Thanks for that. The slow start one sounds like an interesting avenue to explore. I have been told some of the capacitors on these tend to be bad. Finding the capacitor (if it does indeed do a slow start) will be a challenge for me. I have been reverse engineering the schematic a bit, but some of the parts are hard to see as they are hidden by a heatsink which is not easily removed...
In doing the reverse engineering I have spotted a 10R resistor near the output section that is relatively high power that seems to have some physical damage (it is one of the hard to see components). It is marked as 10R and measures 10R (in circuit), but measures 0V across it when the PSU is powered up. I may remove it to check it over. Regards Rob > -----Original Message----- > From: cctech [mailto:cctech-boun...@classiccmp.org] On Behalf Of shadoooo > via cctech > Sent: 18 September 2017 18:42 > To: cct...@classiccmp.org > Subject: Re: H7878 Fails Under Even Moderate Load > > hello, > well it's difficult without the piece in hands to understand the circuit, and > without the schematic, but I could give a try. > > 1) > On a switching PSU, only one output is really stabilized, here it seems 5v. > The other outputs are "unregulated", but in some way related to the stabilized > one, for example via transformer windings ratio. > However, to protect circuits, some voltage range or current range detectors > are added, to avoid these outputs to go outside acceptable values. Maybe > something in these circuits on secondary section detects a false condition of > error, cutting off supply current on primary section. > > 2) > voltage seems to go high too slowly, as if a soft start is going too slow. > Normally SS is done charging via a resistor a capacitor on the feedback, > affecting maximum current on the primary. > If the capacitor is leaking current, it will charge too slowly, and will > never reach > maximum voltage required, so basically you will never have maximum current > in the output. > This could be a small capacitor charged by a done kohm resistor. > > Andrea
