Assistance with single-board computer

[kattbjorn]

I'm still working on this thing, it's not working any more. Instead of showing the pattern, the LEDs stay lit and won't turn off.
I checked the ROM, the right data's loaded. That was the problem I had before, but it's not that this time. I checked every power and chip enable pin, they're all connected correctly. I checked all the data and address lines, they're connected properly...

I did find a problem with the clock timer, the 74LS04. it wasn't producing a wave. I don't have the proper equipment to check, I test it by connecting a LED between the chip and the power, then turning it around to see if its producing both positive and negative. First time I tried, it was only producing a positive output. I replaced the chip with a spare and it seemed to work normally, as far as I could tell. it was producing positive and negative output. The computer wasn't working properly though.

That's about all I can think of at the moment. It's got me stumped, it was working last time I checked and now it's not. Can anyone provide any suggestions? Can you see something obvious that i've missed? Anything?

[ElEctric_EyE]

I remember always having problems making an oscillator work using an 'LS' inverter and a crystal. I had better luck using a CMOS 4069 inverter. But if I were you, I would just avoid that situation altogether and buy a can oscillator. They're much more dependable.

Someone here suggested that you can place a AM radio close to your circuit and listen for the activity of the code racing through its paces.

[kattbjorn]

Okay... what's a can oscillator? Is that the type with 4 pins? and how would I use the radio? Do I have to tune it to the same frequency as the CPU?

[ElEctric_EyE]

kattbjorn wrote:

Okay... what's a can oscillator? Is that the type with 4 pins? and how would I use the radio? Do I have to tune it to the same frequency as the CPU?

Yes can oscillators have 4 pins, usually Power, Gnd, Output Enable, and Signal Out. Make sure you get one that's 5V power.

What size resistor do you have in series with your LED tester?

[GARTHWILSON]

Are you getting a good reset pulse with a clean (debounced) release? Are IRQ\ and NMI\ pulled up? I see you have a VIA there. Make sure its CS1 and CS2\ are true before phase 2 rises, ie, that you didn't bring phase 2 into the select logic. (You probably did it right if it worked before.)

There might be something helpful on the debugging page of the 6502 primer, at http://wilsonminesco.com/6502primer/debug.html . There are a lot of tips at viewtopic.php?t=342

[ElEctric_EyE]

GARTHWILSON wrote:

...(You probably did it right if it worked before.)...

If he was using a LED without a resistor to test signals off the 6502, he may have 'eventually' overloaded the output drivers. This would be my first concern, especially since he had it working. I am unsure of his skill level.

[BigDumbDinosaur]

ElEctric_EyE wrote:

If he was using a LED without a resistor to test signals off the 6502, he may have 'eventually' overloaded the output drivers. This would be my first concern, especially since he had it working. I am unsure of his skill level.

Good point, especially if using 74LS hardware, which doesn't tolerate such overloads for long. A lot of Commodore 64 and 12 owners who tinkered with connecting things to the user port discovered that the hard way when they took out one of the CIAs trying to drive loads without any buffering.

[anders84]

When using a voltage source, always have a resistor. Think about what happens when the voltage changes by a small amount, with no resistor, the LED current would shoot up. If it is current source, then you would not need a series resistor because the LED would run at the current source level.