AU-9900 Function Switch

LBPete

Rolling Along
Subscriber
I'm working on an AU-9900 that came to me with no output at all. It powers up, comes out of protection but no sound from either channel. A quick check show no output from the preamp. I've isolated it to the function switch on the F-2569 input board. The signal path goes directly from the aux input to pins on the function switch but never gets out of the function switch. I have the board out and I'm assuming

Has anyone run into this before? This is one complected input signal path. Before I take the function switch off the board and tear it apart, I want to make sure that the dropout is not being caused elsewhere in this complected signal path.

Here's a clip of the schematic.
Screenshot_2018-11-24 hfe_sansui_au-9900_schematics pdf.png
There are three boards involved. The schematic shows some of these boards twice to depicting the two channels separately.

The function switch and the input RCA terminals are on F-2569. Boards F-2571 and F2572, which have the tape play switches and a separate tuner on off switch, are soldered to F-2569 to form a cluster of three boards.

The function switch has positions for two phono inputs and two aux inputs. The tuner is selected with a separate toggle switch. It doesn't appear to matter what position the function switch is in, flipping the tuner switch appears to bypasses it. Can a problem with the tuner switch drag down the input signal?

- Pete
 
Talking about these three PCB's, I worked on several AU’s-9900/11000 and CA’s-2000,and I had,.if remember correctly, one similar situation with CA which completely disappeared after disassembly and thorough cleaning and lubrication of all switches.

Pete,If you need photos, I can put them in a few hours.

:smoke:
 
Thanks Pepa. I have the boards out on the bench. I've thoroughly cleaned the switch and it hasn't helped. I'm feeding it a 1k Hz sine wave at the aux input and can see it at the input pin on the switch itself, but not at the output. I don't see any damage to the switch. I can see the mechanism inside move when turning the shaft. It appears to move normally and it doesn't seem to be contaminated with anything. It should work but it doesn't. I'll check it again this morning and if it's still dead, I'll pull it off the board and take it apart. Stay tuned.

- Pete
 
I'll pull it off the board and take it apart. Stay tuned.
You might have a hairline crack on the board perhaps? - or even an oxidised rivet on one of the switch wafers?
If you don't get any results with an oscilloscope and sig gen, personally I would use a multimeter until I get continuity. ;)
 
No cracks, no rivets. The function switch is one of those rotary actuated linear switches. the pins solder directly to the PC board, no crimped on solder lugs. It's mounted directly on the same board that the input RCA jacks are soldered to. There's a single trace for each input between the RCA jack and a pin on the function switch. Those traces are good. There is a pair of shielded wires that run from the function switch to the EQ board. The aux inputs connect to the EQ board through those wires. I see the signal go into the switch pin but nothing at the connection to the shielded wires. Those solder joints look clean and tight. It's got to be inside the switch or the way the switch is soldered to the board.

- Pete
 
Here's an update. Turns out I was totally misreading the schematic and relying too much on experience with other units and not enough on actual observation. I pulled the switch off the board and took it apart. There wasn't anything wrong with it but seeing how the contacts were arranged gave me a new perspective on the signal path.

The switch has four pairs of four contacts with a separate common terminal. The aux input hits the function switch in the same in the same four contact pair as a pair of Blue and Gray shielded wires that go to the EQ board so I incorrectly assumed that the switch contacts connected the aux input to those wires. Nope. Those switch contacts ground the aux input when in the phono position. On like many receivers, the aux input signal doesn't go through the EQ board at all.

Once getting that false start behind me, I started a serious study of the signal path and holy crap, is it complected. It goes back and forth between the F-2573 and F-2574 boards and unlike most other Sansui equipment, it is not consistent using the blue and gray wire to carry the signal. This unit is totally loosing the signal for both channels in the preamp. I think I have it isolated to the F-2574 board, The signal disappears after the volume control. It's looking like the tone defeat switch is the culprit. Seems odd that both channels would be affected but who knows what was sprayed into this thin. I'm hoping it will clean up because to take it off the board a rat's nest of shielded wires needs to be disconnected and soldered back.

- Pete
 
I ran into something like this. I think in fact it may have been the same symptoms. Behind the face plate there is a screw that mounts one of the switches. It provides the ground for some circuits in the amp and if it gets loose you lose audio. Look at my restore thread. I install a ground braid so you never have to worry about the screw.
 
Thanks Warren, I'll take a look later today. I did notice that a couple screws had star lock washers on them. Hadn't seen that before in a Sansui product. I assumed they weren't original but that may be an attempt for a better ground.

The chassis on this thing is a complex bit of work all by itself. The F-2573 board mounts to a removable chassis panel that fits into the front of the chassis. It allows removal of that board without having to undo all the nuts and screws that hold a gang of switches on that board. I'll make sure everything is grounded.

- Pete
 
Warren, is this the screw you are talking about?

screw.jpg

Chassis.jpg

Here's is the trio of boards. That screw is an attachment point for a separate bracket that surrounds this trio of boards. The boards themselves should ground through the big nuts that secure the switch shafts.

Board trio.jpg

I've tightened all the screw and nuts that hold the switches and pots in place, installed jumper wires to ensure the various chassis parts are grounded and there is no difference. Still no output from either channel.

- Pete
 
Dang. Well, that is the screw I was talking about, and it does complete the ground path. Took me a few hours of work to find it. Break out the scope, then.
 
It's just a typical chassis screw but it does secure a sub section of the chassis. I've had issues in the past with a hum through the speakers if the nuts that hold the pots aren't tight. The case of the pot must act as a shield and if it's not grounded, you get a 60Hz hum.

- Pete
 
My bad, that's not the screw I meant. I looked at the picture again. The one I meant is next to the lever switch to the left of the screw that is marked.
 
Here's the screw and braid locations:
yKSR2Y2.jpg

To add the braid, scrape the soldermask off the board and scrape the plating off the inner front panel where shown below:
Oc1jO79.jpg

Now solder a braid to both places. Use a large soldering iron tip or a solder gun but do not overheat the board. Apply solder to the chassis first, using plenty of heat to make sure the solder flows onto the chassis. Tin the braid and solder in place.
9prb0qW.jpg
 
That screw holds the switch that enables the tuner. It's on one of the boards of the trio cluster. Those boards are totally passive, no power supply to any of them. I suppose it is possible that a ground wire from another board gets grounded through that one. In any case, thanks for digging out the pictures.

It's looking now like the signal is getting to the relay on the connect separate board. There's power to the relay but nothing coming through it. I definitely took the long way around to get to this point. I need to play with the relay to see if I can get it to work.

- Pete
 
This thing has taken an interesting turn. Today, it working normally on both channels. That said, there is a story behind it. In the post above, the signal was getting to the reed relay on the connected/separated board. The relay was definitely not turning on. I pulled the relay off the board and tested it on the bench. It turned on and was making contact on both channels. So I put it back in. Then it wouldn't turn on when the connected switch was in the connected position. Turn the connected swicht off and the reed relay would turn on. Switch the connected switch on and the relay would open.

It was starting to look like a power supply problem. The reed relay is wired in parallel with the protection relay. They both should turn on and off simultaneously. Monitoring the voltage across the relay coil it's only getting not quite 18v. The protection relay clicks every time but this little reed relay is more finicky.

The relays are controlled by the F2568 board. TR10 grounds pin 16 based on the operation of the protection circuit. There is 37 volts at pin 14 when the power is first turn on and it drops to 18V when the protection circuit grounds the relay coil putting a load on the circuit. Pin 16 shows a 0.5 volt ground level when the relay is on. That seems high. I would expect it to be closer to zero. I don't know if these values are normal or not. I've checked R21 and it is OK at 10 ohms.

Today it's working every time and plays through both channels. Still seeing the same voltage values. I suppose the reed relay may have been sticky. I plan to replace it to be sure.

- Pete
 
Hi Pete,

Interesting development. A question - have you done any recapping on the unit? The reason I ask is that I have found "flaky" protection circuit operation due to the failure of the filter cap on the main protection circuit supply - causing excessive ripple when the relay is engaged. In this case I would check/replace C607 (33uf 50V) on F2567. You could check the ripple on the supply line (pin 27/28 on F2567 or pin 12/13 on F2568). It is typical for the protection circuit to have a low value filter cap on the main supply (in order for the relay to disengage quickly at turn off) and the cap is therefore working at a high ripple current in normal operation (hope that makes sense!)

Cheers

John
 
hope that makes sense!
Yes, that does make sense John. ;)

If it is working 'on the edge' like that, it's correct operation could also be subject to small (normal) daily/hourly changes in mains voltage some areas experience from time to time.
 
Today it doesn't come out of protection when the connected/separated switch is in the connected position. switch it off, it comes out of protection, then switch it back on and the relays stay engaged. This unit came to me as a repair. It has had caps replaced. The F-2567 has all new electrolytics.

Here's what I'm seeing at pin 12/13 on the F-2568 board. It's about 1.3 volts.

Ripple 1.jpg

- Pete
 
Hi Pete,

Another thought, check the two rectifier diodes associated with the protect relay supply (D605 and D606 on F2567). I have seen similar symptoms when one of these diodes has gone open circuit.

Cheers

John
 
Back
Top Bottom