Amiga 500 Black Screen Issue

[Kaigan]

Hey, everyone,

So, I'm pretty sure that I did something stupid.

I had (had being the key word) a working revision 5 A500 motherboard that I was working on putting into a tower case. However, I'm pretty sure that when I tried a power-on test, I didn't have at least part of the board properly separated from the case and something shorted.

Now I have a non-working A500 motherboard. Here's what I get when I power it on now:

• Both the keyboard power and floppy lights come on and stay on, but dim. With the connector plugged in properly, the floppy light turns on and stays on. No power light.
• The keyboard caps lock light stays off, and the keyboard appears to be unresponsive. The system doesn't seem to respond to a keyboard reset command. The caps lock light blinks one, and responds to presses by turning the LED on or off.
• I can hear the floppy drive click once. With a disk inserted, the motor seems to quietly spin. The disk isn't read, though.
• I receive a black screen (but an actual signal) over the RGB port.
• Nothing further in the boot process seems to be happening.

This is the state with the machine running in either its stock configuration (with a Kickstart 1.3 ROM), or with upgrades to Agnus (8372A 1MB) and Denise (8373 "Super").

I do also have a Vampire 500 V2+, and an Indivision ECS Scandoubler. With the scandoubler installed, I occasionally will receive the boot screen from the Indivision board over VGA, but not always. I do receive a black screen signal over VGA.

With the Vampire and scandoubler installed, the system seems to present the Indivision boot screen more regularly, but I do not receive any output over HDMI, including the Vampire boot screen.

Other things I've tried so far:

• I've tested the power supply. All voltages are good.
• I've reseated all socketed chips, including Agnus.
• I've swapped the two CIAs with each other, but to no visible effect.
• All major chips, RAM, and the LS-series logic chips appear to be receiving the proper voltage.
• I think the clock signals are working properly. My old analog oscilloscope has problems of its own right now, so I can't be 100% certain yet.
• With a Gotek drive connected, the drive will fully initialize and the LCD will display the number of the selected disk image.

No capacitors appear to have blown, and there isn't any obvious damage to the board two resistors have been heated, but still seem to work.

I'd greatly appreciate any information the community can provide in helping me figure out what might be wrong, or even just some next steps of where to look. I feel really stupid for breaking the board, and I'd really like to bring it back to life if I can.

Thank you!

[MigaTech]

Firstly, do not worry, the mistake is easily done if you have not built a tower or tried to retrofit an Amiga system before.

Black screen usually means "CPU NOT DETECTED"

Can you get to early boot menu?

[Kaigan]

Originally Posted by MigaTech

Can you get to early boot menu?

I don't think the early boot menu was working when I tried it with the Vampire installed. I'll check again, with the standard 68000 in place, once I'm home from work.

Thanks for the help!

[chip]

Welcome to the forum Kaigan

[Kaigan]

Originally Posted by chip

Welcome to the forum Kaigan

Thanks for the welcome, Chip!

Unfortunately, I'm not able to access the early boot menu. Regardless of holding the mouse buttons or not, all I get is a black screen.

[MigaTech]

If the cap lock isn't operative there is a good chance the CPU isn't addressing the CIA's. Does it blink once when you first power up?

Although this might not be the overall issue with your A500. You have obviously shorted something out, when you attempted to retrofit in the tower unit. Finding that is what we need to do next.

You do have access to a multi meter and have checked certain things. Now can you see anything visually, like trace breaks or burns? Does any of the components look burnt out?

[Daedalus]

What are the various reset lines doing? If they're being held low by a failed part, the CPU can't start. Check pin 18 of the CPU to see if it goes high and stays there very shortly after power up. When you say you've checked the power supply for the major chips, did that include U8 (Even CIA, pin 20). If that CIA isn't getting power, it will give the continuous floppy motor and click on power up symptoms you see there, and its power comes through some vias in the area which could have hard-to-see damage. That alone won't be enough to stop the machine from booting however...

[Kaigan]

Originally Posted by MigaTech

If the cap lock isn't operative there is a good chance the CPU isn't addressing the CIA's. Does it blink once when you first power up?

No, the caps lock light doesn't turn on at all.

Originally Posted by MigaTech

Now can you see anything visually, like trace breaks or burns? Does any of the components look burnt out?

The two resistors just above the keyboard port look to be burnt to me. I'm not seeing anything else, but I've attached pictures, if it helps.

Pictures:
The two burnt resistors: https://drive.google.com/file/d/1l1Z...ew?usp=sharing
The front of the board: https://drive.google.com/open?id=1vH...qTpcIbmZ9uIN5P
The back of the board: https://drive.google.com/open?id=10r...FnbWYnMErigF_U

[Kaigan]

Originally Posted by Daedalus

What are the various reset lines doing? If they're being held low by a failed part, the CPU can't start. Check pin 18 of the CPU to see if it goes high and stays there very shortly after power up.

Pin 18 on the CPU goes high and stays high.

Originally Posted by Daedalus

When you say you've checked the power supply for the major chips, did that include U8 (Even CIA, pin 20)

Yes. It's reading 5.1 volts.

Thanks again for the help, folks!

[Daedalus]

Those two resistors have certainly had some heavy stress at some point. They're on the 12V rails to the serial port, so have a potential of 24V between them. If they were shorted that would certainly explain the resistors being heated. But the resistors themselves won't stop the machine from booting.

The caps lock LED should still flash once even if there's no communication with the computer (CPU or CIA dead). Again this points at a power issue (check pin 4 of the keyboard port for +5V and pin 6 for GND - the two pins either side of the missing "key" pin). There's also a possibility that something is holding the _KBRESET line low (pin 3 of the keyboard port). Like the CPU reset, this should normally be high.

[Kaigan]

Originally Posted by Daedalus

The caps lock LED should still flash once even if there's no communication with the computer (CPU or CIA dead). Again this points at a power issue (check pin 4 of the keyboard port for +5V and pin 6 for GND - the two pins either side of the missing "key" pin).

Chalk me up for a second silly mistake. My A500 must be old enough that its keyboard connector isn't keyed, and I had it plugged in backwards.

With the keyboard connector plugged in correctly, I get no power light, a solid green floppy drive light, a single flash of the caps lock light, and the caps lock key will respond to presses by turning the LED on or off. No change on the black screen, though.

Originally Posted by Daedalus

There's also a possibility that something is holding the _KBRESET line low (pin 3 of the keyboard port). Like the CPU reset, this should normally be high.

The _KBRESET pin goes high and stays high when the computer is turned on.

[MigaTech]

If the Caps Lock key is responding to presses, try pressing several times, if the Caps LED turns on/off each time, then the CPU is responding to the CIA's correctly. <Which means the fault is somewhere else on the board.

If the Caps LED sticks or doesn't respond, then the CPU is not responding to CIA interrupt request.

BLACK: can also mean CIA error if not booting!

Those damaged resistors will need replacing, see if you get any readings from them. Also check R406 and R215.

[Kaigan]

Originally Posted by MigaTech

If the Caps Lock key is responding to presses, try pressing several times, if the Caps LED turns on/off each time, then the CPU is responding to the CIA's correctly.

If the Caps LED sticks or doesn't respond, then the CPU is not responding to CIA interrupt request.

So, this is really odd. The caps lock key will respond exactly 12 times and then stop working until I power cycle the machine. It doesn't seem to matter how long it takes for me to press the key 12 times.

I'll check the resistors as soon as I can and let you know what I find. I may not be able to respond until sometime tomorrow (PST), though.

[Kaigan]

I managed to sneak in a little time to check the resistors:

R406 is reporting back at 2.9-3.0 Ohms.

The two visibly-damaged resistors are reading between 46.0 - 50.0 Ohms.

I haven't been able to find a resistor labeled as R215. There are two unlabeled resistors on the board, however:

The one on the top, just below the Odd CIA chip reads at 68.4 Ohms.

The one on the bottom, in the corner near the RGB output, reads 73.8 Ohms.

[MigaTech]

Sorry my error, R215 is another Amiga system just forget that one.

R406 should be reading 4.7 so it's a little low.

The Cap Lock seems to be working if able to press on/off at least 10 times. Although maybe we shouldn't rule it out just yet.

@Daedalus, what do you make of R406, could it be low enough to cause no video?

[Daedalus]

Originally Posted by Kaigan

Chalk me up for a second silly mistake. My A500 must be old enough that its keyboard connector isn't keyed, and I had it plugged in backwards.

Oh well, at least you discovered it before something more drastic like replacing the keyboard.

With the keyboard connector plugged in correctly, I get no power light, a solid green floppy drive light, a single flash of the caps lock light, and the caps lock key will respond to presses by turning the LED on or off. No change on the black screen, though.

Are you sure that's not a solid green power light and no floppy light? Is the LED still staying dim, or does it turn bright after a second or two?

So, this is really odd. The caps lock key will respond exactly 12 times and then stop working until I power cycle the machine. It doesn't seem to matter how long it takes for me to press the key 12 times.

Yeah, that's the keyboard buffer filling up and not being able to send any more keypresses because the CPU isn't responding. This tallies with my earlier thought on the CPU not running. There's an outside chance that both CIAs have been destroyed in the same way (the Amiga will still boot with U8 failed, and if only U7 had failed, you would see the power LED working correctly and a boot attempt when you swapped them), but I find that unlikely.

The two visibly-damaged resistors are reading between 46.0 - 50.0 Ohms.

That's about right. So they've been overheated but not destroyed by whatever happened there. There's a good chance that damage was already there from a separate shorting incident in the past - shorting the serial port pins is relatively easy by trying to connect metal-shelled connectors at an angle while the machine is turned on. But as I said, they're not causing the issue you have so don't worry about them for now.

R406 is reporting back at 2.9-3.0 Ohms.

That's fine, though that failing wouldn't stop the machine from booting, just stop the video output working - the exact opposite of your symptoms.

I haven't been able to find a resistor labeled as R215.

Nor have I. I'm not sure what one he's referring to...

The one on the top, just below the Odd CIA chip reads at 68.4 Ohms.

That's fine. It doesn't have a label on the silkscreen so is an afterthought. It's labelled XR1 in the schematic.

The one on the bottom, in the corner near the RGB output, reads 73.8 Ohms.

I don't know this one myself, but likely similar to XR1 above, and probably nothing to worry about if the markings indicate a resistance around that measured value.

My next step would be to see if there's any attempt by the CPU to communicate. Try stick your scope on a couple of the data pins from the CPU and see if you get any response at all. If there is, it will be very brief because it's halting after only a couple of instructions.

[Daedalus]

Originally Posted by MigaTech

R406 should be reading 4.7 so it's a little low.

@Daedalus, what do you make of R406, could it be low enough to cause no video?

Nope. The value isn't really critical, once there's some resistance there. Besides, the lower resistance could also be due to parallel resistances within other parts connected in the same circuit. And, as I said, that alone won't be enough to cause a problem - this machine is giving a video signal and no boot, failure of R406 will give no video signal but should boot fine.

[solidcore]

I had similar issue myself with my a500 rev 6b, so at this point mate I'd buy a CIA (off the bay?) and grab Diagrom if you can, in my case KS would either be black or white screen but diagrom would boot up and tell me one CIA had bad timing however that depends on which order the CIA chips were in, in one configuration they were both fine and then in another only a single CIA was failing - hmmm strange! so I bought a new CIA and swapped the faulty one out and what do you know it no more black or white screen (depending on KS version).

[Kaigan]

Originally Posted by Daedalus

Are you sure that's not a solid green power light and no floppy light? Is the LED still staying dim, or does it turn bright after a second or two?

Yes, it's a solid green floppy light. My keyboard is one of the older models with the red power light and green floppy light. As soon as I power on the Amiga, that light powers on brightly.

Originally Posted by Daedalus

My next step would be to see if there's any attempt by the CPU to communicate. Try stick your scope on a couple of the data pins from the CPU and see if you get any response at all. If there is, it will be very brief because it's halting after only a couple of instructions.

I'll see if I can get anything on the data lines. Hopefully my old scope is still working well enough. I do also have a cheap USB logic analyzer, but I'm not certain that it's good enough to give me much out of the CPU. I'll definitely try, though!

Originally Posted by solidcore

I had similar issue myself with my a500 rev 6b, so at this point mate I'd buy a CIA (off the bay?) and grab Diagrom.

I have a known-working, fully-populated rev 6A motherboard being sent to me from eBay. I got a good deal on it, and I'm going to try very, very hard not to break this one. Once it arrives, I can drop the CIAs from the rev 5 into the working board one at a time (and any other chips worth testing) to see if anything is broken. It'll likely be a good week or so before it arrives, though.

Diagrom definitely looks nifty. I'm not sure if it would display anything for me in this case or not, but it'd be worth having one around regardless. I'll pick one up soon.

[Toni Wilen]

Caps lock (keyboard) stopping responding is normal when initial black (or full white when having some ECS Denise revisions) screen is shown because all interrupts are disabled and system isn't really running until screen turns lighter grey (at least when using KS 1.x).

Initial black only means: something is wrong, any chip or connection that has something to do with data, address or related signals can cause it. It can't be used to narrow down the problem (really, even if somewhere some lists says so). It only works with other color codes.

If your scope/analyzer works, you could check if /ROMEN or /RW or other status lines shows any activity. It would at least confirm that CPU is trying to do something.