Help! The WEIRDEST Amiga issue I have ever come across!

  • Thread starter Thread starter Jameson
  • Start date Start date
  • Replies Replies 9
  • Views Views 450
Jameson

Jameson

Active member
AmiBayer
Joined
Jan 19, 2015
Posts
995
Country
Australia
Region
Queensland
If anyone can help I would be very appreciative. This is without doubt the weirdest problem I have ever come across with the Amiga.

I have just installed a newly recapped (and perfectly working ) A600 motherboard into a tower case. I have a Pistorm 600 installed, 2MB Chip RAM upgrade, RGB2HDMI, Amikey USB keyboard adapter, Gotek drive, DVD Burner (connected to the A600 IDE interface. All of this is powered by an ATX PSU with an inline latching circuit so that the front power button on the case will latch ON and OFF the power.

Everything is working perfectly..... with one perplexing exception - when I boot Amigatestkit from the Gotek I can access all tests fine, and they all complete perfectly..... except the audio test - as soon as I press F5 from the main menu to go into the audio section the system shuts down instantly - it is as if the moment the audio section of the Paula chip is engaged it sends a short to the ATX PSU. I then need to press the soft power on button twice (which in itself is weird) - the first time does nothing, the second time powers up the system and everything is working fine...... except the audio.

As I said, this is so weird and I honestly have no idea what is happening. I have disconnected the audio out from the A600 motherboard to the rear RCA jacks of the case to rule out some issue with grounding etc - no change.

Could it be some weird issue with the audio not playing nice with the latching circuit for the soft power ON/OFF? The latching circuit is quite simple and is just a small PCB that takes the 3.3v and ground of the ATX PSU and uses them to power the PCB that is then connected to the power on switch on the case, which when shorted locks on and sends the 3.3v of the ATX to the ground of the ATX which turns it on. How could this have any interaction with the audio out of the Amiga?

A friend of mine who is doing some recapping for me has looked at it and he is lost too (and he also says it is the weirdest fault he has seen on the Amiga)... help! :)
 
UPDATE. On hunch I tried to load a couple of games from ADF that play music..... and I THINK I have zeroed in on the issue.... but its still weird.

Seems anything that activates the low pass filter causes the ATX PSU to shutdown.

So its not the audio, its the low pass filter..... but why would this happen?

Any ideas?
 
Tried a different (older) PSU and it has no issues.... must be that this newer PSU is more sensitive and the audio filter LED is registering as a partial short?

Anyone able to shed any light on this?
 
Last edited:
I would look carefully at the capacitors that were recently replaced for a bad solder joint or damaged trace on the pcb. Do you know if the audio was working properly before the caps were replaced?
 
Not the A600 motherboard mate - have tried two - same behaviour. Tried a different ATX PSU and the problem is gone and audio is working perfectly. I think this newer PSU is just VERY sensitive to grounding any lines - it treats them as a partial short and shuts everything down... well, my guess anyway until someone has a better explanation.
 
With all the safety features today I'm not surprised somehow, but if it works with another PSU then keep running that.
There is no way of knowing what it is, unless you hook that thing up to a scope
 
Have you tried the other versions of the ATK (1.16 - 1.21) or a newly reinstalled one as a 1 bit change in the code could be the issue? [In college, one byte in my punch card code was mis-read and the program crashed; it then took me an hour of comparing the paper print-out to the punch cards (1975) to figure this out].
Properly wired ATX harness should eliminate any shutdown signal of the PSU, although a Gotek issue may be the fault.
 
bdb said:
Have you tried the other versions of the ATK (1.16 - 1.21) or a newly reinstalled one as a 1 bit change in the code could be the issue? [In college, one byte in my punch card code was mis-read and the program crashed; it then took me an hour of comparing the paper print-out to the punch cards (1975) to figure this out].
Properly wired ATX harness should eliminate any shutdown signal of the PSU, although a Gotek issue may be the fault.
Click to expand...

I actually tried some different game ADFs and everyone caused the same issue as soon as music changed the filter state. So it is definitely something to do with the filter state causing the ATX PSU to detect a partial short. It also does this when I use the reset feature of the Amikey - the entire system goes dead.

Have tried two other PSU's and they are fine, so I guess this newer one is just VERY sensitive.
 
bdb said:
So I suppose the PSU wiring harness (out of the transformer l is at fault?
Click to expand...
Well, the harness SHOULD be okay - the PSU is brand new and the harness itself is visually fine..... I am thinking it is more the control PCB of the new PSU - all ATX PSU's have circuitry to detect shorts and shutdown.... I think this newer PSU simply has more sensitive short detection and is thus detecting certain changes in continuity from voltage outputs to GND as being partial shorts - for instance when you use the reset function of the Amiga it is essentially grounding PIN 63 of Gayle, and I think the PSU is detecting this tiny grounding of a line that is connected to (I assume) one of the +5v lines as a partial short, thus shutting the entire PSU down.

The low pass filter also uses grounding of line to engage, which is also being detected as a partial short by the PSU....

..... that's my theory anyway. :)
 
You must log in or register to reply here.
Back
Top Bottom