Announcement

Collapse
No announcement yet.

Another GK 800RB hard failure

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Another GK 800RB hard failure

    Those that service these aging but usually reliable bass amps no doubt have encountered the hard failure they produce when over-stressed. Two of these just came in from a steady client, both being rental amps. In both cases, the output xstrs of the 100W Amp (normally used for the HF output cabinet along with the 300W Amp driving the Bass Bin) had shorted power xstrs. One was a melt-down. PCB scorched where resistors sat and burned a hole thru the board, with many other resistors burning open. Soot all over the insides, along with little fragments that got airborne inside during the meltdown.

    I've no details in how the failure occurred on the two amps. Only one burned, to the extent of having to first remove the burnt parts, then had to scrape away the carbonized fiberglass enough to allow the remains of the PCB to be coated with Corona Super Dope to seal it and prevent the damaged board from being a high-resistance path between circuit branches.

    Click image for larger version  Name:	RB800-B-Failure-4.jpg Views:	0 Size:	1.59 MB ID:	969332 Click image for larger version  Name:	RB800-B-Failure-5.jpg Views:	0 Size:	1.45 MB ID:	969334 Click image for larger version  Name:	RB800-B-Failure-8.jpg Views:	0 Size:	1.34 MB ID:	969336

    Click image for larger version  Name:	RB800-B-Failure-10.jpg Views:	0 Size:	1.79 MB ID:	969338

    I cleaned up the soot and debris scattered throughout the chassis and the boards. All the damage appeared to be on this 100W amp circuit. But, in checking out the rest of the amp, I found the readings of the power xstrs on the 300W section reading bad, though none were a hard short as was the case on the one pair in the 100W circuit. I de-soldered all of the power xstrs in the 300W amp and verified I had full isolation of the base and emitter terminals passing thru the large solder pad. That done, I checked all eight (drivers and the three output pairs), those now measured nominal using the DMM in Diode Test mode. Checked the short circuit protection devices, bias, emitter resistors, and all was ok over there. So, set about the repair of the PCB first.

    Click image for larger version  Name:	RB800-B-Failure-Cleanup-6.jpg Views:	0 Size:	1.46 MB ID:	969340 Click image for larger version  Name:	RB800-B-Failure-Cleanup-3.jpg Views:	0 Size:	1.53 MB ID:	969342

    After removing all of the failed parts. I had to scrape away all of the carbonized PCB from both sides...mostly on the component side, then coated it with Super Corona Dope and left to dry overnight. I thought I had taken photos of that prior to restoring the circuit components, tack-soldering leads where pads had burned off, and such. No photos of that step.

    Click image for larger version  Name:	RB800-B-Failure-PCB Repair-1.jpg Views:	0 Size:	1.46 MB ID:	969344 Click image for larger version  Name:	RB800-B-Failure-PCB Repair-3.jpg Views:	0 Size:	1.91 MB ID:	969346 Click image for larger version  Name:	RB800-B-Failure-PCB Repair-2.jpg Views:	0 Size:	1.67 MB ID:	969348

    Having also pulled the second amp apart, which didn't have this severe damage to the PCB, I just found after putting this first one back together yesterday evening, so it would be ready to connect probes thruout the two amp sections, composing this thread, I saw the same original emitter resistors on this burnt unit in place, and see it was the other board which had both of these emitter resistors open and were the ones I had to chiesl off of the heat sink. I just checked and see one of the two resistors is open on this amp, so I have to take it apart again and replace the pair of 0.33 ohm/5W resistors that should have been done when it was all apart. Sigh.

    At least I found it before applying power.

    Onward



    Attached Files
    Logic is an organized way of going wrong with confidence

  • #2
    Remember lot's of charring and soot when those went bad.
    An interesting cause of failure I've mentioned before: removing the rubber feet to save space in a rack. The now 'longer' screw right under the speaker jacks shorts out the spkr hot to chassis. They may have re-positioned those jacks in later models but in the early ones that's what would happen.
    Kind of silly to put the screw back in without checking if it's going to hit anything after the foot was removed.
    Originally posted by Enzo
    I have a sign in my shop that says, "Never think up reasons not to check something."


    Comment


    • #3
      I got that 100W Amp's emitter resistors swapped out, it too having them glued to the heat sink, so replacing those, as careful as I tried to be, still broke apart so had to clean up the debris from the cement resistor bodies. Got the amp put back together. Cabled it up with test probes to monitor the voltages across the emitter resistors, along with the outputs for DC offset and power supplies to check their level. This particular amp doesn't have the +/- 85VDC supplies for the 300W amp. Only has +/- 70VDC. 100W amp has +/- 60VDC. After I got the bias set for the nominal 5mV across the emitter resistors, I cycled power on and off several times just to see everything came back as it should. DC offset less than 15mV in either polarity. Idle current/power was 320mA/20.8W @ 120VAC.

      Both amps pass signal.....had it fed with burst pink noise...8 ohm speaker for the 100W amp, 4 ohm speaker for the 300W amp. Crossover works fine, all the controls work fine. I haven't yet plugged my bass into it, but I expect that will be ok, from what I'm seeing/hearing.

      The second GK800RB didn't have this level of damage on the 100W amp. Both of it's emitter resistors were open, and the MJ15022/15023's installed in that were shorted. The 300W amp...I found both MJ21193G, MJ21194G's together with MJ15022G & MJ15023G TO-3 xstrs, though these two were used for Drivers, and one more MJ15023 was used in the Voltage Gain stage (Q12). One each of the 21193/21194's failed shorted in the output stage (all MJ21193/MJ21194. All of it's emitter resistors on that 300W amp remained intact.

      I went to order replacement parts via DigiKey, and got hit with sticker shock. MN21193's are $8.83 (10pc), while MJ21194's are $9.93 (10pc). MJ15022'G's are $9.23 (10pc) and MJ15023G's are $9.83 (10pc). Seems like I was paying more like $3.50-$4.00 for these not too long ago...will have to check my invoices. I had a pair of MJ15022/15023's left, so that cleaned me out of on-hand parts, now waiting for the MJ21193/MJ21194's.
      Logic is an organized way of going wrong with confidence

      Comment


      • #4
        Originally posted by g1 View Post
        Remember lot's of charring and soot when those went bad.
        An interesting cause of failure I've mentioned before: removing the rubber feet to save space in a rack. The now 'longer' screw right under the speaker jacks shorts out the spkr hot to chassis. They may have re-positioned those jacks in later models but in the early ones that's what would happen.
        Kind of silly to put the screw back in without checking if it's going to hit anything after the foot was removed.
        One of these amps DOES have the rubber feet on the bottom. The other one doesn't, but I had forgotten about that stupid engineering mistake at GK. On these two chassis I have here, that being the last Flat-Head-formated cover screw...is aligned with hitting the GND terminal of the 300W Output Jacks, where they soldered the two GND terminals together. Good to know! Thanks!
        Logic is an organized way of going wrong with confidence

        Comment

        Working...
        X