On-screen output |
Checkpoints to LPT ports |
Checkpoints to COM1 |
Speaker beep |
Comment | |
---|---|---|---|---|---|
Option #1: MDA video card | Yes | Yes | Yes | Yes | |
Option #2: CGA video card | Yes | Yes | Yes | Yes | |
Option #3: 4KB RAM at address A0000 | No | Yes | Yes | Yes | |
Option #4: 4KB RAM at address B0000 | No | Yes | Yes | Yes | The diagnostic will think that you have an MDA video card. |
Option #5: 4KB RAM at address B8000 | No | Yes | Yes | Yes | The diagnostic will think that you have a CGA video card. |
Option #6: None of the above | No | Limited | Limited | Yes |
1. | The 'Hot NMI' test is known to fail if either: - Math coprocessor (8087 chip) is absent and you have switch 2 in switch block SW1 in the wrong position for that (off). - Math coprocessor (8087 chip) is present and is faulty in a particular way. |
|||
2. | If there is a faulty RAM chip for parity, it will not be shown until all faulty RAM chip/s for data (in the same bank) have been rectified. | |||
3. | All or some of the 'Check ROM at Fx000' tests may fail because there are no ROM's at the tested addresses. |
Type of EPROM | EPROM Size |
Qty. | Image download |
Comment |
---|---|---|---|---|
2764/27C64 EPROM | 8 KB | 1 | Download | |
27128/27C128 EPROM | 16 KB | 1 | Download | See note 2 below. |
" " " " " | 16 KB | 2 | Download | Odd/even (high/low) pair of ROM's |
27256/27C256 EPROM | 32 KB | 1 | Download | See note 2 below. |
W27E257 EEPROM | 32 KB | 1 | Download | See notes 2 and 3 below. |
Note 1 | It is possible for a RAM chip to fail in such a way that affects only some of the chip's addresses. Imagine such a chip in bank 0, one that fails at say the 5K address. That would result in the diagnostics 'Check first 2 KB of RAM' test passing. But the diagnostic will report an error later when testing the remainder of RAM. |
Note 2 | Ruud's Diagnostic ROM for the PC and XT is actually only 8 KB in size. To create a 16 KB sized image, the 8 KB image was simply doubled. To create a 32 KB sized image, the 8 KB image was simply quadrupled. A related diagram is at here. |
Note 3 | Regarding the programming/writing of an image into the W27E257: Even though the W27E257 is an EEPROM, rather than a 'traditional' EPROM, you will probably still need an EPROM programmer (or other) to program/write to the W27E257. That is because it is highly unlikely that the motherboard of a PC clone or XT clone supports programming/writing to EEPROM's in its sockets. |
Note 4 | It is possible for a RAM chip to fail in such a way that not all of its addresses are affected by the failure. |
Note 5 | For most of its work, Ruud's Diagnostic ROM (RDR) needs some RAM for variables and for the stack. The motherboard RAM cannot be trusted (a common use of RDR being because there is bad RAM in bank 0). For that reason, RDR uses either: - If an MDA video card is present, the unused video RAM on that; or - If an CGA video card is present, the unused video RAM on that; or - If 4 KB of RAM at address A0000, some of that RAM. (Version 4.3 or later of RDR required.) - If 4 KB of RAM at address B0000, some of that RAM. - If 4 KB of RAM at address B8000, some of that RAM. |
Note 6 | On XT clone computers that have turbo functionality: • Ruud's Diagnostic ROM will attempt to turn off turbo mode, but it is an attempt only. • Determining 4.77 MHz operation (non-turbo) can be done by using an oscilloscope to verify a 4.77 MHz clock on the clock pin of the 8088/V20 CPU. |
Note 7 | Some people may be puzzled as to why there is no mention of the checkpoints also being sent to I/O port 80h, the port monitored by an ISA POST card. Although this diagnostic does also send the checkpoints to I/O port 80h, rarely does an ISA POST card work in an IBM 5150/5160 clone. Do not be surprised if it does not work. More information about that is at here. |