Abit KR7A-R DDR Issues

DIMM Slot Positions effect memory stability – Joe

SUMMARY: Interference patterns with multiple DDR modules may account for problems with Abit’s KR7A-R.

Josh sent me an email about RAM problems with Abit’s KR7A-R. In short, users are finding that, while running at 133 MHz, they get memory errors; it appears that it shows up as system instability and while running Memtest86. His take on the problem:

I have come to believe this problem is due to the 4 DIMM slots and their timing algorithms (Abit is the only board I know of that has 4 DIMMS on a KT266A, and also the only board with this problem on the KT266A chipset¹).

Here are some forum threads that are following this problem:

Via Hardware
Via Hardware
Amdmb.com

What’s interesting to me is that this is dejá vu all over again. Back in the stone ages, we saw similar problems with Abit’s BH6. It was generally concluded that there were interference patterns (multipath signals) which resulted in memory problems. This was more apparent with multiple RAM modules.

This “fix” was to locate RAM modules in certain DIMM slots which, it was thought, cancelled out these “reflection errors” – it worked for many BH6 users.

Well, did Abit learn anything, or are we repeating a problem?

To add fuel to the fire, I tested various DDR modules on the Abit KR7A-R using RAM Stress Test Professional (R.S.T. Pro). This is a stand-alone diagnostic card that plugs into a PCI slot. The card is self-booting – there is no OS to get in the way of results. According to UXD, the RST Pro has

“…the capability of testing and validating RAM modules in a system. On many occasions, memory modules that pass standalone testers will fail using R.S.T. Pro. This is caused by the system chipset, noise, and several other factors. R.S.T. Pro in this environment, is the best, easiest, and fastest way to validate, and test system RAM.”

I have used RST Pro before with great results – it tests stability and high FSBs more rigorously than running benchmarks in Windows. With this card, not only can you test system RAM, it will tell you which chip on a memory module fails – more on this later.

To test DDR modules, I set up the KR7A-R using an Athlon XP, running it at a 10 multiplier. I watercooled the CPU and mounted a small 60mm fan over the power regulators to cool them – it did not want results influenced by heat or overclocking. I used the most conservative memory settings in BIOS – “AUTO” and “SPD”.

I used the following 256 MB memory modules:

• Crucial DDR 266, CAS 2.5
• Kingmax DDR 266, CAS 2
• Princeton DDR 266, CAS 2

I found that I could run any SINGLE stick of RAM for hours without any errors, so individually each DDR module was OK. However, as soon as another module was added, DIMM positioning became the controlling factor. When errors occurred, they showed up within the first 10 minutes of testing (note: each test cycle takes about 30 minutes).

When RAM failed in multiple installations, only ONE RAM module showed errors; errors were massive – usually every chip on the module in the first DIMM showed errors. This strongly suggested that DIMM positioning was determining stability.

As the table shows, I reached memory stability at 133 MHz with RAM modules in DIMMS 1 & 2, 2& 3, and 1, 2 & 3 – other combinations failed quickly and decisively. Further, upping FSBs did not improve stability, although some others seem to have luck with settings around 136-138 MHz (strongly suggesting signal interference).

I would suggest those with RAM problems on this board try different DIMM locations, keep RAM modules together (no DIMM gaps) and perhaps run the same size and brand. Abit does mention in their manual that “we suggest that you be populate DIMM1 to DIMM4 in order”, suggesting that Abit has at least some awareness of this issue.

Indulge me a rant: I’m getting sick and tired of seeing products shipping with systemic problems such as this. There are too many manufacturers that ship “beta” products to consumers. In my book, you pay good money for products that work; if manufacturers want us to field test their products, they should pay us, not punish us with malfunctioning products.

¹Michael emailed: “The Shuttle Ak31 Ver3.X has 4 DIMMS on a KT266A chip set.”

Email Joe