|
Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment. |
|
Thread Tools |
28th Apr 2022, 9:30 am | #1 |
Diode
Join Date: Apr 2022
Location: Guildford, Surrey, UK.
Posts: 3
|
Musical Fidelity A220 (1998) repair
So I've had this amp since new in 1998 and it's been driving a pair of castle harlechs (8ohm) all it's life. That was until yesterday when I tried a pair of 4ohm speakers and had the board attached to the scope. The probes were on the transformer secondary pre-fuse and on the mute start resistors that drop the voltage from 50V to 24V for the mute relay circuit (the PCB under the zener here shows head damage radiating from the diode.
No schematics but this amp is a toaster oven by reputation. On a second start to capture the mute start voltages with improved settings the main +V rail fuse (250V T6.3A) provided a lighting show. Tracing the obvious dead fuse after removal showed the 400V 10A bridge rectifier has a shorted diode between the AC1 and + pins. The power toroidal secondaries still show 104Vpp without the fuses in - so that's good. So the next component is a 15000uF 50V Jamicon. The rails then split for the channel with each channel has a 3300uF 50V caps so in total I've calculated about 47,000uF hangs off the bridge rectifier without a soft start. In circuit the output MOSFETs seem fine, as do the 4ohm speakers. The one second from the right I have removed from circuit and it tests fine. I think that discolouration is from when it was factory soldered. The C23 cap from circuit holds charge but I don't have a capacitance tester so I can't see if the capacitance is good or a rapid charging capacitance DCR tester (it's 15000uF so a climbing reading could take a while with a multimeter). So today's task when I have time is to remove the main caps and devices to test each one. I'm not too fussed if the amp is unrepairable - I'll repurpose it with some valves however I suspect the bridge rectifier let go out of age. Possibly a near failing cap or the additional 4ohm tipped it over the edge.. It seems to be power side rather then output device side at this stage. |
9th May 2022, 6:03 pm | #2 |
Diode
Join Date: Apr 2022
Location: Guildford, Surrey, UK.
Posts: 3
|
Re: Musical Fidelity A220 (1998) repair
An update - good news.
So with the blown fuse, the bridge rectified tested as a short over the AC-Positive rail. This pointed squarely at the FET's or the caps. The FET's tested good out of circuit so that means it's a cap issue. I assume that the capacitance had dropped low enough that the ripple current and inrush exceeded 6.3A. So regardless of them blocking DC, they needed changing. So here are the changes I have made so far: Preamp 2x 10uF 10V -> Panasonic 10uF 50V EUU FR low ESR these decouple the MC33079 op amp power rails, so I figured 105degC pieces would age better. Main power amp rails etc 2x 0.22uF 35V electrolytic -> Panasonic ECW 0.22uF 400V MPP cap. Ewww - electrolytic coupling cap so that gets replaced, I may change this to a WIMA or increase to 1uF depending on the listening. 2x 15000uF 50V -> CDE 380LX 15000uF 50V 85degC. I picked these because these provide a ripple current of 6A and ESR of 0.020 ohms. 4x 3300uF 50V -> CD 381LX 3300uF 50V 105degC. Picked again due to their 105 rating and lowish ESR for the price. I made a boo-boo here in that they are wider than the originals so they have to be back mounted with hook up solid wire I had at hand. Filter: 2x 470uF 35V -> Panasonic EUU FM 470uF 50V low ESR but not really needed due to it being a filter, however as they're cooked.. they get replaced with a 105degC piece. Mute: 100uF 16V - Wurth 100uF 50V. Good name and £0.20 for the mute circuit. 10uF -> Rubicon 10uF 50V. Good name and £0.20 for the mute circuit. Bridge Rectifier BR104 400V 10A -> SiC BR104 400V 10A. I also have the direct non-SiC replacement pair in case it went pop. Also replaced the thermal pads on the BR and two of the MOSFET's. The fuses. Result - it works After testing with the 8ohm 100W load resistors, it got setup on the Castle Harlechs and it certainly sounds better. Deeper taught base, clear mid range and the treble is good but a little harsh. This is very likely down to the MPP replacement needing time to burn in. My WIMA FKP1 and MKP10s have taken time to burn in before so give it some time. My little python softwareawg worked nicely using the apple mini output as a signal generator, controlled by the SDS1104X-E. At 1/4 turn on the volume: At full volume: Now it's not all good news.. This is the noise I have caused by the mains feed in. So I took the liberty of temporary installing a Schaffner mains filter EIC socket module to see if it made much of a difference - a massive difference. So I will order a Schaffner FN-9262-6-06 line in filter socket that fits in the A220 that doesn't foul against the transformer. The LTSpice model is unstable. I've compared this against the components and the measurements of those components. However I'm thinking that some pF in the right spot may help the design. If I remove the VAS section it all becomes happy.. so the design isn't great to start with If the FET's had blown it would have become a nice solid state circlotron driven by tube and tube front end Only issue doing that is I only use the N-channel and a P-channel for a DC servo. So I get less power.. |