1. Time for some upgrades in server hardware and software to enhance security and take AK to the next level. Please contribute what you can to sales@audiokarma.org at PayPal.com - Thanks from the AK Team
    Dismiss Notice

AU-717 PSU voltages

Discussion in 'Exclusively Sansui' started by afroaudio, Jun 9, 2018.

  1. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Hi gang,

    I am in the process of restoring my AU717. I have so far completed the main amp boards (amp tested good after these were completed)

    This week I managed to get my 2663 power supply board out and recapped. I also decided to replace all diodes with UF4004/1N4148 where appropriate as well as the Zeners (paid attention to 5.6, 6.2 and 13V versions) and the VD1212. Also replaced TR603 and the relay with an Omron MY02. Main filter caps also replaced with 15000/63 Nichicons.

    I managed to get the PSU board and main caps back into the chassis and wired up today.. took me the entire morning! Powered up on a dim bulb which looked good but now measuring voltages at the top of the F-2663 board I am getting the following:

    Between point 21 (gnd) and 23: - 34.8V
    Between point 21 (gnd) and 19: +35.8V

    Between point 22 (gnd) and 24: -15V varying to - 18V approx
    Between point 22 (gnd) and 20: +15V varying to +18V approx

    I checked the DC at the speaker terminals and set to near zero mV.. all good
    Bias adjusts fine and is set at 20mV as per service manual.

    Both channels of the amp play fine but the volume on Left channel is much lower than Right. I have a horrible feeling I will be needing to undo all my work today and take the 2663 board out again and figure out what went wrong there. Any pointers at trouble shooting would be great!
     

     

    Please register to disable this ad.

  2. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    Check your work in the PSU with low output, usual things - capacitor polarity, semiconductor orientation and so on, and also check that you put the grounds back in the right place when you changed the main PSU capacitors? The regulated PSU depends on having correct ground connections for it to work properly, these are located at the Main PSU capacitors, as you have both supply rails +ve & -ve in trouble this is my bet for what is wrong. (Probably not the issue here as you say it plays ok, but make sure all the connectors to the various boards are plugged in too - as there is a ground path completed when they are all plugged in).

    Note that the negative regulator is a 'tracking regulator' - and will only work if the positive regulator is working. Therefore concentrate your efforts in getting the positive regulator working first. I am going to 'stick my neck out' here to say that it is probably a ground fault as described.

    Let us know how you get on...
     
    Last edited: Jun 9, 2018
    afroaudio and SansuiJavier like this.
  3. matejp

    matejp New Member

    Messages:
    3
    Hi,
    Im buying au717 for 650€.
    Seller recaped the nice looking unit with this componets:
    Elna Silmic II
    NIchicon FG,KG
    Wima FKP2, MKP10
    Are they any good?
    Thank you for replies.
     
  4. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    The components are OK - but does the seller know where/how to use them, and what is his/her workmanship like?

    You should start your own thread - this thread is about a PSU issue - nothing to do with buying a recapped AU-717 - thank you.
     
    afroaudio likes this.
  5. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Hi John and other readers

    I took a look this morning with fresher eyes and followed all the hook ups I made yesterday for the power supply which all seemed to be correct.
    I decided to power on again to take another set of measurements. Now to my surprise all the voltages seemed to be right! between +/- 34V to 35V each side.

    I then hooked up speakers and a source into AUX and played music for 5 - 10 minutes.. I then attach the meter again and bang.. the one side of the supply is back to +/- 18V :(

    This is making me think I have some bad diodes or transistors in the 2663 board? Not sure what could cause this behaviour!
    I had already swapped out TR1/2/3/4 for MJE15032/MJE15033 and apart from those and a new KSC945 (TR603) all other transistors were left original.

    Jonathan
     
  6. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    I assume there wasn't an actual 'bang' when you lost the +/-18V ? - if you can't see anything burned - then I suspect a dry joint, dirty connector, or something similar. Check the connections from the transformer '02' & '06' to the bridge rectifier comprising D06/D08/D10/D12 - and/or measure the output from the bridge rectifier?

    Try tapping the board with the blunt end of a screwdriver? - if you get a reaction from this, try gently 'nudging' each of the components on the failing regulator to see if you get a reaction.
     
    Last edited: Jun 10, 2018
    afroaudio likes this.

     

    Please register to disable this ad.

  7. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Thankfully no actual BANG :)

    Appreciate the pointers John.. I have gone over everything and poked around on the board and those voltages keep fluctuating on the bad side..anywhere between 18V and 33V. I am going to unplug clean and replug all the connectors and have removed, cleaned and refitted the two grounding tags which screw directly to the chassis. Praying that its a bad plug contact or something simple I might have missed and I can avoid the pain of pulling that power supply board out again..
     
  8. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Reporting back with more frustrations unfortunately!
    I checked and rechecked the wiring of all grounds and connections to the main amp boards/power supply board/transformers and all connections between boards. Nothing changed..
    After this check I powered up again on dim bulb which operated as expected. I then bypassed the dim bulb and powered up again intent on measuring voltages once again to see if had changed. Thats when the smoke happened! Coming from the right side of the power supply board (ie the side that was measuring correctly) very quick power down of course.

    Yesterday I removed the power suppy board again to recheck all diode cap and transistor orientation and noted that the 2 x 100ohm MOX resistors R05 and R15 looked slightly dark and the most likely suspects for the smoke. These were then pulled and replaced on both sides. Everything else on that board looks as it should be.. no cracks, no solder bridges, everything correct orientation... I decided to replace all the small transistors while I was in there with KSC1845/KSA992. The only original transistor remaining is 2SA733 which I pulled and measured fine.

    After reinstalling the board, double, triple checking all wiring... I power up again on dim bulb, now the protection LED flashes for longer than it used to, about 10 seconds.. then goes solid BUT no relay click! Even on dim bulb I can also detect there is some heat coming from the right side of the 2663 board again and the resistors R05/R15 are quite warm to the touch (left side totally cold)

    Feeling a little out of my depth trouble shooting this... Once again any guidance is most welcome! :)

    [​IMG]20180607_145622 by Bregtje Cals, on Flickr
     
  9. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    All I can suggest is that you connect the PSU board up, but not attach it to the chassis with the brackets, and lay it over somehow, (component side up) so you can use your multimeter to compare voltages of the working regulator to the non-working one - make sure nothing can short out. You have changed so many components that it would take a genius or a psychic to know where you might have made a mistake.

    I recommend you disconnect the +ve and -ve outputs of both PSU's while you compare voltages (pin17/19, 23/25 & pin18/20, 24/26) - but make sure you connect the grounds (pin21/22). And only reconnect those PSU outputs when you are sure things are working correctly, i.e. both with the same output voltages.

    The increased time for the protection light to stop flashing suggests a low voltage at R601 (390Ω 2W), supplied to the protection circuit. Enough to run the multivibrator TR606/607 but not enough to fire the relay when TR603 turns on.

    As usual, let us know how you get on.
     
    afroaudio likes this.
  10. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    Transistor orientaion next to c16 looks incorrect
     
  11. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    In fact I would check them all, most modern transistors looking left to right across the flat with legs towards you are e c b
    Need to verify against board screening. I do not recall any board screen errors on the 717.
     
    afroaudio and Hyperion like this.

     

    Please register to disable this ad.

  12. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    Now that I've realised I can magnify the picture :rolleyes: there does seem to be some transistor orientation inconsistencies.

    You wait ages for a genius or a psychic and then both come along at the same time in the shape of Lee :D
     
    afroaudio, richardpp and Overundr1 like this.
  13. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    Once those transistors are verified I would carefully inspect the soldering around C23 and C25, perhaps even removing them to be sure there are no cap legs touching or solder bridges on the masking side of the board. Cooking R05 indicates something downstream of TR01 is pulling too much current. Board work looks very nice btw :)
    John touched on the ground scheme, here is another way to measure those voltages---

    The +/-33vdc feeds to the phono and tone board are the red/blue and yellow/gray wires visible at the top left and right of the power supply board and are referenced through the grounding plane across those two boards, when you have everything torn out of the unit testing for voltage you will get incorrect values. The method I used was to clip a mini-grabber to the ground pin on each side and a solid chassis ground, THEN measure those wires for proper voltage
     
    afroaudio and richardpp like this.
  14. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Gentlemen thanks so much for your detailed replies and the time taken!

    I went back in this morning and checked my transistor orientations on the regulator board (F-2663) All were in fact correct.. all ECB. I then pulled the C23-C26 caps (220uF 50V) no bridges or dry joints there.. I measured the caps and the ESR seemed a bit high (0.13ohm) on all of them.. these are Panasonic FC's I had some 330uF Panasonic FMs that measured ESR of 0.03ohm... I put those in instead. Following this I checked all my wiring for the umpteenth time against photos I'd taken and checked that the grounds were all intact.

    Before taking the 2663 board out again I just decided to power up on dim bulb again 30 min ago. This time the relay clicked in... good sign.. I then quickly threw my meter onto the speaker terminals and noticed the DC offsets were wildly off.. 200 odd mV. I set these back to 0 or very close and then went on to measure the voltages on 2663 board. I now am getting (still on dim bulb) a voltage of +/-22.5V on both sides.

    I am feeling like things are looking better here, but a bit concerned that I was not able to zero in on what the issues were previously.
    So I have the amp playing now ( still on dim bulb) and sounding fine but am still @*&^% scared to give it full mains voltage!

    ***As a side note I can feel a little heat coming off the F-2723 Equalizer board.. nothing alarming, but this is not a board I expect to get any heat off of... EDIT.. the heat is coming off TR13 (2sc1400) on this board and it is hot enough to burn your finger tips
     
    Last edited: Jun 16, 2018
  15. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    48,563
    Location:
    Hertfordshire, UK
    Well, nothing bad was happening before (no burning) - just a sudden intermittent drop in voltages from one of the regulators. However, I don't remember the EQ Board giving out a particularly noticeable amount of heat - more from the driver boards, and yet more from the PSU/Protector board, as I recall.

    If you take out the DBT, do check the DC offset again - and if you set bias on DBT, that will first need turning down to be on the safe side - and re-setting when on full mains. ;)

    Don't set too much store by what you 'hear' on DBT, its use is mainly for getting the DC conditions right.
     
    Last edited: Jun 16, 2018
    afroaudio and Overundr1 like this.
  16. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    This issue on F -2723 Equalizer board has stalled my progress. TR13 (2SC1400) on this board was getting hot enough to burn your finger tips within a minute of power up. Starting new thread for this issue and will get back to the power supply voltages etc once this is sorted.
     

     

    Please register to disable this ad.

  17. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    If you remove the molex connector next to tr13 does it still overheat? Need to see if the issue is downstream from f2723 board. Will also help explain the power supply voltage sag and no that transistor should not heat up at all.
    While doing that take a moment and check that interconnect cable looking for any abrasions that would allow an internal wire to touch ground
     
  18. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    If it still overheats you will need to follow the +/- wires that feed this board back to the power supply and un solder them, then re-measure your power supply voltages with them out of circuit betting that they will fall in line.
    Then we dig into the reason why
    -Lee
     
  19. afroaudio

    afroaudio Active Member

    Messages:
    157
    Location:
    Cape Town, South Africa
    Hey Lee

    Thanks for your comments here! Looking closer there are 3 Molex type connectors on the EQ board:

    a 3 pin adjacent to TR13 (Grey, White Yellow) connects to top right of regulator board.. all connections to regulator board are solid, no shorts
    a 3 pin lower down (Red, Black Blue) connects to top left of regulatorboard.. all connections solid, no shorts
    a 2 pin (white, red) which connects at top center of regulator board... all connections solid, no shorts

    Just wondering if I should pull both those 3 pins and then power up under those conditions and then check my regulator board voltages and then add them back one at a time? Assuming the 2 pin should stay connected?
     
  20. Overundr1

    Overundr1 AK Subscriber Subscriber

    Messages:
    2,515
    Location:
    Mississippi
    The grey white yellow and the red black blue need to be pulled, then re-measure your power supply voltages and see if they are now within specification. I do not recall which side color wise was giving you trouble, might take before and after measurements on that side. Also, need to be back on the dbt for these tests just in case. If you use a 100w bulb your voltages should be pretty close and still have a failsafe against over-current.
     
    Hyperion likes this.

Share This Page