In the meantime, have you re-soldered the power connections to the analog board? There's a group of them, large-ish pins (might be a connector body, don't totally recall), which tend to develop near-invisible cracks in the joints where they meet the board. AFAIK the issue is common to all of the classic compact Macs, and it can cause all sorts of odd problems. On Fri, Mar 20, 2015 at 1:27 AM, Terry Stewart <terry at webweavers.co.nz> wrote:

On 03/20/2015 02:08 PM, David Griffith wrote: To me, it sounds as if there's an arc going on in the FBT. But that's just from general knowledge--I have no experience with Mac SEs. --Chuck

On Fri, 20 Mar 2015, drlegendre . wrote: Now for capacitor choice: would it be better to use surface-mount tantalums or through-hole electrolytics surface-mounted? -- David Griffith dave at 661.org A: Because it fouls the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail?

On Thu, 26 Mar 2015, dave at 661.org wrote: IMO SMD type tantalums are just way too temperamental for these applications. They are very intolerant of ripple and can only safely be used at a maximum of half of their rated voltage (according to the product engineers at both Kemet and AVX). I have some early 1990s equipment made up of 3 board stacks, and each of the boards in each device are absolutely loaded with tons of SMD tantalum capacitors. Each board has at least one SMD tantalum which is either browning or has failed catastrophically (thankfully the damage to the pc board substrate looks to be minimal). There is no particular pattern to the failures either, parts of different capacitance and voltages with different case sizes. Rather than solid tantalum, Panasonic SP-Cap or similar aluminum polymer capacitors are another possibility. These are available in the same SMD case sizes as traditional solid tantalum parts. For that fact, you might even be able to find OS-CON parts with the same SMD footprint as the original SMD aluminum electrolytic parts, too. For boards that originally used SMD aluminum electrolytic capacitors though, why not just use new SMD aluminum electrolytic parts? I've been doing this with good results in other equipment, including industrial control boards that need to last at least 10+ years in the field. Newer SMD aluminum electrolytic parts don't suffer from the problems the SMD electrolytics made in the early to mid 1990s had with faulty rubber seals (some of which was also due to the use of chlorinated cleaning chemicals during the manufacture of the boards).