V for reaching 280Mhz,Quote:
Originally Posted by jien_xex
with 3.2V from the DFI he got 260Mhz
Well one of my sticks just died on me after 4 months use. Been running @ 237Mhz 2-2-2-5 1T 3.2V with active cooling. TMSP 3200 BH5-UTT Week 9 Year 05 44B 14AT chips. I guess it was just bad luck really. :( Now I'm gonna have to run 512mb for a week + whilst I get my return. Added to that, my DVDRW drive packed up and is in an RMA now and it broke my windows CD. Not a great few weeks I've had with computers.
My twinmos 2x256 UTT 3.3V 262Mhz 2.2.2.5
i have 2x512 atm AA4t
and on chips is mtec
05084
ttd7608f8e50d
what should i expect?
AA4T isn't that ch-utt?
what can you expect.. well you'll have to find out for yourself ;)
That rocks.... got pics? :D :thumbsup: :toast:Quote:
Originally Posted by Troman
im afraid littebit, dfi killed my 2x256mb ch5( they where very good :mad: )Quote:
Originally Posted by Garrett
used 510-2 bios, is 704-2 safe to use?
Just got RMA from TwinMOS Taiwan and the PC3500 SP they sent back are STILL 1A4T :D
szukalski... giv me an email at work or PM me here before Sunday night... my flatmate has a laptop so i can log into XS during the weekends again.
if anyone is interested... the new pc3500 is Week 31 2005 BH-UTT
could you post some results please? :D
will my twinmos cheap memory (m.tec) going to die when using high voltge such as 3.3v :confused:
omanhardware,
If the part number on your M.Tecs ends in AA4T or 1A4T should be fine at 3.3V.
If it's AADT or CADT, it might not like it.
What chips are 16ATT... not on the speed premium, but the PC3200 TwinMos standard. Looked in my local store, searching for bh-5 or ch-5 (1A4T or AA4T)
Can anybody tell me the max voltage these chips support and
what results can I expect when overclocking:
TwinMos PC3200 Dual channel Kit 2x 512 Mb - old Winbond CH-5
Markings on the ships are:
W942508CH-5
034251
433088900 (one side)/ 43312700 (other side)
Label:
P.N. M2S9J16AJAWB9F0811B-T
Model No.: M2S9j16A-WB
S.N. 344M35M33630676
Manufacturing Date (by Cpu-Z): Week 44, Year 03
my AA4t can do 260 max with 3,4v 2-2-2-5 , with 3,5v same clocks
2x512 stick, goint to sell and get some bh5 (real one)
what about this?
the right one is twinmos with AA4T on stiker
the left one is m.tec, with AA4T on stiker
the left one doesn't boot at 200 2-2-2-5, is UTT?
http://img146.imageshack.us/img146/4...twinmos0pl.jpg
It´s on auction in Sweden, can´t find any info on them, anyone :cool:
http://www.tradera.com/auction/aid_22263729
Left stick is Powerchip/PCS "garbage" :( (partnumber in SPD ends in CAD). I guess you've bought it second hand?Quote:
Originally Posted by Malachia
My Twinmos Speed Premium pc 400 AA4T do about 255 2 2 2 5 at 3.2V. Not really tested the max of them yet, but on nf2 they only did 250 no matter what voltage (tried up to 3.7v). Was the mem controller of the board which was already maxed out I guess.
I jus got some of that acmt stuff. came packaged with bh5 on the label but no joy. Its the psc stuff. sum1 here said that the acmt stuff was relabelled hynix d43?? how well do these things work? what sort of volts can i give them? They are twinmos SP 3500
I dont know if they are Hynix D43 or not, but if they are, my old Vitesta DDR500 with Hynix D5 reached its top @ 2.9v, and anything higher than that doesnt report any benefit... ;) I supose that the D43 stuff performs similar to the D5... :confused:Quote:
Originally Posted by Postal Dude
Also put 3-4-4-8 if you want to reach high FSB´s with Hynix chips, from my experience, anything tighter than that was a no go with Hynix stuff... :stick:
Lord.
3-4-4-8? is the latency increase worth the extra mhz?
worth a try at least. im runnin 240mhz 3-3-3-8 @ 2.8v i think. burnin in prime95 custom torture between 1024 and 4096k ffts
You will have to try if it worths for you...Quote:
Originally Posted by Postal Dude
I have sold my vitesta 6 months ago, so I dont remember what config I used, because it was with a 754 3200+... ;)
Give it a try, and tell us....
Lord.
Will see if its passed the 240Mhz 3-3-3-8 P95 1st. Im at work at the mo. Will check it when i get in. Change timings. Up to 250 and re-run P95 while im asleep. I kno they can hold 250Mhz 3-3-3-8 all day and all night in memtest. But windows dont like it. Will find maximum and burn them in while im on night shift
Ok guys. The PSC crap stuff that noone likes. 2.8v 3-3-3-8 240Mhz 11 hrs Prime Custon torture stable. currently doing torture at 2.9v 3-4-4-8 250Mhz.
Torture im doing is a custom one with 1024k for smallest FFT and 4096 for largest FFT. 15 minutes each FFT. off to bed now. See how its doin in 8 hrs :D
Since they are the chips that Adata Vitesta DDR500 mounts, they should be 100% stable @ 250 3-4-4-8. My old Vitesta reached abuot 280 3-4-4-8, I think that I have some screens... ;)Quote:
Originally Posted by Postal Dude
Lord.
9 hrs prime stable at 250 3-4-4-8 2.9v. They wont do it at 2.8 i jus get a windows freeze up. Jus done a Spi 32M with them at that. 29 minutes. Gonna see if i can get 3-4-3-8 with them tonight. See if i can tighten up the timings just a little bit. For £70 half decent timings ddr500 i cant complain really!!
if anyone has a dead stick or problems with 2 sticks working together you could try the prog in this thread: http://www.xtremesystems.org/forums/...ad.php?t=85533 . i couldn't even get 200mhz stable with my 2 BH-5's no matter what voltage. i wrote the SPD from one stick to the other and now can do a stable 240mhz @3.1v's . my board only goes to 3.2v's and reads 3.11v's actual in the bios.
gonna see what they can do when i get my 3700+ back from rma in my DFI 250gb with the options to go at least to 3.6v's(up to 4v's in the bios).
god when are u guys going to learn those days are over .even the origanal guy has forgot bout this post.i think ur last chance was when they had twinmos speed premium on sale at egg.when i hit 100 posts mine will b on sale for $50 for 2 256 mb sp that hit 255 on an35.theres no more cheap mem.utt or tccd.im just gonna wait it out and get some better mem when prices drop.its amazing.these guys are worse than gas sellers when they find out there products work.otherwise they try to give it away. period
original guy still around - but not fighting for TwinMOS anymore :nono:Quote:
Originally Posted by chester
Alonso,
I think he was referring to "esdee".
So if you are not with TM anymore, are you with someone else?
well i was just trying to help those who may have good but appearing dead good TMSP's left over. waste of good ram if you got some of the good ones that could be revived with the app i mentioned. if for anything else they could sell the now revived ram and get something else. money is money and i'm sure everyone could use some more:D
Hahaha, all the world needs is... money?Quote:
Originally Posted by omga14
I have a stick of TMSP that just might qualify for your method omga14 so your post wasn't in vain!
I'm keen to try it out once I get some spare time on my hands, and if it solves the problem then I owe you a beer :toast:
stay tuned until March then the cards will be mixed new again ;)Quote:
Originally Posted by Reefa_Madness
well if you owe me a beer for this you could end up on the winning end cause you'd have to come to hawaii to buy me one:D but the real credit goes to Waus-Mod who started the thread.
That thread started by Waus-Mod would probably have saved at least half of the TwinMOS ram that was reported dead in this thread...I bet those that would "die" and then come back to life after a few days might have benefited from having their SPD info reprogrammed...certainly couldn't have been any worse than making keychains out of them...or worse...RMA'ing them and getting PowerChip based modules back. :) I'm sure quite a few of them were run with both ClockGen and CPU-Z running at the same time. No reason to believe the problem could only occur when running an Asus/CT-479/Pentium-M combination.Quote:
Originally Posted by omga14
******************
Alonso,
I'll keep an eye out for you in March...but only if you give me a hint/taste of what's to come :D
@ Reefa
will be something with fast memory - thats for sure! ;)
ok an update on my stupid powerchip or hynix modules. Whichever they turn out to be (either way theyre not the BH-5 i wanted from twinmos :( )
at 2.95v i can hit 270Mhz by loosening the tras to 10 (will this hit performance too much)
Still prime blending the things now. Think i gonna keep prime blending at this voltage to see if i can burn them in some more. If i can get 280 at this timing and voltage then my 144 im purchasing monday will at least be able to run (hopefully) FX-57 speeds at 1:1
Cmon cheap memory!!!!! Do it for meeeee
that is Hynix for sure... nice results BTW. :up:
but didn't you say your ICs were marked PSC? or you said your retailer told you they were PSC... :confused:
Hello, I wanted to ask if you know at what timings and what clock would 2 pieces of Twinmos 512 3200 DDR go? (The chips are Twinmos 0349D TMD7608F8E50D and they have a half-cyrcle on the edge near the pins)
OH, forgot to tell that they are over 2 years old.
DH-5?Quote:
Originally Posted by Alonso
So, anyone knows if my ram is any good? could i achieve at least 220Mhz?
why don´t you just test...?
I cant, dont have the mobo atm and thinking about selling the ram, too...
A pal of mine got these memories:
M2G9JA8ATT9F081AA4
ends with: 1AA4
Now: He's able to do 2-2-2-x at 247MHz at 3.63V
What kind of chips could that be?
winbond chips should end with xxxx1A4TQuote:
Originally Posted by Aleman
is he able to run 2-2-2-5 at any volts at all?
could also be that they are switching codes from time to time...Quote:
Originally Posted by esdee
could be interesting when these sticks are produced -> fully info of sticker needed
i believe its Winbond - either "oldschool" or nex UTT (0,175 - BH)
the 1A is standing fpr 0.175 process and i think noone here is still producing ICs in 0.175 ;)
but like esdee said: why not just testing if 2-2-2-5 are possible with 2,7-2,8V then these should be Winbond based
Quote:
Originally Posted by Alonso
nice to see you around alex :)
Alonso....remember this:
I'll keep an eye out for you in March...but only if you give me a hint/taste of what's to come
March is almost here. :)
hi guys,
i wanna ask..
my Twinmos SP3200 AA4T UTT 512x2 have error when i test that with Memtest86+
even only in 2-2-2-11 1T 200 Mhz @3v .., that AA4T have 55 error count! what the??!!!
and then when i set back toi standard timing 2.5-3-3-8.. that error still occur
so i think it's not about timing.. but bad memory modules
what should i do ? can i fix it before i choose RMA solution..because when i RMA, i dont think so that i 'll get winbond chipset again..:rolleyes: :rolleyes: :(
Need more info.Quote:
Originally Posted by dhetz
Mobo, voltages,PSU and bios settings.:toast:
ok ..i get it
Mobo DFI Lanparty Ultra B
i put vdimm @ 2.8 (std timing) and 3v (2-2-2-11)
my PSU .. SIlverstone ST405 400 watt
rail 12+ @ 11.95
5 v @ 5.1
(from multi tester)
bios setting standard.. only FSB 200 Mhz with vcore 1.6
thx :toast:
I think I had the best twinmos SP when they used BH5, mine did 1.5-2-2-6 @ 250 1T 1:1, i was happy, then i got some TCCD :D
http://img242.imageshack.us/img242/9030/memstd2.png
is this bh5 or winbond?
bh5 and winbond is the same...
hard to tell from cpuz screen - possible to up a pic of the chips?
TMD7608F8E50D is what chips?are they capable of 2-2-2-5?what oc +-?
skinnie,
http://www.xtremesystems.org/forums/...26&postcount=1.Quote:
DIMMS Found With UTT
-----------------------
(*) Twinmos SpeedPremium (04464 production week,a 4 at the end help too)
(*) Twinmos\Twinmos Twinmos\Mtec chips (cheap ones) (04514 ,same as above)
Make sure if you can take a look at the chips of the RAM that there are
---------------------------------------------------------------------
(*) Two Dots marking
(*) Metal Dots (two in each side!)
(*) AA4T [or 1A4T] (not AADT [or ACMT])
(*) week 0xxx4 (the ending 4 is winbond chips)