Boot Rom Oddness

Hello,

I am trying to figure out why a single board computer isn't booting. I am a noob when it comes to processor boards so, here we go. I haven't mentioned much about the actual board because this is more of a general processor board question. Well, I think it is Just looking to see if I am way off. If i may be on to something please say.

The bootrom(PC28F640J3D-75) output enable pin is being strobed by the processor bridge indefinitely. I had my scope probe on it for 5 minutes and it doesn't stop. I am thinking that the board is getting reset every time it tries to load the boot rom data because it must fail the load and then it resets itself to try again. I'm thinking this is what happens but not sure. I am going to add some fly wires to it to see if the address and data are actually toggling. Plus I can check the resets of the CPU to verify it is indeed resetting itself.

What do you guys think?

Any processor that is running needs to access the rom constantly even if it's just in an infinite loop. What you're describing sounds totally normal to me.

Mike.

What should the computer do if it boots up?
I have a board that just sits there until it gets a key from the keyboard.

The /OE pin is often just connected to the CPU /READ line or derived from the READ /WRITE signal, depending on the CPU and hardware setup.

You need to look at the chip enable signals - and all of them in combination, on a device such as this that has several - to know if it is really being accessed.
/OE has no effect unless the device is also enabled.

Edit: Only pin B4 enable is actually relevant in your board as the other enables are tied to a pullup resistor, if the diagram you posted is for that board.


You don't say what type of computer board it is.
If the ROM is the only only program storage for the system it will be continuously accessed anyway, as Pommie says.

If it's supposed to be displaying something or booting from other media, it may be stuck in a loop waiting for whichever peripheral to be connected or become ready.

If it's a PC board of some sort, do you have a speaker / sounder directly connected to it, to hear any error code beep signals it may be producing?

A BIOS diagnostic card - aka "POST Card" - is also very useful for faultfinding with problems such as that; many PC BIOSs will output status codes to an I/O port as they go through the stages of the hardware testing and boot sequence & those status codes can be viewed with a simple plug-in card that has a LED or LCD display.

This ebay listing illustrates various versions for different PC bus types:


Beep codes list, if the board has a speaker connection:

More about the POST card numeric codes:

I'm with the rest, it sounds perfectly normal - and as others have already suggested, it's normal for systems to do checks during boot-up (called POST on PC's) and it's quite possibly sat in a loop waiting for a response from some peripheral or other.

In my previous job as a TV engineer we used to sell and repair (as far as we could) Sky TV satellite systems. For many years Sky had a contractual obligation to have the box connected to the telephone line for the first year, or permanently for multi-room boxes (more than one box in a house).

Unfortunately this left them wide open to lightning damage via the phone line, which destroyed the modem and various components in that part of the circuit. The effect on the unit was 'stuck in standby', as during boot-up the box was waiting for a response from the modem chips, and never received it. Presumably it was holding the I2C bus LOW, as if you removed the modem chips the boxes usually worked perfectly well, and the modem wasn't really used for anything anyway.

So that's an example of what might be happening with you, the processor is waiting for a replay from 'something'.

If the above is true then there ought to be some activity on the address lines.
Another possibility is the processor being held in reset by a supervisor ic or something, and and the proc is going to its reset vector.

dr pepper said:

If the above is true then there ought to be some activity on the address lines.

Click to expand...

It certainly should, but he's not checked that yet, or the reset pin.

I made a mistake. I actually see the strobe that never stops on the chip enable pin. I'm going to measure output enable now. Also, I will measure the resets to the processor because there is a CPLD in this chain and maybe it doesn't like the bootrom it gets and just keeps on resetting itself. That said, I have read that the processor can reset itself internally because it knows it got a garbage instruction from the bootrom.

More testing today. Results to follow.