> On 20 Aug 2017, at 23:20, Chuck Guzis via cctalk <cctalk@classiccmp.org> > wrote: > > On 08/20/2017 12:12 PM, Brent Hilpert via cctalk wrote: > >> If the transformer were faulty and heating up internally, it would seem >> unlikely that it would cool off that quickly to recover, IME it takes some >> time for a transformer to cool off. > > I have an open-frame linear PSU that started behaving that way. Turned > out to be a current-sense resistor with a bad internal connection. As > things warmed up, the resistance would rise until the current overload > IC kicked it and shut things down. It was pretty frustrating because > the entire time, the output looked pretty stable.
Chuck wins some 16-bit air, or I suppose technically 16/32 bit air. I changed the reference resistors to the NE5561N and the PSU is now stable, no picture drift on the monitor and the 5V rail doesn’t move from 5.4V. If I could send an e-beer I would :) http://www.binarydinosaurs.co.uk/atariSTPSUfixed.jpg (dodgy disk in the A drive there, it’s currently running Bubble Bobble) — Adrian/Witchy Binary Dinosaurs - Celebrating Computing History from 1972 onwards