Blank screen on Osborne executive boot up

[generic486]

Hello everyone,

I've been messing around with an Osborne Executive lately trying to get it going. At some stage the power supply had been replaced with an ATX one but I've checked all the voltages and it is all good.

The problem is most of the time, when I boot the computer up I am left with a blank screen. I press the reset button and I do get the boot screen, however, the boot disks I've made refuse to work. (I have cleaned the drives, perhaps alignment?)

Sometimes, less than 5% of the time, the computer will do something different. The computer will do one of two things. First and more common is it will boot up and run self test diagnostics. It will then report the computer is not working properly with a P in the top left hand corner of the screen. I can press reset and get to the familiar boot screen. However, in less than 1% of the times I've tried, I can get it to boot up with no problems. It won't show anything on the screen but it will beep and then the boot screen will appear automatically.

I have begun the tedious task of checking the logic board as the amber CRT gives me the same results as an external monitor I have connected. I have reseated all socketed chips but it gives me the same results.

I have start with the reset switch circuitry as this appears to be the solution to all the problems I'm having as it will display the boot screen after being depressed, no matter what every I get.

However, I was wondering if anyone is privy to possible cause of this problem so I could narrow down my task.

 

[daver2]

So (if I understand your issue correctly) when you power up the Osborne you have problems - but if you press the "reset" button it boots up just fine. Is this correct?

If so - I would look at C57 (47 uF capacitor) and/or CR7 (1N4001) and/or the associated gates up to the RST* pin of the Z80 microprocessor. This pin should be low (briefly) when you power up the machine (or press the RESET button).

Look at "http://www.textfiles.com/bitsavers/pdf/osborne/executive/OCC2_Schematic.pdf" sheet 3 for the schematics.

Can you read electronic schematics and do you have any test equipment at all?

Dave

 

[generic486]

Exactly. The screen will appear every single time, however, I feel as though it should prevent me because there is obviously a serious problem and even if I get the drives aligned, I'd run into problems.

I'm right with you. I've already replaced c57 as well as c58 but no difference. I did test those old caps on a leakage tester and it was most acceptable. CR7 I'll have to double check. I'll most likely have to remove it from the circuit but I did an in situ test and I believe it to be okay.

Yea, I should be fine reading them. I've been checking the inverters using an oscilloscope and they all appear good so far. I'll update you when I get to the processor chip.

 

[generic486]

Very good news. I tried running the system without some chips in it that were connected to the processor and it's reset circuit. I pulled out a p8253 at UD1 and the computer will boot up fine with self test showing no faults. I can then use the B: as an A: to boot up to CP/M with no apparent problems. My questions are, do I need to replace this chip or is it non-crucial as it appear on the serial page (I assume for use with serial port), Also, what is the best way to go about aligning the A:?

 

[JDallas]

...I pulled out a p8253...and the computer will boot up fine...

Osborne Executive - P8253 (UD1) Problem:
The P8253 chip is a simple programmable interval timer with three outputs.

The boot lock-up is likely due to the code initializing the P8253 chip to generate its three output signals and then reading one of the counters back in a loop until it has counted past a delay interval. If the chip is bad reading the count is unreliable and may never change. If the clock signal for that timer is bad, the count never changes from the initialized value. In both cases the boot-code will never get out of its delay loop and effectively fail to boot up the system.

When you removed the chip, the boot code writes to the empty socket, but when it reads the count value in the delay loop, the data bus has no output driver (chip is removed), so the data bus floats and the micro gets what it thinks is the correct logic level from the data bus and continues with the boot-up process. Note the interval was not correctly generated.

The Osborne-2 (aka Osborne Executive) schematics show that the P8253's (UD1) three outputs are used for two RS232 data clocks (unimportant to boot-up) and for creating a signal called "Spindle Clock" that goes to the Osborne's floppy drive electronics. Its possible that during boot-up that none of these functions are yet in need so the boot code uses the counter for a quick delay function loop, terminated when a read count value crosses a threshold.

That could be a delay to let drive A: spin up before proceeding with booting from a floppy.

When you removed the chip, that signal would be removed too and when a counter value is read by the micro from the P8253, the data bus would have no logic level driver an the read-in value would not be valid. But if the boot code is merely looking for a bit in the count to toggle, the invalid data may be just right enough some of the time for the code to exit the loop.

On the clock input side, there is a 74S161 (UC25) that appears not to be important as it just provides a clock to the RS232 supporting signals - so I don't think that 74S161 is the problem. However, it the delay code just chose timer0 or timer1 inside the P8253 then it would require that clock signal from the 74S161 to be valid.

The third P8253 output signal is "SpindleClk" which goes to the Osborne unique floppy drive electronics board. That appears to be clocked by a signal named 2Mhz which is probably used by other chips; I'll see what all uses it, but I'd guess that that signal is ok because if not, every chip using it would also have troubles.

What remains that the P8253 is a likely failure. Inspect the pins to make sure one is not bent under and failing to go into the socket correctly. If you have a scope or a pulse counter, confirm that the three clock inputs to the P8253 are reasonably clocking; two are tied together being the output of the 74S161 chip.

Regarding the history of this computer, its interesting that the power supply was replaced; maybe a previous owner just bet that was the problem. I found this quote interesting:

...less than 5% of the time, the computer will do something different...

This fits a scenario above wherein the P8253 is dead and the boot-code is getting gibberish when it tries to read the status byte after initializing that chip.

 

[generic486]

Thanks for the info. I will certainly have to get a replacement chip as when I was testing some of the pins, they broke off. I've managed to align A: with my floppy drive know how and have reassembled it and it works a charm.

When I was doing checks, gate, out and clock 0 all would allow the computer to boot fine, however, when doing the same with 1 and 2 on the chip, the system would either give no display or it would tell me the computer is not functioning properly.

I assume that most, if not all, DIP-24 interval timers will work in this socket?