View Single Post
Old 01 June 2012, 18:43   #1298
Join Date: Jun 2010
Location: out in the wild
Posts: 1,251
Originally Posted by paulo_becas View Post
Can anyone back me up on this?
Yes, I can.

It was said and written numerous times, and I'm willing to repeat it until everyone understands it:

You cannot tell from the board revision if your board needs a fix or not. Commodore has never used anything that comes close to today's quality management. In everyday business, this resulted in boards being assembled with any combination of chipsets, using any combination of filter components that may or may not fix all issues.

Fixing a board to work in all screenmodes is a trial-and-error job. There are no instructions that are equal for equal board revisions. A scope helps, but what helps the most is a basic understanding of digital circuits. Visible Glitches are mostly caused by violated data-hold times in the data path from memory to Lisa. Measure the end of the Ram-CAS signal in reference to CCK and make sure that the CCK edge comes well before the rising CAS edge.

It's sometimes even necessary to exchange the memory chips on the A1200 mainboard. Note that memory is not purely digital, and memory chips do age. They get slower over time, which may result in Lisa not catching the data, but CPU never finding a problem with chipmem (for example using a memtest program).

Commodore produced a huge mess with all those chipset combinations. Some A1200 boards are even equipped with memory chips that require the address inputs to be stable during the CAS cycle, which is NOT the case for 2M Agnus and Lisa chips. The A4000 has a latch in the chipmem address lines in order to make most SIMMs compatible, but the A1200 mainboard connects Alice address lines directly to the memory chips.

I'm aware that this is getting way beyond the capabilities of a normal solder-junkie, but that's the exact reason why I'm writing this: If you can't get it to work, please seek help from an expert. One of the experts I'm thinking of is registered in this forum by the name "FOL". He works for AmigaKit and I consider him an expert in finding a fault like that.

Schoenfeld is offline  
Page generated in 0.05961 seconds with 10 queries