The AU-X1 is, like, so hot right now.

Discussion in 'Exclusively Sansui' started by slimecity, Sep 24, 2017.

  1. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    Another question -

    Im thinking the best way to identify any issues with the power modules would be to mount them where they fit in the chassis one at a time. That way you can also get probe access to the components on the board to test voltages etc.

    Given this amplifier's propensity for oscillating - could anyone forsee an issue with doing this - ie: if one preamplifier channel is not connected to a power module - could that cause issues?

    thanks
     
  2. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    I decided this was not something that could be done easily, and resorted to a combination of careful visual inspection, ohmeter & 'diode' checks, and removal, testing, and changing of components (i.e. transistors) based on the fault being displayed. All this with the respective driver assembly removed - and fully reinstalling for a functional check afterwards - albeit with 'tack' solder joints for speed of removal again if needed.

    I should say that both channels of my X1 were somewhat functional when received, apart from those four 390ohm resistors ;) - but probing with limited access caused me to damage a driver transistor - which I diagnosed and resolved with the above method.

    There didn't seem to be any other viable option as I saw it. :) - and it gave my fault finding skills a bit of a workout too. :D
     
    Last edited: Oct 30, 2017
    Willy6 and slimecity like this.
  3. kevzep

    kevzep Its all about the Music Subscriber

    I just wouldn't be probing around in that amp's driver boards with it powered up to be honest. Too many risks, one slip and its all over....

    I'd roll with John's suggestion.

    Luckily I didn't need to probe around in the one I did....
     
    Hyperion, slimecity and Willy6 like this.
  4. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    Kev you were very lucky - for sure.....I am still having issues. Although Im prepared to be VERY patient as this is a special amplifier.....

    So I have sacrificial outputs mounted in the bias monitoring sockets on both sides - FYI.

    I fired it up again and started taking measurements. The good news is that now I am getting stable bias on both modules.

    The bad news is the crazy jumping around DC offset on both sides - from 0 up to 4 or 5 volts. Once side is slightly more erratic than the other.

    I also disconnected the grey wire leads 12 and 13 to the triac - however this had no effect. Seems I have the same issue which is independant and on both sides.

    I then got annoyed that the DBT was preventing me getting more accurate readings, so I bypassed it - it wasnt glowing bright anyway. After that I got exactly 100VAC being fed into the amp.

    Before I look into this DC offset issue further however - I revisited the power supplies to the power modules and one of these supplies is not right. My understanding is that there are TWO DC powersupplies going to each module:

    - The first one is the supply from the large capacitors - this is fine - around 65-66VDC (from each rail in relation to ground), but

    - The other supply - going to pins 6,7,8 on both modules are both down - Im getting 50VDC on each one (from each rail in relation to ground). I take it this supply is also supposed to be 68V?

    The other thing is that I took measurements on TR01 on the main powersupply/protect board F2781 - instead of having 23V on the emitter of TR01, I am getting 14V. All other readings (for R601 and TR602 are fine).

    Any suggestions on where to look first for the reduced powersupply voltages - if these are indeed down.....this isnt great - Im a bit worried I have a damaged trafo - I will test those secondaries and the bridges tomorrow.....
     
    Last edited: Nov 4, 2017
  5. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    Yes, this (and the supply for the other channel) comes from the rectified & smoothed output from two bridge rectifiers on the PSU/Protector PCB, (D01 & D02). You should be able to measure the AC coming in to the PSU/Protector PCB (from the transformer) and then the rectified DC out again (via 2 x Molex) to the Auxiliary PSU (chassis mounted) smoothing capacitors next to the small transformer, but also on 2 Molex connectors plugged onto the Driver Board PCB's. It isn't unheard of to have a failed bridge rectifier - and I am hoping this is the problem you are seeing, but maybe also check that the connectors are plugged onto the edge of the PSU/Protector PCB in the correct order. I can help you with that if you can't find a picture. ;)
     
    Last edited: Nov 4, 2017
  6. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    OK - thanks a lot John.

    Supplies for the large trafo so far look fine. The smaller trafo is the one I am checking at this point. There are two pairs of secondary leads coming off this trafo, the yellow leads go to pins 10 & 11 (with the white CT going to pin 9), and the blue paired leads go to pins 49 and 50 of the board 2781, with the black CT lead going to pin 51. These are all hooked up correctly and I numbered these plugs as well.

    Note that someone has been in this board before me - they had already numbered some of these leads and C1 and C2 on this board are clearly (correct) later replacements.

    I am getting the correct 52-0-52VAC on all the pairs of secondaries! Yay, that is great. As you can see, there is naff-all for these connections to go thru before the DC voltages are sent off to the relevant boards. Without any DC connections made to the board, I am getting around 100VDC between the plus and minus pins of D01 and D02.

    My understanding is that this should be at least 136VDC - correct?

    Am I right in thinking I have not one, but 2 failed RB-152 bridge rectifiers here, showing exactly the same failure symptoms? This is fairly bizarre if so......

    Oopps = perhaps I should be measuring at the smoothing caps like suggested - I will get onto that tomorrow and advise.

    These are sanken parts - rated at approx 250V and 1.5A.
     
    Last edited: Nov 5, 2017
  7. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    That^^^ seems right (+/-68VDC).......but you get only 100VDC ?.... Note: with the connectors (to the chassis mount electrolytics) pulled off the PSU/Prot board this will remove most of the smoothing, making the meter misread, and thus maybe explaining that apparently low voltage?

    Incidentally, someone did some measurements regarding RB152's and I believe found them to have a surprisingly 'dirty' output waveform. (reverse recovery). It may have only affected the gear or sample tested, however, I now regard them with some suspicion in locations where a really clean supply is needed.
     
    Last edited: Nov 5, 2017
    slimecity likes this.
  8. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    OK - I have double-checked all power supplies to the main boards and they appear to be fine.

    I have set up a testing rig and had each board connected one at a time via a wire harness to the chassis.

    The voltages I got on each module are as attached. Any thoughts on these? Green is good and red is bad. The symptoms are mainly common to both boards. My guess is that I have smoked input FETs (unusual I know) and Trs 02-09 (2SC1854s & KSA992's) seem to be the most out of spec. Could be worth replacing all of these - as I have a stockpile of these handy....

    Capture.JPG
     

    Attached Files:

    Last edited: Nov 6, 2017
  9. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    bump.....
     
  10. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    If ZD01 is OK (27V measured) then I would change TR02, TR03 - (this one might be ok) and if still no go, TR04, TR05. (might be good to check these after removal).

    Something very weird with the DC measurements you are seeing.
     
  11. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    Thanks John

    I think its odd too - plus the fact that I am seeing similar issues on both sides. I am wondering if some kind of spike or something came through from the preamp section, to equally affect both power modules. That's why Im wondering about the 2SK97 FET.

    I have a new meter and its really good - and I trust it.

    I cant see any harm in refreshing all TR02s-09 with new hFe matched parts.

    Note that the two zeners are new - I replaced them with 27 and 13v zeners respectively.
     
  12. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    Do one channel at a time, one step at a time, until the voltages come right, this will teach you what you need to do for the other channel. At the moment I don't suspect the FET(s) but acknowledge that it can look like they are faulty, however it may just be DC coupling feedback effect.
     
    slimecity likes this.
  13. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    No probs - cheers John.

    I am keen to not rush this one obviously.......
     
  14. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    Has anyone had to replace any of the NIR resistors on these power amp modules? Or are these generally OK? I note many of these are large, and I have read that this is more likely to mean you get less trouble from them.....
     
    Last edited: Nov 8, 2017
  15. Hyperion

    Hyperion Roobarb & Custard Subscriber

    Messages:
    40,190
    Location:
    Hertfordshire, UK
    I usually measure them, but can't remember ever changing one.
     
    slimecity likes this.
  16. kevzep

    kevzep Its all about the Music Subscriber

    Nah those NIR's don't generally go bad at all.
    Wouldn't worry about them.
     
    slimecity likes this.
  17. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    Bugger this thing. I've had enough.

    Cheers all - especially Hyperion.

    Replaced TR02-09 - still had issues - the same as before.

    Replaced all s2c2071 with MJE340 and all 2sa939 with MJE 350 - ensured all pinouts were correct. Then I had huge voltages everywhere including 75v on the bias pins and resistor 36 started smoking.

    I have pretty much no idea what has gone wrong here. Really gutted as I seem to be getting more issues than others had on their units. Sorry for the tanty.

    I will get another one and try again. Expensive but bugger it. I think I need to start again with functioning power modules. This one has just become a very expensive parts unit.
     
    Last edited: Nov 10, 2017
  18. Ronito6

    Ronito6 Super Member

    Messages:
    1,001
    X1's challenge the most experienced.

    WOW!!

    I would just put back in all the stuff I removed and bring it back to square one...if possible.
     
    slimecity likes this.
  19. slimecity

    slimecity Super Member

    Messages:
    1,900
    Location:
    New Zealand
    I think I need to follow your plan and have more than one unit Ronito.

    I may get back to this when Im in a better frame of mind - but I havent had the power modules working correctly from the start, thats been my main problem. So I havent had a working frame of reference to start with.
     
    Last edited: Nov 10, 2017
  20. kevzep

    kevzep Its all about the Music Subscriber

    Send it up to me, I'll get it going for you.
    MJE350/340, not good substitutes by the way, they have no published FT, and low gain, not good for this amp.

    KSA1220 and KSC2690 would be a better option.
     
    Ronito6 and Hyperion like this.

Share This Page