Componentschip.com is a Electronic Components Distributor.

Componentschip.com is a Electronic Components Distributor.

広告

※このエリアは、60日間投稿が無い場合に表示されます。記事を投稿すると、表示されなくなります。

Tech You How To Test RAM Fault

2017-06-30 15:18:50 | 日記

In a variety of single-chip microcomputer application system, the chip memory is directly related to the normal operation of the system. In order to improve the reliability of the system, the reliability of the system is very necessary to test. Through the test can effectively find and solve the memory failure caused by the destruction of the system. This paper introduces several commonly used methods of RAM test for single chip microcomputer system, and puts forward a kind of RAM fault test method based on seed and bitwise reversal.

Method 1: a method of testing the system RAM is to check in two steps, has to the entire data area into the # 00H and # FFH, and then read out to compare, if not the same, then the error.

Method 2: Method 1 does not completely check out the RAM error, and in the reference analysis, a standard algorithm MARCH-G for RAM detection is presented. MARCH-G algorithm can provide very good fault coverage, but the required test time is great. The MARCH-G algorithm needs to traverse the full address space three times. Set the address line as "root, the CPU need to access the RAM 6 × 2n times.

Method 3: A method of performing a test by address signal shift. On the basis of the address signal for the whole O, only the signal of the address line Ai is inverted one time at a time while keeping the signal of the other non-detection address line Aj (i ≠ j) unchanged, so that the bit from the low bit to the high bit And then the signal of the address line Ai is inverted only once while the signal of the other non-detection address line Aj (i ≠ j) is maintained at the same time, and the signal from the lower High bit by bit. So the address signal shift is in accordance with 2K (K for the integer, the maximum address bus width) non-linear addressing, the entire address range can be seen as all 0 and all 1 as the background and then by the shift produced. Write different pseudo-random data to the corresponding memory cell while the address is changing. After the above write unit operation is completed, the address signal is shifted in reverse order to read out the written pseudo-random data and detect. If the address line is n, the CPU accesses only 2n + 2 memory cells in the system RAM. For example, how to test ds18b20 and how to test cr123a batteries.

Test methods based on seed and bitwise reversal are obtained on the basis of method 3 for further improvement. Method 3 is mainly using the whole O and all 1 two background number to shift the expansion, compared with the MARCH-G algorithm to obtain the fault coverage is slightly lower, but the use of fewer address units. Here we refer to the number of backgrounds in Method 3 as "seeds". To address the line for the 8-bit RAM, for example, the seeds were taken 00000000 and 11111111 two numbers, take 00000000,11111111,0000llll and llll0000 four numbers, and take 00000000,11111111,00001111,11110000,00110011,1100llOO, 01010101 and 10101010 Eight numbers to shift the expansion test, the fault coverage is not the same. The improvement method of seed number 2 is lower than that of MARCH-G algorithm. The improved method of seed number 4 is equivalent to MARCH-G algorithm. The improved method of seed number 8 can exceed the effect of MARCH-G algorithm. The improved method based on seed and bit-by-bit reversal can replace the MARCH-G algorithm as a whole, but the number of addresses required for different seed numbers is different. Set the address line for the n, the number of seeds is 2 when the need to access a total of 4 4 +4 times, the number of seeds is 4 when the need to access a total of 8n +8 times the number of seeds 8 need to access the total RAM 16n + And the MARCH-G algorithm requires a total of 6 × 2n times to access the RAM.It can be seen that the improved method based on seed and bit-by-bit reversal is much less than the MARCH-G algorithm's test time overhead, and the fault coverage increases with the number of seeds In addition, the test time overhead is different when the number of seeds is different, and the appropriate number of seeds should be selected according to the test time and the test fault coverage rate in order to achieve satisfactory results. If you want to learn more information about chip parts,please go to componentschip.com.





ジャンル:
ウェブログ
コメント   この記事についてブログを書く
この記事をはてなブックマークに追加
«  The Advantage Of Linear Te... | トップ | NXP Semiconductors Launched... »

コメントを投稿

日記」カテゴリの最新記事

トラックバック

この記事のトラックバック  Ping-URL
  • 30日以上前の記事に対するトラックバックは受け取らないよう設定されております。
  • 送信元の記事内容が半角英数のみのトラックバックは受け取らないよう設定されております。
  • このブログへのリンクがない記事からのトラックバックは受け取らないよう設定されております。
  • ※ブログ管理者のみ、編集画面で設定の変更が可能です。