B-2102 fault finding

[Hyperion]

I have owned a Sansui B-2102 for about 5 years, and to be honest haven't used it much. I have given it a somewhat patchy restore, but there was a problem, not serious, with setting DC offset on one channel.

I had a few cracks at solving it when I have had time, so recently I decided I would have another go. My investigations focussed on one of the F-4618 pre-driver boards. I did a lot of comparing the Left and Right channel boards - which are identical, and changed quite a few components on the faulty one - without success. The problem showed itself by exhibiting too much DC offset (~240mV) on the offending channel, without triggering protection and after adjustment, it would creep down to less than 20mV over 10-15 minutes.

Yesterday I found the problem - a factory fit wrong value resistor on the faulty F-4618 board, fitted was a 1.2KΩ when it should have been 1KΩ. Note that I found the fault 'visually' just by studying the board closely.

Here is the board layout - the wrong value resistor location (R41) is ringed in red - its opposite number (R42) is ringed in green.



And this is the schematic



Notice how this balanced arrangement of components to deliver 0V at the output would have been influenced by a wrong value - thus unbalancing the circuit voltages and causing the DC offset setting with the trimmer VR1 to be incorrect. What made it difficult to find was that the wrong value was only slightly wrong, meaning that the difference could be adjusted out. But the tell tale was in the warm-up drift characteristic - plainly different when comparing channels.

That's all folks.

 

[MIJ_Pioneer]

Nice one!

I believe most would have been stumped by this. Even if you removed a leg on each resistor and measure resistance would the ~200ohm have raised a flag? I think not for many. Especially since our friends at Sansui are known to make changes on the fly during production. It shows you have an excellent understanding of the circuit.

Thanks for sharing!

 

[Sansuiman]

That's the mark of a true tech, versus a mere parts swapper. Almost anyone can do the latter, the former requires the ability to "simulate" the circuit function in one"s mind to know intuitively that there is an improper component in play. And a sharp eye doesn't hurt.

I'm surprised at the error, I wonder how often that sort of thing happened, especially where hand assembly was involved in producing the boards versus machine compoment insertion. I don't think Sansui ever did the latter, but have no way of being certain.

 

[GrisPato]

So many thanks for sharing this with us. Especially for me as a beginner this is valuable information. Although it's really good to learn from your own search and mistakes it's good too learn from the experience (s) from the more experienced ones.
And I am for sure going to check this in my B2102 when it's his time. .
Appreciated

 

[Hyperion]

Thank you, - although I often do visual checks for faults this mistake somehow eluded me much to my annoyance. Due I think, to its position on the PCB and some jumper wires obscuring the view - well that's my story anyway.

 

[GrisPato]

By the way seems like you have a thing with dc-offset and components in Sansui's. I was reading up for my next project, my CA-F1. In your restoration of one I was reading you encountered a wrongly installed trimmer also after a thorough visual inspection.

 

[Hyperion]

By the way seems like you have a thing with dc-offset and components in Sansui's

Yes I do - the trimmer issue was once again a factory installation error where the trimmer had been installed incorrectly - rotated by 120° so the three pins for the track and slider were presented in the wrong positions on the board. When I work on amplifiers I make absolutely sure the DC conditions are correct first, i.e. all currents and voltages, before even thinking about the AC = 'signal' conditions.

 

[Ray Gianelli]

Thank you, - although I often do visual checks for faults this mistake somehow eluded me much to my annoyance. Due I think, to its position on the PCB and some jumper wires obscuring the view - well that's my story anyway.

It's really not the kind of thing you'd expect. My visual examinations are usually for burned parts or defective solder joints. If there's no fresh soldering I wouldn't suspect a wrong value of component. Good catch John!

 

[Skywatcher]

Nice one!

I believe most would have been stumped by this. Even if you removed a leg on each resistor and measure resistance would the ~200ohm have raised a flag? I think not for many. Especially since our friends at Sansui are known to make changes on the fly during production. It shows you have an excellent understanding of the circuit.

Thanks for sharing!

It would have for me. I would have also checked it's "brother" to verify. 200 Ohms is about 18% of 1.2k. A bit more than 5%.