8088 board, problems booting up

[otacon14112]

Hi guys, I've been working on an 8088 board system on and off for about a year, and one thing that I can't figure out after all this time is why the system seems to work sometimes, and not work other times when I power it up. I've tried using different-valued capacitors, and snooping around with my oscilloscope, but I just can't seem to figure out the problem. It's like sometimes there are random signals on the various lines when I turn it on, and in those circumtances, it doesn't work right. Usually, all the LEDs on 8255 port 1 just light up. The LEDs appear brighter on the higher end of the bits. For example, bit 7 appears brighter than bit 0, and it fades dimmer with each lower bit.

I suspected maybe the clocking was faulty, so I checked out the processor clock pin, 8284 pin 8 CLK, and it had the same waveform both when the system worked correctly, and when it didn't. I also tried other 8284s and crystals, and the problem still persisted.

I think it's a problem with the hardware, but I haven't totally ruled out software. I'm not advanced enough in x86 assembly to be able to see if there is a logical error. Still, I think it's hardware-related.

Here are some pictures I took of the system:
8284 pin 8 CLK when the system is not working: http://www.flickr.com/photos/9114793@N03/sets/72157634880600574/
8088 pin 25 ALE when the system is not working: http://www.flickr.com/photos/9114793@N03/sets/72157634874156595/
8088 pin 25 ALE when the system is working: http://www.flickr.com/photos/9114793@N03/sets/72157634874292649/

Schematics:
http://www.flickr.com/photos/9114793@N03/sets/72157634881520510/

A sidenote: This has been happening from the very beginning. It didn't just start happening when I added an extra chip or something. The issue has existed ever since the beginning. I just figured it was because the thing didn't have enough chips to make it work right, but it's at the point where I know it shouldn't be doing this.

I really didn't want to ask for help on this. I wanted to isolate the problem, diagnose it, and learn along the way, but if I scratch my head anymore, I'll go bald.

 

[Chuck(G)]

I note that you have a reset button on the 8284. Without cycling the power, can you reset the system from a non-working condition to a working one?

Does each IC have a 0.1 uF (104) capacitor from +Vcc to ground?

 

[otacon14112]

1. Sometimes
2. Yes

I did notice that the ground wire was unhooked from the LEDs, so after I plugged it back in, the LEDs don't light up unless the program boots up and actually runs, but the problem of the system working sometimes and not working other times still happens.

 

[SpidersWeb]

I'm watching this thread closely because I have two machines that do this also.
Power ON - sometimes they wont start but a reset always works as does flicking the PSU on/off very very quickly (as in 0.5 seconds tops).

I'm no where near as skilled or as technical as you guys, but I pondered up a theory - let me know if it sounds plausible or garbage:

Is it possible, that these boards are just powering up a tiny bit too soon, before the PSU is stable?
I remember discussions about the "Power Good" line being ignored in a lot of clone boards - so I wondered if a timer, that relied on a cap, was used?

 

[otacon14112]

I tried holding down the reset button before I turned on the power, and held the button for a couple seconds after the power was turned on, and it did not help one bit. The board still had the same problem . Also, I was monitoring the CLK line with my oscilloscope the whole time, while holding down the reset button, turning on the power while still holding down the reset button (the oscilloscope showed that the line was low the whole time I held the reset button in), and then releasing the button.

 

[lucasdaytona]

I'm watching this thread closely because I have two machines that do this also.
Power ON - sometimes they wont start but a reset always works as does flicking the PSU on/off very very quickly (as in 0.5 seconds tops).

I'm no where near as skilled or as technical as you guys, but I pondered up a theory - let me know if it sounds plausible or garbage:

Is it possible, that these boards are just powering up a tiny bit too soon, before the PSU is stable?
I remember discussions about the "Power Good" line being ignored in a lot of clone boards - so I wondered if a timer, that relied on a cap, was used?

I had boards that did the same thing! They all needed a reset (not all the time, but very often) to work ok, but none of them was a 8088 system. The cause on some of them was bad solder joints, in another was a faulty cpu. And in others I never figured the reason. Usually the problem was gone if I replaced some parts (cpu and ram), a funny thing is that the "bad" parts (the parts that introduced the crazy behaviour) worked perfectly on others systems. I have a 486 cpu that can't pass on Himem.sys (windows 95 first version) test, but if I use it in another board it works 100%, and if I put another identical 486 cpu on the first board (even same revision) the himem.sys tests works fine. This is madness. Maybe your reset issue is related with "incompatible" parts, sounds crazy, but I saw that sometimes.

Anyway, otacon14112, did you tried different cpu and ram?

 

[otacon14112]

I tried a different 8284, different 6116 (2KB RAM), different crystal, different capacitors, different 373 latches, different 245 data transceiver, but not a different CPU because there are too many wires around it, and it would be a major pain to unhook all of them. I could try bending them around it, but there are too many, and it's just pushing everything around when I try to do that. I might try it, though.

 

[otacon14112]

I'm watching this thread closely because I have two machines that do this also.
Power ON - sometimes they wont start but a reset always works as does flicking the PSU on/off very very quickly (as in 0.5 seconds tops).

I'm no where near as skilled or as technical as you guys, but I pondered up a theory - let me know if it sounds plausible or garbage:

Is it possible, that these boards are just powering up a tiny bit too soon, before the PSU is stable?
I remember discussions about the "Power Good" line being ignored in a lot of clone boards - so I wondered if a timer, that relied on a cap, was used?

I do know that crystals require a certain amount of time before the frequencies they generate become stable. I never thought about that possibly being the problem, though. I'm glad you mentioned it. I might put a PIC microcontroller on the board and program it to cause a very small time delay (which will give the crystal time to stablize), before sending the appropriate 8284 line high or low. I don't remember what logic level it is. I may have to AND it with the line.

 

[otacon14112]

I discovered the problem. It was being caused by bad code. Although the board still doesn't start correctly on power-up, after I push the reset button, it always starts correctly now.

Unfortunately, I have another problem, and that is that the 7-segment LEDs take about 30 seconds before they gradually start to light up. I can't figure this one out. I've tried different-valued resistors, but I've gone by the schematics in the book. When I get a day off, I'm probably just going to hook up my 20x4 LCD screen to it and ditch the segment LEDs. But I still can't figure out why the 7-segment LEDs would take so long to come on. I mean the obvious reason is they aren't getting enough voltage until a while has passed - I confirmed that with my oscilloscope, the question now is why is it taking so long for the voltage to rise. Oh well, something else for me to tinker with