User avatar
dansalvato
Boise, ID

Posted Sat Feb 01, 2020 9:44 pm

Hello, I'm trying to troubleshoot a strange issue that arose with my A2000. I've searched around online and can't find anyone else describing something similar to this.

Recently, in order to install an Indivision ECS onto the board, I replaced a nearby capacitor with one that is smaller in dimensions, so that the Indivision ECS can fit. Since then, whenever I power on the A2000, the fan spins for about 1 second before shutting off completely. (I'm using an original A2000 PSU, not an ATX replacement.)

If I remove the 68000 from its socket, this behavior doesn't occur - the PSU fan spins and continues to supply power to the board until it's switched off again. Strangely, if I keep the 68000 inserted and instead remove Gary, the PSU fan remains spinning for about 4 seconds before shutting off. I also tried using a different 68000 from another Amiga of mine, and the issue persists.

Given the behavior described above and the fact that this only began happening after working on the motherboard, I don't think the PSU is the issue here. The behavior also shouldn't be related to the capacitor I replaced, because it continues to occur even switching between the old and the new capacitor. It could be that handling the board while replacing the capacitor caused something else to break.

Does anyone have suggestions on what else I can test to determine the cause of this? My skills are fairly basic, so if you recommend any tests with a multimeter then I may need some guidance. Thanks in advance for your help, and I very much hope I can get this machine back up and running.

Edit: Forgot to mention that there is some corrosion damage around where the battery used to be - this never stopped the Amiga from fully working before, but I suppose it could have caused something unpredictable to happen while handling the board, if some of the traces were weakened by it.

Edit 2: Like with the 68000, removing the Kickstart ROM and then powering on also sustains power (video outputs a black screen). Replacing the Kickstart ROM once again causes the PSU to power off after 1 second. Tried with two different ROMs. Also, removing the right CIA chip boots the Amiga to a yellow screen for ~5 seconds before the video signal disappears. Replacing it (tried two different chips) once again causes the PSU to power off after 1 second.

User avatar
obitus1990
USA

Posted Sun Feb 02, 2020 7:52 pm

Is the polarity on the capacitor you replaced correct? When you say it is of smaller dimensions, I am assuming you mean it’s physically smaller and you matched the capacitance value and voltage (or went a little higher with the voltage of you couldn’t find the exact voltage). Are you sure the top of the cap (where metal is exposed) isn’t touching the underside of the indivision and creating a short? I usually bend the legs and lay caps on their sides if they might interfere with expansions. If that is the only thing you swapped out, and the problem began immediately after that, that’s the first thing I’d check. Also, so you have any other expansion cards in the Amiga? If so, remove them all when trying to troubleshoot.

User avatar
dansalvato
Boise, ID

Posted Sun Feb 02, 2020 8:47 pm

obitus1990 wrote:
Sun Feb 02, 2020 7:52 pm
Is the polarity on the capacitor you replaced correct? When you say it is of smaller dimensions, I am assuming you mean it’s physically smaller and you matched the capacitance value and voltage (or went a little higher with the voltage of you couldn’t find the exact voltage). Are you sure the top of the cap (where metal is exposed) isn’t touching the underside of the indivision and creating a short? I usually bend the legs and lay caps on their sides if they might interfere with expansions. If that is the only thing you swapped out, and the problem began immediately after that, that’s the first thing I’d check. Also, so you have any other expansion cards in the Amiga? If so, remove them all when trying to troubleshoot.
The polarity is correct, and it's not shorting with the Indivision because I've been troubleshooting with both the Indivision and Denise with the same results. All other expansion cards are removed.

I tried swapping Agnus out for a different one, and the machine no longer powers down immediately, but instead gives me a dark yellow or brown screen for a few seconds before the video loses sync. That's my best lead so far - it might have to do with the Agnus socket or something.

Edit: I'm a little more confident now that it's the Angus socket, as I seem to get slightly different results whenever I reseat or wiggle the chip. I think the next thing that I'll try is to replace that socket, in which case I'll need to wait a few days for it to arrive.

User avatar
dansalvato
Boise, ID

Posted Wed Feb 05, 2020 4:54 pm

Quick update on this - I successfully desoldered the Agnus socket and have a new one on the way. It's a PLCC 84-pin socket which is really cheap and easy to find.

I was worried about the desoldering since I only have a basic desoldering pump, but it worked pretty well. After desoldering each pin, I took tweezers and gently wiggled each pin to ensure that it was completely free. If it didn't wiggle, then more desoldering was required. Only after I confirmed every pin was loose did I try to remove the socket from the board.

Next I just have to give the area around the socket a good clean, and hope that soldering the new socket also goes smoothly and fixes my issue. FYI, the motherboard is revision 4.4. I believe most of the A2000 boards label Agnus as "Fat Lady".
1.jpg
Last edited by dansalvato on Tue Feb 11, 2020 7:44 pm, edited 1 time in total.

User avatar
Christian

Posted Sun Feb 09, 2020 11:04 am

Did you buy the capacitor on eBay and/or is it from China? Did it have long leads or were the leads short (then probably used)? Either way it may be an old and/or used capacitor. For that reason I only buy capacitors from distributors such as DigiKey, Newark, Arrow or Reichelt.

User avatar
obitus1990
USA

Posted Tue Feb 11, 2020 7:02 pm

Pulling on incompletely desoldered pins isn't a good thing to do. Especially after having subjected the vias to the stresses of desoldering. You can easily rip out a plated through hole doing that. Apply more solder, then flux, and heat and desolder again. Glad you got it out without hurting anything.

User avatar
dansalvato
Boise, ID

Posted Tue Feb 11, 2020 7:50 pm

obitus1990 wrote:
Tue Feb 11, 2020 7:02 pm
Pulling on incompletely desoldered pins isn't a good thing to do. Especially after having subjected the vias to the stresses of desoldering. You can easily rip out a plated through hole doing that. Apply more solder, then flux, and heat and desolder again. Glad you got it out without hurting anything.
Sorry, I went back and realized how poorly-worded my post was - I didn't mean to imply that I tried forcing off a component that wasn't fully desoldered. I just meant that wiggling the underside of each pin to ensure it was loose was effectively the strategy that I used to make sure I could remove the component without having to force it off. I edited the wording such as not to mislead anyone who might stumble upon the thread when attempting a repair of their own.

User avatar
dansalvato
Boise, ID

Posted Tue Feb 18, 2020 8:30 am

I have another update to this - the issue took a bit of a bizarre turn, but I have it solved for now.

The culprit was the PSU after all. I measured the voltage on each of the pins and discovered that the voltage on the 12V pins was below the acceptable threshold. It's kind of strange that it worked perfectly up until the point I removed the motherboard, and then failed immediately after. Based on the timing of this, I had previously assumed it had to have been a motherboard issue, but I was wrong.

There's a little more to this. In the past, I had replaced the old PSU fan with a new Noctua fan. In order to reduce the fan noise, I installed the fan's included "low-noise adapter" which reduces the RPM on the fan a bit. I think it's just a cable with a resistor in it that lowers the voltage going to the fan. It makes a huge difference in noise output.

Well, after removing the fan's low-noise adapter, the PSU started working again. I'm baffled as to how the low-noise adapter never caused me problems before, but is all of a sudden causing the PSU voltage to drop. Why would the amount of voltage going to the fan even have an effect on the voltage going out to the computer? Could it be that my PSU is actually going bad, and removing the low-noise adapter is only buying time? I don't have enough expertise to speculate any further than that, but any insight would be greatly appreciated, if only to understand what could have happened here.

User avatar
Christian

Posted Mon Mar 16, 2020 10:37 pm

Some big box Amiga power supplies monitor the current draw of the fan. If the fan is disconnected or drawing too little power, the power supply shuts off.

Is your power supply perhaps an Elec&Eltec or Phi Hong. One (or both) of those brands have the above mentioned safety function.

User avatar
dansalvato
Boise, ID

Posted Sat Mar 21, 2020 7:31 pm

Christian wrote:
Mon Mar 16, 2020 10:37 pm
Some big box Amiga power supplies monitor the current draw of the fan. If the fan is disconnected or drawing too little power, the power supply shuts off.

Is your power supply perhaps an Elec&Eltec or Phi Hong. One (or both) of those brands have the above mentioned safety function.
This must be it! Pretty sure it's an Elec & Eltek PSU, so that makes sense. But it's still very odd that the system worked for so long with the low-noise adapter, only to suddenly stop. It must have just barely been above the safety threshold. I suppose I'll eventually replace it with a passively-cooled ATX PSU, but for the time being I have the system up and running, and that's good enough.





Return to “Hardware”