Sansui G-22000 Restoration/Upgrade Thread

Discussion in 'Exclusively Sansui' started by mattsd, Oct 28, 2018.

  1. bktheking

    bktheking Gitter Done! Subscriber

    Messages:
    12,991
    Location:
    Ottawa Ontario
    Nice, ya after I wrote you I was thinking relays, then i just changed the 2 on the right channel lol.
     
    kevzep likes this.

     

    Please register to disable this ad.

  2. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Here are the motherboards!

    Motherboard for Flat Amp Boards (F-2888)

    The motherboards are pretty simple, they hold the local decoupling caps along with the bypass caps for aforementioned caps.

    [​IMG]

    [​IMG]

    The local decoupling caps for the flat amp boards are C07 and C08, they are 470uF 35v. They were upgraded to Panasonic FM 680uF 50v caps. The decoupling bypass caps are .1uF mylar film and were replaced with same value Wima MKP10 polypropylenes. C901 is a 470uF 6.3v and was replaced with a 470uF 16v Panasonic FM.

    [​IMG]

    [​IMG]

    Motherboard for the EQ Amp Boards (F-2884)

    Very similar setup to the flat amp motherboard with the two decoupling caps and bypass caps, however this board has the rotary switches for the function and impedance selectors.

    [​IMG]

    The rotary switches for this unit were pretty heavily oxidized, so I went ahead and pulled them both and cleaned them with a ultrasonic cleaner.

    Before:

    [​IMG]

    [​IMG]

    And after! I wasn't able to get them perfectly shiny, but they look a lot better than before.

    [​IMG]

    [​IMG]

    Now for the impedance selector switch. Before:

    [​IMG]

    [​IMG]

    And after! Again, not perfectly shiny but much better than before.

    [​IMG]

    [​IMG]

    And the motherboard after! The decoupling caps C01 and C02 were 470uF 35v like the other board and were upgraded to 680uF 50v. The decoupling bypass caps (C03 and C04) are .1uF mylar film and were replaced with .1uF Wima MKP10.

    [​IMG]

    Up in the next post, the big power supply board!
     
    Last edited: Dec 12, 2018
  3. kevzep

    kevzep Its all about the Music Subscriber

    Looking good Matt, this is where I am headed with my G22K, the good old rotary and lever switches...I suspect mine are heavily oxidised like that....
     
    Ronald.C likes this.
  4. txturbo

    txturbo AK Subscriber Subscriber

    Messages:
    930
    Location:
    Austin, TX
    Nice! I have one for another AK'er thats ready to get the treatment.

    :lurk::lurk::lurk:
     
  5. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Power Supply Board (F-2891)

    This board is the hardest one to get to out of all the boards in this unit and requires a considerable amount of disassembly to be able to access properly. The board itself is fairly easy to work on however.

    [​IMG]

    C03, C10, C11 and C17 are 1000uF 50v, replacements were 1000uF 50v Panasonic FM. Interestingly, C10 and C11 in the +34v regulator are 1000uF 50v, but only C03 (C10) in the -34v regulator is a 1000uF 50v. C04 (C11) was only a 470uF 50v, even though the setup is identical between the regulators. C04 was upgraded to a 1000uF 50v to match C03. The regulator transistors were replaced with KSB596 and KSD526. The caps shunting the zener diodes, C06, C13 and C21 are all 100uF 10v and were replaced with 100uF 25v Panasonic FM. C07 and C14 were 1uF 50v electrolytic and were replaced with 1uF 50v Wima film. C28 and C29 were 2.2uF 50v and were replaced the same value Nichicon PW. C18 was a 470uF 50v and its replacement was of the same value. C19 is a 220uF 35v and was replaced with a 220uF 50v. C22 was a 3.3uF 50v and was replaced with the same value. C27 was a 220uF 50v and was replaced with the same value. C30 was a 100uF 16v and was replaced with a 100uF 25v. C31 and C32 were 100uF 35v and were replaced with 100uF 50v. The zeners ZD01, 02 and 03 are all 6.2v 500mW and were replaced with Vishay 2% zeners. Relay replacement was a Omron MY4. 3 glue damaged metal oxide resistors were replaced, R01, 08 and 17, also replaced was R27, the 330 ohm 1/2w fuse resistor that is a common cause of the power meters not working in this unit. Replacement was a 1/2w metal oxide.

    [​IMG]

    Bonus, cleaning the pre-main switch. These switches are notorious for causing drop outs in one channel or the other, even after being cleaned with Deoxit. Only true way to make sure they don't cause trouble is to take them apart and clean the contacts manually.

    [​IMG]

    [​IMG]

    [​IMG]

    [​IMG]

    [​IMG]

    Just two boards left in this preamp unit! Tuner board and the meter amp board, after that we will move on to the power amp section. Stay tuned!
     
  6. kevzep

    kevzep Its all about the Music Subscriber

    How on earth did you get the separated/connected switch out? I am very impressed!!
     

     

    Please register to disable this ad.

  7. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Very carefully, it's a pain in the butt to get to, that's for sure. You first have to take off the big gray trim/frame piece and the black metal faceplate panel, then with the edge of a jewelers flat-blade work the 4 black push pins out that hold that board on. This is delicate work and it's not hard to damage or destroy them, thankfully I've had to remove several before on Sherwood and Pioneer receivers, so I've had practice. Even after those are all loose it doesn't have much slack to move, you have to create it. After that there just enough wiggle to turn the board to pull the switch out after you de-solder it. Funny enough taking apart the switch itself is the easiest part. One note though, be very sure you note exactly how it comes apart, you can put it together backwards.
     
    Hyperion likes this.
  8. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    49,307
    Location:
    Hertfordshire, UK
    Yeap, the overall procedure sounds familiar, although I have never worked on a G22K. :) I even have a stock of those black plastic 'push pin' fixings just in case I manage to damage one. :)
     
    Last edited: Nov 18, 2018
  9. kevzep

    kevzep Its all about the Music Subscriber

    Thanks for the explanation, I will be doing this when my pre-amp comes out of service for a bit of maintenance...
    I figured it would not be easy.
     
  10. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Last two boards!

    Tuner Board (F-2920)

    [​IMG]

    And after. This board is fairly straight forward, I upgraded all the signal path caps to film where possible and all caps 1uF and under were also changed to film.

    [​IMG]

    Peak Meter Board (F-2892)

    [​IMG]

    And after! Panasonic FM and FC were used for the blue caps and Nichicon KL was used for the orange low noise caps. Note the trimpots I replaced after the photo was taken since I forgot to include them in my parts list. :oops:

    [​IMG]

    I went ahead and fired it up by itself just to test function, everything worked great with no problems noted, distortion measured .0022%! out of the preamp with a 1KHz tone. I also aligned the tuner at this point and yes @kevzep, this thing has got a stupendous tuner and it's very easy to align, this one was a joy as it appears it's never been touched and it was very clean as well, so no sticky or intermittent coils. I was able to improve the sensitivity a bit and get the dial aligned back to normal. But the distortion. Let us talk about the distortion. HOLY FUDGEPICKLES. After fine tuning it as far as it would go, at 65dBf with a 1KHz tone, the distortion was .0162% THD+N. :eek2: I didn't even know a tuner could get that low! Stereo separation was 50.57dB. Trying it out on actual stations, the selectivity and sensitivity is amazing, this is easily the best tuner I own. Can't wait to get the power amp done!
     
    70'sMusic, Hyperion and Hipocrates like this.
  11. Hipocrates

    Hipocrates Anti-Muppet Subscriber

    Messages:
    1,862
    Location:
    Costa Rica
    Man... what a joy of write up, thanks!
     

     

    Please register to disable this ad.

  12. kevzep

    kevzep Its all about the Music Subscriber

    Oh my word yes it does!! I had the same experience, I thought something was wrong with my Distortion analyser, I actually checked it as I could not believe what I was seeing.....Its an extraordinary tuner....easy to align.....I like the way the Zero adjust works on the tune meter, "move the stub".
    Yeah agreed, the rest is pretty easy and it responds well to adjustments...

    Awesome write up Matt.....:thumbsup:
     
  13. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Starting on the power amp! We will start with the driver boards.

    Driver Amp Board L-ch and R-ch (F-2900, F-2901)

    [​IMG]

    [​IMG]

    [​IMG]

    [​IMG]

    There is a lot of the glue on the driver boards, it unfortunately ate some resistors, so they will be getting replaced. Below is the right amp board, but both were about as bad.

    [​IMG]

    [​IMG]

    Note on the G-22000 and 33000 the thermal tracking is done on the drivers, so it's a good idea to redo the thermal grease on the drivers. I noted much better thermal tracking on the left channel when I replaced the bad drivers compared to the right channel, which was stock. And no wonder:

    [​IMG]

    [​IMG]

    And after! The local smoothing/decoupling caps C05 and C06 were originally 220uF 100v and were upgraded to 680uF 100v. C04 shunts ZD01 and it is rated at 100uF 50v. It was replaced with the same value. C03 shunts ZD02 and is rated at 100uF 16v. It was replaced with a 100uF 25v. C26 was a 1uF 50v and was replaced with a Wima 1uF film. The film bypass C14 was removed. C901 and C902 which shunt D01 and D02 are mounted on the back of the board and are 1uF 50v. Replacements were Wima 1uF film, the bypass caps C23 and C24 were removed and the 1uF caps installed in their place. The decoupling/bypass film caps C07, 08, 19 and 20 are all rated at .47uF 250v. They were replaced with a Wima MKP10 of the same value.

    ZD01 is a 27v 500mW zener and ZD02 is a 12v 500mW zener, replacements were both Vishay 2% tolerance zeners. VR01 and VR02 were replaced with Bourns single turn pots. Several resistors affected by the glue were also replaced.

    [​IMG]

    [​IMG]

    Up next, the two power supply boards and the filter caps. Getting close!
     
    Ds2000, Hipocrates, cnyel1967 and 2 others like this.
  14. Verde

    Verde AK Subscriber Subscriber

    Messages:
    18
    Mattsd,
    Thanks for doing the great write up with photos! It's very interesting to see and understand what makes the bid G special.
    Verde
     
  15. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    Power Supply Board (L-CH and R-CH) (F-2902 and F-2903)

    [​IMG]

    [​IMG]

    The left and right power supply boards on this unit were a bit of a mess, a lot of dust and dirt along with glue issues. These boards aren't too difficult to access but it requires you to unplug all the connectors that can be unplugged and two more de-soldered to be able to flip the board over. These boards have a couple of unusual caps, C14 and C19 which are the local de-coupling/smoothing caps and are rated at 470uF 120v. Somewhat hard to find as the overall height of the cap cannot be more than 40mm. 45mm would put it uncomfortably close to the bottom panel. I was able to find some Nichicon 470uF 160v caps that worked perfectly.

    For the other caps, I went with stock values, upping voltages here and there. For the left and right channel PS boards, C15, C20, C29 and C30 are 100uF 100v and were replaced with the same value. C17 and C22 which shunt ZD01 and ZD02 were 100uF 35v and were replaced with 100uF 50v. C18 and C23 are 3.3uF 100v and were replaced with the same value. C24 on the left channel is a 220uF 50v and was replaced with the same value. C24 on the right channel is a 1uF 50v and was replaced with a 1uF Wima film. C27 on both boards was a 100uF 6.3v bi-polar and was replaced with a 100uF 16v Nichicon EP bi-polar. I used a EP as it has a 105C temperature rating vs the 85C rating of the ES series. C31 is only on the left channel, it shunts ZD03 and is rated at 47uF 50v and was replaced with the same value. In addition, the right channel adds C11, C12 and C13. C11 and C12 are 100uF 25v and were replaced with the same value. C13 was a 470uF 10v and was replaced with a 470uF 16v.

    There are 3 zener diodes per board, the left channel has three 24v diodes and the right channel has two 24v and one 6.2v. All were replaced with Vishay 2% tolerance zeners. The right channel also has a MV-12 diode that was replaced with two 1N4148 diodes in series. Both boards have 3 fuse resistors, R08, R16 and R44. R08 and 16 are 47 ohm 1/2w and were replaced with the same value metal film. R44 is a 4.7 ohm 1/4w and was replaced with the same value metal film. Several glue damaged resistors and a polystyrene cap were also replaced. All the fuses were also replaced with ceramic equivalents.

    [​IMG]

    [​IMG]

    The filter caps are straight forward, pretty easy to do, though the space is a little cramped.

    [​IMG]

    [​IMG]

    Two of the filter caps showed significant leakage, both look like the one below:

    [​IMG]

    And after! I also cleaned the chassis, it looks much, much better than before.

    [​IMG]

    [​IMG]

    Details on initial power up and impressions coming shortly! I will also be doing a full performance test, I recently got a new toy to help with that. (This is in loopback, but yes, that is .0003% distortion)

    [​IMG]
     
    Last edited: Dec 12, 2018
  16. ossodiseppia

    ossodiseppia Well-Known Member

    Messages:
    801
    Location:
    Broomfield, Colorado U.S.A
    This is some really nice work. @mattsd, I was wondering if you are planning on flowing or re-flowing the molex connectors on the boards. I know the flat amp boards hang upside down on these things and can jiggle around when the unit is moved or jostled.
     

     

    Please register to disable this ad.

  17. kevzep

    kevzep Its all about the Music Subscriber

    I have never found the Molex connectors to be much of an issue on the Sansui's but if I'm in there I often do it anyway.
    Very nice job Matt, she was a bit dirty inside right!! Hilarious THD!! Just a testament to the Sansui designers.....
    Off topic, have you found a place to get Pansonic service manuals for the line of gear like your THD analyser there? I have a Panasonic Multiplex generator of the same vintage, and although its functioning fine, I'd like to find a service manual.....
     
  18. LBPete

    LBPete Rolling Along Subscriber

    Messages:
    11,640
    Location:
    Long Beach, CA
    Absolutely outstanding work. A real pleasure to read this thread.

    - Pete
     
    70'sMusic likes this.
  19. ossodiseppia

    ossodiseppia Well-Known Member

    Messages:
    801
    Location:
    Broomfield, Colorado U.S.A
    Thanks, Kevin. I was just wondering how most approach this.
     
  20. mattsd

    mattsd AK Subscriber Subscriber

    Messages:
    1,662
    Location:
    Cincinnati, Ohio
    I redid the power amp board connectors, but I didn't touch any of the others. For the preamp, each of the boards that plug into the motherboards has a foam strip in line with it attached to the bottom plate that pushes down on the top of those boards and keeps them in place, as well as damping any vibration.

    I forgot to mention that distortion measurement is with the analyzer in loopback lol. That would be amazing if it was that low from the receiver though! I'll PM you on the generator.
     

Share This Page