Switching PSU Theory of Operation / Repair Tips

From: Douglas Quebbeman <dhquebbeman_at_theestopinalgroup.com>
Date: Sun Mar 17 08:25:34 2002

Is this a follow-up to my thread about my dead Prime's PSU?
if so, thanks, but that was a long time ago, I bought a replacement
PSU rather than fix the old one.
Damn thing still doesn't work but will maybe soon, we're debugging
through it...

-----Original Message-----
From: John Salmon [mailto:reversebias_at_talk21.com]
Sent: Sunday, March 17, 2002 6:38 AM
To: classiccmp_at_classiccmp.org
Subject: Re: Switching PSU Theory of Operation / Repair Tips

Re Where does the 600v come from?
The SMPSU in question is probably of the 'Flyback' or of the 'Forward
Converter' type. In either case the converter O/P Tx primary winding is
connected between the +300vdc rail and the collector of the switching
transistor. The emitter of the switching transistor is connected by a small
resistor to the 0v (-tve) rail.
During the conducting period of the switching transistor (lets call it TR4 -
it often is) the potential across TR4 is very low, a few volts. Hence
virtually all of the 300vdc appears across the primary winding of the O/P
converter Tx (lets call it Tx1). The collector of TR4 is therefore at -300vdc.

When TR4 is driven into its 'off' period Lenz's Law causes the primary
inductance of Tx1 to try to maintain the primary current. The resultant
induced emf (due to collapse of primary flux) now produces a +tve potential at
the Tr4 collector and a -tve potential at the +dc rail end of Tx1 primary.
Hence the Tr4 collector is now 300vdc positive to the +300vdc rail.
Therefore the collector of Tr4 is now 600v +tve to the 0v (-tve) rail ofv the
PSU. Don't touch its Heatsink while the PSU is switched on. Also allow 4 or 5
mins for the voltages to bleed down after switching the PSU off before
touching internal components. (It should give this warning on its nameplate)
Hope this is clear and useful - it is 3.30am in the morning here in UK.
John Salmon
Q&R Engineer
Received on Sun Mar 17 2002 - 08:25:34 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:35:12 BST