KRK V6 Series 2 Speaker repair

[Luciano2572]

Hello everyone!

Here I am again with another one, which has priority over my HP 6274b power supply 

This is a 6" active monitor speaker, this week out of nothing it made a loud noise and suddenly let out the magic smoke, a LOT of it. Opened it up and oh god, the smell. The tweeter measured open right off the bat, so that's dead. I didn't test the woofer yet but it measured 1.9ohms, however shorting the leads also gives 1.9ohms... Didn't get to measure impedance or test it with another amp.

So I took the amplifier board out of the cabinet (forgot to take pictures of the whole speaker  but I attached one from the internet), no obvious signs on it of anything bad other than some goo to fix capacitors and parts in place. They used hot glue a lot, but I question the yellowish circles around C10, C9...  theres a picture with a zoom on them. Are those some adhesive or leakage? I have an el cheapo transistor tester from ebay which I guess can help me test capacitors along with fluke 179 capacitance meter.

So I went to power up the board. The supply rails for output ICs seem fine at around +/- 29,5V each.
Feeding a 0.3Vpp sinwave and checked the output with the osciloscope (blue = woofer output, yellow = tweeter output, purple = signal gen)
I swear I had both tweeter and woofer having a sinwave at one point, and their amplitude changed as I changed the frequency, tweeter going higher as amplitude got higher, inverse for woofer. But that didn't last long. All of a sudden both outputs were at 14VDC.
Powered off the unit and after powering it on again, woofer had a sinewave and tweeter not. I hooked up an old woofer to it and it sounded pretty nice, so I guess that rules off amplifier IC.

Then measured the two linear regulators on the board for +15v/-15V. +15V one was hot to the touch, while the -15V was warm.

Here comes the interesting part:
+15V reg (HOT): Vin = 25.5V, Vout = 15.08V
-15V reg: Vin = -26,2V, Vout = -13,74V and dropping a bit more... until it went to -0.67V. At this point, the woofer output which was doing fine went to a fixed DC level again!

So I'm guessing theres something wrong on one the opamps (theres 5 TL072CN ) which is dragging the rail voltage down or... the capacitors?

Can anyone shed some light? I tried looking for a schematic to no success.
As a sidenote, we have 5 of these in the same studio, they were stored for 5-10 years in the box and put into usage at the start of this year, so they are sorta new old stock 
More & Higher res pictures: https://drive.google.com/open?id=1-53U3BguL3Rvb073eWsImBWB0wsSNNfE

Output ICs are LM3886TF, one for each


Any input is appreciated!

Yours,
Luciano

[Luciano2572]

More pictures due to forum limitation:

[coromonadalix]

Dang they put  gunk glue every where,  well the fun part is :

You have standard power supply,  an xformer some fixed regulators, known number parts ...  i think you have an switched supply or switched speakers with the medium sized black relay, and maybe an muting or another supply rail  with the black smallest relay.

I would check c7 c8 c9 c10      your output  ic's  dont forget to put back the screws to help them dissipate on the heatsink,  they may have thermal shutdown.

http://www.ti.com/lit/ds/symlink/lm3886.pdf

You want to do some tests  you have to put an load at the output IE: some 8 ohms fixed resistors, you never test an amplifier without an fixed load, and connect an scope in parallel of the loads  ...

If all the supplies are good   the +15 and -15 volts and the two mains voltages, you can try to inject some small audio signal to the lm3886  (the pdf link will help) and see if something will get out, ( dont forget the relay(s)

The -15v  dropping   maybe due to capacitors,  not sure for the op amps,  you can remove them, put ic sockets and check if the -15v  will stay solid,  lucky you thoses regulators are normally short proof.

those op amps  cost nothing   dont know the 14 or 16 pins ic number

And finally  if you can,   remove all the hot glue, the gunk  ... when cold hot glue should'nt be too hard to remove it

[Luciano2572]

Thanks for the guidelines, @coromonadalix !

I brought my thermal camera today and took a picture of the board -  KA7815 was around 70-80C but providing a solid +15V.  opamps were cool as was the KA7915.But it's voltage kept dropping. As the input voltage was fine, I went ahead and replaced it with a LM7915. Steady -14.90V output, didn't drop over time and I guess problem solved.. I stuck a heatsink on the KA7815 to lower it's temp a little since it's gonna be enclosed in the cabinet. I saw a picture of it's bigger brother, the V8 speaker, and it had heatsinking on both regulators. I left the KA7915 without since it's running cool.

Attached thermal picture before KA7915 replacement and normal picture after KA7915 replacement + KA7815 heatsink bodged

When I removed the 7915, I powered it up just to see what it would do. >14V dc on the output with huge >20V peaks. Even with a sinewave on the input, it changed nothing. As soon as I soldered the 7915 replacement, it was fine.

Stress testing the amp with an old speaker+tweeter I had here, since both were blown (looking at 120$ for the replacement tweeter+woofer combo  ) and everything seems fine 

Yours,
Luciano

[coromonadalix]

Good idea for the heatsink, dont forget to bolt the outputs ic's on the heatink  i did not see the screws on the photo.

I would say check for the ac component on the two main capacitors if the ac ripple is too high or not ? 

Glad you got it repaired, and lucky you with your thermal camera  , i do use an Flir at my job, very practical. But i'm afraid to buy one for myself  loll

[Luciano2572]

Good idea for the heatsink, dont forget to bolt the outputs ic's on the heatink  i did not see the screws on the photo.

I would say check for the ac component on the two main capacitors if the ac ripple is too high or not ? 

Glad you got it repaired, and lucky you with your thermal camera  , i do use an Flir at my job, very practical. But i'm afraid to buy one for myself  loll

While bolting the unit back to the back panel I bolted the output IC's as well, almost forgot about them   

I took C7 - C10 out and tested them on my cheapo LCR meter (since they are on the input/output of the 7815/7915 regulators), got the following results:

• 1000uF 35V#1: 968,2uF - 0,09 ESR - 0,5% Vloss
• 1000uF 35V#2: 954,7uF - 0,09 ESR - 0,8% Vloss
• 470uF 25V#1: 465,8uF - 0,13 ESR - 1,4% Vloss
• 470uF 25V#2: 467,4uF - 0,12 ESR - 1,4% Vloss

Checked a ESR table and the values seem pretty good so I put those back in.

+39,2V rail ripple (measured with HP 34401A in ACV): jumpy around 22mv-31mV but most of the time in the 22-25mV area
-39.2V rail ripple: 22mV-26mV

I'm tempted to open a good V6 and see if the 7815 is hot as this one, the 15V line gets routed to a small board with the limiter components but still the difference in temperature from the 7815 and 7915 is huge

As for the thermal camera, it's the """"cheap"""" flir one for phones - not like the pro ones but hey for the cost I can't say anything bad about it   

Again, thanks for the support on this one Coromona!

[Luciano2572]

Okay, think I found the reason for the 7815 to be "hot"
The main relay is powered from it, it's a FTR-F1AA012V. It's rated Power Consumption is 530 mW.
At 25.5V input, the regulator is dropping 10.5V in heat.
P = UxI -> 0,53 = 12xI -> Irelay = 0.04416A
Pdissipated by KA7815 = 10.5x0.4416 = 0.46375
From KA7815 datasheet: Thermal Resistance Junction-Air (TO-220) RθJA 65 °C/W
0.46375*65 = 30C!

So if my math is right, just to power the relay alone that TO-220 is going to rise 30C over ambient, not counting other IC's on the limiter board. So this explains the 70C reading it was giving me the other day. To confirm this, I left the unit powered on with the switch on the back (which turns on the speaker, has 3 positions: ON, OFF or AUTO which powers on as soon as theres a signal) on the OFF position, keeping the relay off. The KA7815 was a bit warmer than the 7915 but not by much sooo that's done! 

I went to a local construction supply shop and bought some high temperature silicone, put some on the heatsink to help support it to the board since there's no fixing tabs for it. Pics attached, not the prettiest application but it should work 

[coromonadalix]