Amiga 2000 Will not boot.

[Pheonix]

I'm at my wits end 2 different MBs and the problem remains. It will not boot, all I get is a blank screen and a flashing power light. The first thing I did was remove all the expansions. All zorro slots empty, removed the 2320, removed the kickstart switcher, and the problem remains. It fails with both the 1.3 and the 2.0 ROMs.

I guess I should be a little more clear. The "first" thing I did was purchase a replacement MB, as the web search stated an error with the on board Chip RAM (green screen, flashing power light.) With the latest tests, the green screen is gone (just a plain white screen.)

I cannot figure out what to try next. The battery was removed from my first board years ago (and it was still working just fine back then.) The new board (well, new to me,) also has the battery removed.

I cannot think what is left to test Any help would be appreciated.

[cpiac64]

verify if 68000 is mounted right

use a tester for verify if the kickstart have all pin (only one is not connected) connected to 68000

[hooverphonique]

power supply?

[8bitbubsy]

Are there any traces of battery leakage on the motherboard? The battery could have leaked before it was removed...

[Pheonix]

I removed the battery on both MBs, and neither one was leaking at the time. The area on both also appear clean. Checked the 68000 and it is mounted correctly, but I don't have access to a tester to check connection of KS to CPU.

Right now, the PSU seems the most likely culprit, and thinking back it is also the one that is making me the most nervous. When I first booted up the replacement MB, I got the insert WB screen, but with a strong green tint to the screen. Then the screen when all green, and it never came back. The 2320 isn't currently in the machine, I'm connecting using the A/V jacks.

If it is the PSU, then it looks like it may have also damaged something on the MB. It also looks like the 2320 has taken damage as well (the green tint.) So its looking like I'm going to have to replace MB, PSU, & 2320 unless I can track down exactly what on the board was damaged by the PSU.

[Pheonix]

It was pointed out to me (in a roundabout way,) that I have one other device that was common to both MB's and still connected. That was the PC floppy drive connected through a Kylwalda adaptor. I don't see how that could cause a problem, but I will be trying one last time with it removed. I am currently working on my 4000 (separate thread,) so the 2000 is put away again. I'll try to test it out and get back with the results later (maybe tomorrow.)

[Pheonix]

Nope, even with the Floppy drive disconnected, or an Amiga drive installed (Kylwalda removed,) the same thing occurs. I can only conclude that it is most likely the PSU.... Is there a way to test it, since I don't have access to another one right now? Assuming it is the PSU, is there a way to figure out what it might have damaged on the motherboards (assuming it might have done so, as the replacement worked for a short time?)

[TTD]

If I remember correct green screen is something with chip memory, but as I understand you have switch the old motherboard with another and it worked for a short time, and then white color and nothing more, or is it still green?
Can you test other chips like Agnus and/or Denise CIAs Odd/Even and GAry is these set or mounted correctly?
Just some ideas, you proparly allready have done that.

[Pheonix]

When I first replaced the MB (came with all chips,) it booted to the insert Workbench screen, but there was a strong green tint to the screen. On a reboot, I just got the green screen. I had powered down, and checked the 2320's mounting then powered back up.

If I boot now with the 2320, I get a solid green screen, if I switch to composite it's a solid white screen. The 2320 also switches to a white screen if I flip the flicker fixer switch. I no longer know which position is on & which is off, though.

The 2320, though, works just fine in my A4000, so I don't think there is anything wrong with it. This would, I believe, point to the Denise chip, but each MB had their own chips. So, the same thing applies. If it's the Denise, then it's 2 Denise chips with the same problem.

[cpiac64]

the composite is black and white, no colors on this machine, only on rev 6.4 if i remember well

i can repair your mobo but shipment from USA and return is very high

search a friend for burn a special kickrom for test

[Pheonix]

Wouldn't green produce more of a gray screen over a white one? Also, switching the flicker fix on 2320 on & off switches between green & white screen. The only difference is one is flicker fixed and one isn't. Not sure how that would happen, except that the tick line from the PSU (60 Hz,) is necessary for the flicker fixer to work, and if the PSU is faulty (the only common piece from both MB's - I removed everything else,) then that could effect the 2320 as well, when in flicker fixed mode.

On a side note, did the A2000 really depend on the 50hz/60hz AC frequency of the PSU instead of a crystal oscillator for timing? While many countries regulate the AC frequency, there are also many that cannot be relied on (from what I understand.)

[hooverphonique]

iirc, there is a jumper on the mobo where you can change it to not rely on the mains tick - can't remember the details though..

[Pheonix]

That would be J200. One position has MB use the graphics chip for timing, the other uses the PSU. The system won't boot in either position, and I had the 2320 removed while I was testing that, and put the jumper back to PSU timing before I put the 2320 back in to the machine.

Right now, I'm waiting for a replacement PSU. I removed the PSU, and attempted to measure the voltage across the connections. One +5V line registered no voltage, one was registering around 2.5V, and when I tried the third +5V line the PSU started smoking and is now pretty much toast

Seems the PSU is the culprit here, I just hope it hasn't damaged the MBs. I'll find out when I get the replacement. I have been looking into the ATX power adapter, and the Aminet schematics for the tick replacement circuit. I don't have the skills to build that circuit, but after studying it, I also found something else that might be interesting. My modified design is below (sorry about the quality, used a free CAD program to build it.

In the original design, there is only a single crystal oscillator and you would chose either a 3.93216 Mhz for NTSC or 3.2768 for PAL. Since I can't find a crystal on the board, I was wondering if the 2000 used this timer line for everything. If you noticed, in my modification, the single crystal is replaced by 2 crystals with a switch (double pole single throw,) to select which crystal would be used. With this, the machine could be set (with correct timing,) to either PAL or NTSC without breaking flicker fixing.

The file on Aminet is mix_and_tick.lha and comes in Amiga Guide format. The schematics there were a little difficult to read (being completely text based.) I'm sure, at some point, someone will attempt to build it with the switch mounted on the outside box. Personally, I would advise against it, as switching while the machine is active could be risky. Not sure how the machine would react, but there would be a brief moment when neither crystal is in the circuit, and the changed signal could effect other parts of the machine. Plus, you would need to set J102 appropriately as well. This assumes that your Agnus is the type that is compatible with both PAL & NTSC. The 8375 (I believe,) came in separate PAL & NTSC versions.

[Pheonix]

Well, it doesn't seem to have been the PSU I have now, literally replaced every last bit of my Amiga 2000. 2 different mother boards, each with their own set of chips. 2 different PSU's, with & without the A2320, with Kickstart 1.3 & 2.0 ROM chips (with and without my kick-switcher.) The kick-switch switcher seems to work fine (I get the high/low tones when I'm supposed to.) The only thing I can think of, is that the PSU went out, and when I got a replacement MB, the PSU blew that out as well.

This is rather frustrating

[EugeneNine]

I picked up an A2000 a bit ago and it had a leaky battery. I clipped off the battery and found that someone had pulled out and replaced the 68000 as one pin was bent up and out of the socket. When I pulled the 68000 out I found that a couple pins of the socket closest to the battery were badly corroded. I have now replaced the socket but get a green screen at boot up also.
I'm looking for more info on troubleshooting such as schematics, pinouts, etc.

[hese]

@EugeneNine

When you get green screen, I would suggest to check the Fat Agnus and Gary. Check that the pins of the ICs and sockets aren't oxidized/dirty and also that the traces aren't broken. And of course that the traces near the 68000 socket are still conducting. Are any of the DRAMs burning hot?

[Pheonix]

How easy is it to blow either of those chips? Is there a way to test them without buying replacements? If push comes to shove, I'll just buy both, if the price isn't too extravagant. I would need to buy the 2 meg version of the Agnus for my MegaCHIP, I guess. Though, one of my old 1 Meg Agnus chips may still be good (assuming it's the Agnus that got blown by the old PSU.)

All this assumes that I'm right about what happened. To me, it seems the only reasonable explanation. But I'm still hoping for a more definitive method of finding the exact issue.

[Pheonix]

Well, it turns out there's nothing left that's wrong. With all the testing I did before finding out the PSU was bad, I had ended up moving some jumpers to the wrong position. In searching for a means of tracking down what was wrong now, I had come across some better information on the blink codes I hadn't found the first time. It turns out that 10 short + 1 long blink of the power LED with a green screen means - clean & re-seat the Agnus chip. So I removed my MegaCHIP cleaned the socket & the pins, re-seated it and tried again. Now I got the 2 grey screens (dark grey, followed by light grey,) then a different blink code (5 short + 1 long,) and then the system started over.

This was new behavior, and I started looking for what could cause that. During that, I came across the proper jumper settings & discovered that one of my jumpers was set wrong. J500 was closed, when it was supposed to be open. I removed the jumper and it booted right to the insert workbench screen. At this time, all my cards and drives are in the machine, and it boots just fine to both Windows 1.3 and Windows 2.1

[StarQuake]

Good to hear it works now!

I also would be very interested to know how you got Windows to work on your A2000

[TroyWilkins]

Originally Posted by Pheonix

At this time, all my cards and drives are in the machine, and it boots just fine to both Windows 1.3 and Windows 2.1

Originally Posted by StarQuake

Good to hear it works now!

I also would be very interested to know how you got Windows to work on your A2000

I was wondering the same thing, I know there is such a thing as Windows 2.1, but I didn't know there was a Windows 1.3.

Anyway, in all seriousness, great work getting it going again my friend, bet you're happy to have it back up and running?