Can't get Athlon MP 1800+ working right on K7S5A.

Don't ask how to over-clock.
Post Reply
bizzybody
BIOS Rookie
Posts: 59
Joined: Sat Oct 20, 2007 1:30 pm

On my PC-Chips 830LR, flashed with CheepoMan's modded final release of the ECS K7S5A (non-Pro) BIOS, it will not POST with the FSB/RAM set to 133/133 with an Athlon MP 1800+.

An Athlon MP 1600+ works just fine at that setting, it doesn't POST if the FSB/RAM are *not* the same frequency.

When I swapped CPUs, the setting switched itself to 100/100. I don't know why when the only difference between these two Palomino core CPUs is the 1600+ is 10.5x multiplier and the 1800+ is 11.5x multiplier.

133*11.5=1529.5 = the actual Mhz of an Athlon 1800+ CPU, so why does it not work?

Would one of the older HoneyX modded BIOS versions work? The later ones removed support for slower CPUs to make room for the newer Thorton and Barton chips.
Denniss
BIOS Guru
Posts: 3153
Joined: Thu Mar 21, 2002 8:16 pm
Location: Near Hannover (CEBIT) Germany
Contact:

Does it work if you try 112/112 or 124/124 MHz FSB/memory ?
Dou you use SDRAM or DDR-SDRAM, one or two sticks ?
edwin
The Hardware Archivist
Posts: 6286
Joined: Wed Mar 20, 2002 7:11 pm
Location: Netherlands
Contact:

what are the exact markings on that CPU?
edwin/evasive

Do not assume anything

System error, strike any user to continue...
bizzybody
BIOS Rookie
Posts: 59
Joined: Sat Oct 20, 2007 1:30 pm

The 1600+ is

AMP1600DMS3C
AGNGA 02023EPBW

This one is on CPU World.

The 1800+ is

AMP1800DMT3C
AGOIA 0229WPDW

CPU-Z says it's
Family 6, Model 6, Stepping 2
Ext. Family 7, Ext. Model 6, Revision A5

It's not listed on CPU World, yet. I need to take a picture of it to send there. Appears to have been only sold as retail with heatsink and fan.
Denniss
BIOS Guru
Posts: 3153
Joined: Thu Mar 21, 2002 8:16 pm
Location: Near Hannover (CEBIT) Germany
Contact:

What about answering my questions regarding the differnet FSB options ? (I don't use cheepo's Bios but the HoneyX Bios has those setting and I assume Cheepos's too)
Does the 1800+ MP work at all on the K7S5A or any other board ?
AFAIR HoneyX did not remove support for those CPU but only removed their name string from the Bios. They may be detected as unknown AthlonXP but they should work without problem.

The K7S5A is known to eat everything available for Socket A but limitet to 100 and 133 MHz FSB. It's also known to be somewhat tricky if SDRAM is used and problematic if two sticks of SDRAM are used.
bizzybody
BIOS Rookie
Posts: 59
Joined: Sat Oct 20, 2007 1:30 pm

The 1800+ POSTs at 100/100 (IDs as 1200Mhz XP)
112/112 (IDs as a XP 1500+)
and 124/124 (IDs as a 1600+) but nothing happens when set to 133/133 or H/W Trap and I have to clear the CMOS.

RAM doesn't make a difference, one or two sticks of PC2100 or PC2700.

At any setting it will POST at, XP Pro SP3 runs fine, but if it's not going to run at 133/133 then I'll stick the 1600+ back in.

This 1800+ is now listed on CPU World, I'm going to send in a picture of it now.
Denniss
BIOS Guru
Posts: 3153
Joined: Thu Mar 21, 2002 8:16 pm
Location: Near Hannover (CEBIT) Germany
Contact:

How strong is your power supply ? The K7S5A needs lots of power at the +5V rail thus not all PSU work with CPU needing a lot of power. How are the voltage readings in Bios with your 1800+ and 124/124 ?

Please try 138/138 MHz, Some K7S5A had a strange behaviour: No post at 133/133 but posting happily at 138/138 ......
Please make sure your K7S5A is not affected by bad caps - may result in really strange behaviour as well. -> www.badcaps.net

If nothing works please try the HoneyX OC030327 Bios, it's based on a Beta-Bios several months older than the latest official release but this latest Bios was always reported as acting strange sometimes (not to call it buggy).
bizzybody
BIOS Rookie
Posts: 59
Joined: Sat Oct 20, 2007 1:30 pm

It has a 300W Antec, so I believe the 300 watt rating. ;)

I'll try a faster speed, but I don't think this BIOS has anything between 133 and 147. If it'll run (stable) at 147 that'll be a big hellz yeah.

The caps are fine on the board. I've had plenty of experience with showing people their fubar old boards with swelled and leaking caps, then explaining how it'll be just as cheap to go for an all new box as it would be to buy a new board for their old CPU and RAM. (And way way wayyyy faster PC.)

That's how I acquired this board.
5+ year old hard drive croaked so I sold the customer on a new dual core Celeron box with a gig of RAM, launches MS Word instantly. Cost me a whopping $220 for the box, ready to assemble and install Windows XP. (They've got cheap enough I can once again make some profit!)

I pull everything apart to clean the dust out and discover the old no-name power supply is about to go BANG and let the magic smoke out of the 8 or so leaking caps. Pulled the Antec out of my PSU pile, plug it in and all good to go, no problems at all- except for this CPU that refuses to work at its rated speed. (The 1600+ was in it with no troubles for over 5 years.)

This was the last PC I put together 'on spec' and barely made any profit because of how cheap components had become- but the big box stores could still undercut me on complete systems. (Damn those huge Windows volume license discounts the OEMs get!)

I know this PSU is fine because I had a retired US Navy radioman check it out. When he was in the Navy he had to build the commo gear before he could use it. He fixes electronics nobody else will bother with, even had him do a cap replacement on a couple of not too old motherboards. He has walls full of bins of every kind of electronic component there is. Best thing is he only charges for the parts! :)
Agent24
BIOS Newbie
Posts: 30
Joined: Mon Dec 24, 2007 3:02 am

If that board was running with, as you say, a no-brand PSU with leaking caps then it's still possible that high levels of ripple could have damaged the caps or other components on the board.

Things can fail or be marginal in value/capability and yet not have any apparent signs of this
Post Reply