Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: Matt_H on April 09, 2012, 12:36:28 AM
-
Just when I think I've got my rebuilt A2000 stable! Arg.
There's nothing connected to the joystick port, but lowlevel.library seems to think that the red button is being held down, as does stuff run through WHDLoad. Anyone have any ideas? I'm a little scared to plug in a CD32 pad (to see if the presence of an actual joystick without any buttons pressed would make a difference), just in case something gets zapped.
This machine used to be configured as a Toaster system. Could there have been a hardware modification for setting up a GPI trigger?
Thanks in advance!
Edit: Hardware details - 2630+2632, AD516, OneStopMusicShop, 2091, Indivision ECS, MegAChip
-
Your left Commodore Amiga keys are stuck down. Unstick them.
-
Has the board been worked on at all? I know someone whose Joyport pin was being grounded and their port was acting wonky.
-
Your left Commodore Amiga keys are stuck down. Unstick them.
Unplugged keyboard. No effect. It's not this. Thanks, though.
-
Has the board been worked on at all? I know someone whose Joyport pin was being grounded and their port was acting wonky.
It hasn't been worked on since I owned it, don't know about its prior history. I'll have to take a closer look next time it's open.
-
In that case: You are having a nightmare. Wake up. :D
-
Have you let ur womenfolk play with it?
One time one of my girlfriends somehow plugged in a joystick or mouse into the A2000 control port in such a way that it smashed one of the pins way backwards into the motherboard. This made the port malfunction but I can't remember how.
Yes, I gave her a spanking :D
-
Yes, I gave her a spanking :D
Did you ever think that she did it, 'cos she *WANTED* a spanking??? ;)
-
Try swapping the 8520's the fire signals go through U300.
-
Try swapping the 8520's the fire signals go through U300.
I like this idea the most, but I swapped the chips and didn't get any change. Could they both be borked? What symptoms should I be looking for if U301 is fried, too?
Regardless, I've ordered some replacements. Always good to have spare 8520s around :)
-
I like this idea the most, but I swapped the chips and didn't get any change. Could they both be borked? What symptoms should I be looking for if U301 is fried, too?
Regardless, I've ordered some replacements. Always good to have spare 8520s around :)
Maybe you'll need to poke around with an oscilloscope and look at the levels on the port.
There's an A2000 schematic here, but unfortunately it's resolution isn't that good. http://amiga-manuals.xiik.net/schematics/index.php
Try the A500 schematic as I would guess it's probably the same around the ports.
-
I like this idea the most, but I swapped the chips and didn't get any change. Could they both be borked? What symptoms should I be looking for if U301 is fried, too?
Regardless, I've ordered some replacements. Always good to have spare 8520s around :)
Well, as I read the schematics, the same failing bit in U301 would show up as a constant RTS or DTR from your modem. You could check the emi filters at emi238 or emi241. Sometimes they short to ground. I even saw one where loosening the mobo mounting screw at the mouse port fixed a problem.
-
If you can rule out the CIA it can only be the EMI or a short on the board. Theoretically, Agnus' _LPEN input could also cause a problem (pull J200 to make sure) - my bet is on the EMI or a loose screw.
Oh yes, early B2000s have discrete ferrite bead and capacitor instead of an EMI.
-
Thanks for the advice, everyone.
I don't like the direction this is going. New CIAs - no difference. Removed J200 - no difference.
I pulled the entire motherboard - found a few solder-side legs that had snapped and cleared them away. Added some electrical tape between the plastic insulation and the RF shielding. Still no difference. I'll pull the board again for a closer look, but, assuming that fails, what are my options? What exactly are these EMI things? (i.e., what's the part called that I should use as a replacement?)
EDIT: Solved, finally!
Slathered the RF shielding in loads of electrical tape to add some additional insulation. Phew! What a mess this was.
EDIT2: Just kidding, not solved, actually. New symptoms, though. Works correctly with no accelerator installed, works correctly with the accelerator installed and disabled in software (A2630 right mouse button boot menu), does *not* work with the accelerator installed and fully enabled in 030 mode. I didn't add extra insulation directly under the CPU slot, so will try that next, otherwise... any more ideas?
-
Thanks for the advice, everyone.
What exactly are these EMI things? (i.e., what's the part called that I should use as a replacement?)
EMI: ElectroMagnetic Interference. Basically there to keep your machine from spewing out radio frequency interference thru it's connectors. It's a passthru connection with a capacitor to ground to shunt off RF noise. Some have ferrite beads on them but these are 6800pF capacitors to ground, according to the schematics. If I found one bad, I'd just remove it, and jumper across. ;-) But that's just me, the FCC scofflaw. ;-)
-
Matt did you mention the Revision your B2000 Motherboard is, I remember there being compatibility problems with Earlier B2000 revisions.
-
Just when I think I've got my rebuilt A2000 stable! Arg.
There's nothing connected to the joystick port, but lowlevel.library seems to think that the red button is being held down, as does stuff run through WHDLoad. Anyone have any ideas? I'm a little scared to plug in a CD32 pad (to see if the presence of an actual joystick without any buttons pressed would make a difference), just in case something gets zapped.
This machine used to be configured as a Toaster system. Could there have been a hardware modification for setting up a GPI trigger?
Thanks in advance!
Edit: Hardware details - 2630+2632, AD516, OneStopMusicShop, 2091, Indivision ECS, MegAChip
pull and reseat paula chip. if thats not it try changing with a known good one. if that doesnt work check for a bad socket with internal break.
-mech-
-
Matt did you mention the Revision your B2000 Motherboard is, I remember there being compatibility problems with Earlier B2000 revisions.
It's a 6.2, which I think is pretty safe.
-
pull and reseat paula chip. if thats not it try changing with a known good one. if that doesnt work check for a bad socket with internal break.
-mech-
Would Paula explain the weird accelerator symptoms?
-
You have checked the joy port pins ? If corroded the pins can sometimes catch and get pushed backward when inserting a joystick plug. looking at the arms on the back of the socket, where they enter the mother board, will tell you.
-
You have checked the joy port pins ? If corroded the pins can sometimes catch and get pushed backward when inserting a joystick plug. looking at the arms on the back of the socket, where they enter the mother board, will tell you.
The pins all look good.
-
Would Paula explain the weird accelerator symptoms?
Reseated Paula, no effect.
-
Ok, you'll all hate me for this being so simple, but not as much as I hate myself. Plugged in a CD32 controller. Now all is well.
This is - by far - the strangest Amiga problem and solution I have ever encountered. I hope I never deal with something like this again.
So, for the record, A2630 + lowlevel.library + no joypad = problem. Madness. Absolute madness.
-
I've noticed strange behavior from the lowlevel.library on a 3000 also. I installed this new lowlevel.library and it fixed all my problems:
http://aminet.net/driver/input/NewLowLevel.lha