Need help with Mitsubishi DA-A30 repair

Discussion in 'Solid State' started by kotofei, Aug 31, 2017.

  1. kotofei

    kotofei permanent gear evolution Subscriber

    Looking for help with Mitsubishi DA-A30 amp: one channel died during idling current adjustment without apparent reason.

    I’ve got this amp in a stock form a while ago and as always decided to do some upgrades.

    The Mitsu has four pots per channel, two for center potential adjustment and two for idling current. All old and single-turn; changed three pairs of them to multiturns. Powered on, amp seems fine, adjusted center potential as in SM, all good. Then went ahead to adjustments of idling current; for this, the SM asks to rotate both idling current pots counterclockwise, connect DMM between emitters of two output transistors, power the amp and first adjust VR104 pot to 20 mV, then wait and re-adjust it to 40-50 mV, then adjust VR105 to get 80-100 mV (see attached pic with description from the SM).

    Did it for the right channel, amp seems fine. However when I attempted to do same for the left channel amp emitted cracking sound and produced some smoke. The DMM showed 0 mV and the reading didn’t change with rotation of VR104 or VR105 pots; for right channel, it was around 1.3 V in the beginning but quickly went to 0 with VR104 rotation.

    Found that the PCB is charred at the R144 and the resistor itself is severely burned although still keeps 220 Ohm. Two other resistors, R145 and R184, also look burned but slightly and also are of expected resistance. Checked fuses-nothing blown.

    What could go wrong? The VR104 pot is at the base of Q110 and VR105 is at the base of Q118; R184-R144-R145 connect base of Q113 to +58V power rail. I’m afraid to turn the amp on and check voltages since something else could go up in smoking.

    Here is the link to schematics and Service Manual:
    http://www.vintageshifi.com/repertoire-pdf/Mitsubishi.php

    Thank you.
     

    Attached Files:

    Last edited: Sep 1, 2017
  2. slimecity

    slimecity Super Member

    Messages:
    1,548
    Location:
    New Zealand
    If that resistor blew - chances are that a semi close to it did also.

    Is R114 a fusible? Are there any fusibles in this amp?

    I'd be pulling and checking Q's 111-114. Have you also checked the outputs on that side?

    Are you using the original trimpots? In my view the trimpot may have sent a spike in voltage thru the circuit - either that or a transistor involved in these operations was close to letting go and just let go. I would replace the adjuster trimpot as well - especially if it looks manky.

    I take it you have a DBT, if not, get onto this IMO......If you connected it and powered it up now, and it goes bright, chances are one/some of your outputs are gone.
     
  3. kotofei

    kotofei permanent gear evolution Subscriber

    R114A is not a fusible. Others-need to look closely.

    I quickly checked outputs to find if they are shorted. They all look shorted i.e. reading of 0 Ohms. None of the reading shows open while in the right (good) channel all readings were open or in the MegOhm range. I'll double check them later but now looks outputs are gone.

     
    Last edited: Sep 1, 2017
  4. slimecity

    slimecity Super Member

    Messages:
    1,548
    Location:
    New Zealand
    Are you happy with how you are measuring these, and your meter is OK? You had one side up and running OK. Plus its unusual fo rall outputs to go under any circumstances I would have thought......

    Dont want to sound condescending - but did you try your meter on a gud transistor first?
     
  5. kotofei

    kotofei permanent gear evolution Subscriber

    Meter seems OK, at least for resistance measurements. I checked few good transistors on the right side and their counterparts on the damaged left side. Here's what I got ( all 2xx are from good R channel):

    Q111 E-C 158 Ohm, E-B 42 KOhm, C-B 0.61Mohm
    O211 E-C 10M, E-B 580 KOhm, C-B 10 Mohm

    Q112 E-C 1 K, E-B 126 KOhm, B-C 96K
    O212 E-C 151 Ohm, E-B 580 KOhm, B-C 20 Mohm

    Q113 E-C 1.3 K, E-B 1.2 K,
    O213 E-C 10M, E-B 8M,

    Q114 E-C 5 Ohm, E-B 1 K,
    O214 E-C 8M, E-B 8.4M,

    Outputs:

    Q121 E-C 1.3 K, E-B 14 Ohm,
    O221 E-C 1M, E-B 0.8M,

    Q122 E-C 1.3 K, E-B 13 Ohm,
    O222 E-C 1M, E-B 0.8M,

    Q123 E-C 9.3 Ohm, E-B 12.6 Ohm,
    O223 E-C 1.2 K, E-B 0.8M,

    Q124 E-C 10.4 Ohm, E-B 15.2 Ohm,
    O224 E-C 1.21 K, E-B 0.82M.

    I'll pull all of them out but now Q111-114 and Q121-124 look shorted for me.
     
  6. slimecity

    slimecity Super Member

    Messages:
    1,548
    Location:
    New Zealand
    Do you have a diode test function on your meter? Should be a diode symbol. This is the best way to test transistor function
     
  7. kotofei

    kotofei permanent gear evolution Subscriber

    Soldered out Q111-Q114 and outputs, Q121-Q124 from the amp and measured them with diode function of my DMM.

    Q 111 and Q112 are good as are all of the four outputs:). The Q113 and Q114 are dead (opened). Need to look at my transistors for possible replacements.

    Measured also R144 and R184 out of circuit and they are within specs.

    What else might go wrong there? How emitter resistors survive such a burnout: I don't have a way to measure 0.47 Ohm except open/continues? Could Q115-Q120 be affected? I'll also check the zener D112 since it's located in the middle of "action".
     
  8. slimecity

    slimecity Super Member

    Messages:
    1,548
    Location:
    New Zealand
    OK, this sounds more like it.

    Replace the trannies that are obviously dead with good replacements. i would also replace the bias trimpots in this case, as the issue could be coming form there.....

    Set up your DBT and power the amp up.

    If it glows bright - you still have a problem.
     
  9. kotofei

    kotofei permanent gear evolution Subscriber

    Found two replacements for Q113 and Q114: MJE15030 and MJE15031. Soldered them in as well as outputs and resistors. In the process lifted one leg of D107 and D108, checked- both good. Measured resistance in-circuit of R160, R143, R142, R161, R141- same as right channel counterparts. Changed old 100K trimpots VR105 and VR205 to new multiturn 100K.

    Set up DBT and powered the amp up- glows bright. Removed fuse F11, effectively powering of left channel: DBT glows bright then dim. That means DMM works as supposed on this amp. Swapping F12 for F11 i.e. leaving right channel only resulted in DBT staying bright. Not good.

    Rechecked Q113 and Q114 lifting one leg of R143 and R145- both looks fine with diode DMM. Interestingly, resistance between E and C of Q114 is 4 Ohm ( red lead is on E). If I unsolder wire 7 from the PCB- this wire goes from E of Q114 to R167 and base of the output Q123- the E-C resistance of Q114 increases to 115 KOhm. Can't really explain this.
     
  10. elnaldo

    elnaldo Addicted Member

    Messages:
    6,500
    Location:
    Buenos Aires GMT-3
    Hello. I can't follow all the resistance readings, but you need to find the short from scratch again. Of course 4 ohm resistance between emitter and collector means a problem there.

    What I'd do: to remove the output and driver transistors, and check with the DBT. Compare the DBT brightness between channels (removing fuses), or use an Ammeter. If everything looks fine, no smoke, no high current in the bad channel, install the drivers and check the DBT again with drivers installed and no outputs. Check voltages at the output transistors connection points. it must be near zero at the base and emitters, and +V and -V at the collectors. If you have high voltage at the bases, something is still wrong.

    Small signal transistors are possible fails, and so the Zener and diodes in that part of the circuit, I'd replace them if in doubt. Some transistors check OK to the low voltage of the DMM, but fail under the high voltage in circuit. Check Q109 and 110.

    I wouldn't discard a problem at the "over power detect" circuit, Q125.

    Also, measuring the voltage drop across emitter resistors at each of those transistors, and comparing to the good channel, can point you to a bad one, conducting more, shorted, or not conducting. A hot resistor means a large current flow across that resistor, from +V or -V to ground, so follow the path from +V or -V to ground through those resistors.
     
    Last edited: Sep 5, 2017
    slimecity likes this.
  11. kotofei

    kotofei permanent gear evolution Subscriber

    Removed outputs again, all 4 of them. Then decided to plug it in with drivers installed through the DBT reasoning if bulb will glow bright turn off and remove drivers then.

    Accidentally, plugged the amp not in DBT sockets but in main AC i.e. before DBT. Powered on- smoke, sparks... #$%&!

    Both drivers are gone in a violent way, the emitter of Q113 was just blowed off the transistor's body! Never saw anything like this. Q117 is bad, others in Q115-Q120 group looks OK by diode test in circuit. Soldered out Q116, Q119 and they tested good. The resistor R156 is burned up and is partially open, 33 ohm instead of 150.

    I'll look if I have R156 and Q117 subs; however, is it safe to test the amp using the DBT, even without drivers?
     
  12. elnaldo

    elnaldo Addicted Member

    Messages:
    6,500
    Location:
    Buenos Aires GMT-3
    you can test the amp without drivers. Don't expect sound at the output, or bias circuits working, but you can check for shorts.

    But I suspect you have some physical or mechanical short somewhere, do a visual inspection for bare wires touching the chassis, some screw pressing one wire, some mica insulation missing, some insulator (washer, whatever) missing, some solder bridge, some component lead touching something it shouldn't.

    What you was doing (passive tests searching low resistance) is a good practice, plug the unit to the DBT ONLY when passive tests doesn't show a short. If you read low resistance between +V or -V and ground, it won't fix by itself when you power the unit, make sure the passive tests are OK before powering ON again .

    Do a mental list of things you trust 100% and things you don't trust.

    You reported a short that disappear when one board was disconnected, narrow your search there.
     
    Last edited: Sep 7, 2017
  13. Goldie99

    Goldie99 Active Member

    Messages:
    490
    Location:
    The original Boston, UK
    It's not going to help at the moment, but you should also review the procedure you used for bias adjustment - if what you typed above is what you actually did, that may have caused, or at least contributed to, the original problem:
    The manual states to rotate both idle current pots (VR104, VR105) fully clockwise, then adjust VR105 to 20mV, then 40-50mV, and finally to adjust VR104 to the final setting.

    If any of the pot substitutions you made were in this area, you should also check that they rotate in the same sense as the originals, e.g., clockwise to increase or decrease, whichever the case is, but the same as the originals, otherwise, the whole procedure changes.
     
  14. elnaldo

    elnaldo Addicted Member

    Messages:
    6,500
    Location:
    Buenos Aires GMT-3
    That's why I NEVER follow those kind of instructions on the service manuals, I never rotate a trimpot fully to one side, I always set them in the middle, and power the amp on the DBT, reading the voltage for a brief second. If OK, I power it at full power , also for one or 2 seconds just to check if bias is within a reasonable value, and move the trimpot from there.
     
  15. Binkman

    Binkman Addicted Member

    Messages:
    5,964
    Location:
    Raleigh, NC
    Perhaps a little help w/transistors. Avionics page with pics. Make sure batts. are good as constant use the dmm can drift some.

    http://audiokarma.org/forums/index.php?threads/testing-transistors.201303/#post-2362176

    ---aside;
    Vpots.. can be erroneous to clockwise or counter-clock wise to zero. elnaldo is correct that why move the pot if the previous setting is assumed accurate. then slightly adjust to see +/- mv change. Or if those cheapie plastic pots? junk. get bournes and save the aggravation.

    Made same mistake on latter model sony assuming same as older sony's ccw to zero? was cwise to zero. magic smoke.. took out the drivers and half the outputs.
     
  16. kotofei

    kotofei permanent gear evolution Subscriber

    Checked resistance from +V and -V to ground, it's not zero and approximately same in both channels. Powered up with DBT-bulb is dim and remains dim.

    Soldered in small transistors ( Q116, Q119) and re-tested with DBT-same as before. Found then subs for Q117 and R156 and soldered them in. Retested with DBP- same as before.

    Since I don't have extra drivers just took them from other (good) channel and soldered in the bad one. Tested again with DBT. Bulb is dim at power up and remains dim, doesn't change brightness by eye. Powering with DBT, measured current with ammeter-inserting it instead of a fuse. Current is 0.266 AC amperes.

    Now I need to power the amp on with DBT and check voltages at the output transistors connection points. Right?
     
  17. Binkman

    Binkman Addicted Member

    Messages:
    5,964
    Location:
    Raleigh, NC
    Noted above... Bias pots!! you have to be sure these are reliable CC or CCW make sure of that.

    I'd check the drivers for proper input v's.before the outputs cold. pull if you haven't.

    Don't need the dim bulb if okay "dim". just check the emitter and or "0xx) or whatever the sm says for proper mvs. You won't get any adjustment on bias as noted above so driver transistors are important.
     
  18. kotofei

    kotofei permanent gear evolution Subscriber

    With the amp plugged thru DBT, tested voltages at output transistors connection points. The +V is 34.3V and the -V is -34.3V ; SM calls for +/- %* V but this is without BDT. Before measurements, moved pots into central position.


    For Q121 and Q122, E is +3.05 V, C is +34.1 V and B is +13.5 V.

    For Q123 and Q124, E is 3.09 V, C is -34.1 V and B is +13.5 V.

    All measurements against chassis ground.
     
  19. elnaldo

    elnaldo Addicted Member

    Messages:
    6,500
    Location:
    Buenos Aires GMT-3
    Emitter voltages look OK, but bases at +13.5V look too high to me (I know you don't have current flow since the output are not installed, but anyway seems high to me). And the fact both are positive, doesn't look right either, they seem shorted (tied togheter).

    Perhaps the DBT is making some transistor to stay off (you have 34V, and the schematic calls for 58V). Try with a larger bulb, just a brief test, to see the +V getting closer to 50V, and check the bases again.

    Is R146 OK? and the Zener D112 OK ? You should have a voltage across it, if I'm right, 6V (MZ306 looks like a 6V zener).

    What are the voltages at Q113 and Q114? If both are conducting, you should have a voltage drop at R146. They are biased by Q111 and Q112, so if they don't conduct, check back at Q111 and Q112.

    Look at the schematic. Output bases are connected to pins 4 and 7 of the board where the drivers are. Those points are not electrically connected, so it's not normal to have the same voltage at both points. I'd expect a difference.
     
  20. kotofei

    kotofei permanent gear evolution Subscriber

    Using the same bulb (it's 43 Watts; not sure I have another especially around 100W) found no voltage across 146 (0 mV) and 13.36V to ground at each its end.
    R146 looks fine by eye. D112 also looks fine.

    For D112, voltage across it is also 0 volts and same 13.36V to ground at each its end.

    At Q113, E is 13.4 V, C is +34.6 V and B is 14V
    At Q114, E is 13.4 V, C is -34.5 V and B is -34.3 V

    Q 111 is E 17.8 V, C is 34.5V and B is 18.5 V
    Q112 is -19.8V, C is -34.5V and B is -9.3 V.

    I will try to find high-wattage bulb and measure something with it.
     

Share This Page