Kaiju Right Channel Out

[bombay317]

My right Kaiju channel started getting quiet, and now it wont even make noise.  I flipped it over and D1 and D2 on the B side are not lighting up on the right channel.

Voltage Checks Right:
IA=517.5V
OA=357V <--looks bad
IB=364.4V <--looks bad
OB=-0.039V <--looks bad

Voltage Checks Left:
IA=516.0V
OA=297.0V
IB=297V
OB=162.5V

Are Q1 and 2 out? i was hooked into the ground of input AC.  Should i try reflowing the C4S board?

 

[bombay317]

I did some reflowing of the board and havent gotten the right channel to come up, it was a cursory effort and I will try more in depth work tomorrow.

 

[Paul Birkeland]

OA=357V <--looks bad
IB=364.4V <--looks bad

These are wired together and can't be different.

OB=-0.039V <--looks bad

Yes, this is bad.

What's the Kreg voltage on the B side?

 

[bombay317]

Right Channel
KregA: 30.1mV
KregB: 14.9V

Left Channel
KregA: 3.5V
KregB: 7.2V

 

[Paul Birkeland]

Swap the 5670 tubes to be sure the issue doesn't follow a tube.

I would suspect an issue with the string of black zener diodes that snake around on the terminal strip under the PC board.  Sometimes long uncut leads can create shorts. 

 

[bombay317]

I did do the swap and it did not follow the tubes when it was at my house.  I can do it again in the lab to make sure.

I will inspect the zeners.

 

[bombay317]

I inspected the zeners and they appear to all be clear.  I also checked them with the voltmeter and they are all good.

I also reflowed everything and am still getting the weird performance.  I am thinking that I should check the other components on the board and potentially replace some of the fets and stuff.

 

[Paul Birkeland]

Did you reflow the joints on the PC board? 

How are the OA and OB voltages on the offending board?

 

[bombay317]

OA: ~364V
OB: -0.087V

reference connected to -reg. some other voltages:
Q1C (B): 350V
Q1B (B): 364V
Q2B (B): 364V
breg (B): 364V
Kreg (B): 15V
Kreg (A): 30mV

Of note, D1 (A) and D2 (A) are lit, D1 (B) and D2 (B) are not lit, and they shouldnt be based upon the voltages.  they are not conducting for some reason.  I guess also Rdb and or Cdelay could be shorted, i can measure them too.  It does seem like Q1/2(B) are not conducting.

 

[Paul Birkeland]

We have some odd circumstances here.  KregB has 15V, which is plenty to juice up the HV regulator and get your OA voltage down where it's supposed to be, but that's not happening.

If we look at the 5670 plate curves, 15V on the cathode and 364V on the plate means that no current is flowing.  My suggestion would be to remove the wire going into IB, then recheck the voltages.


I strongly suspect a soldering issue on that PC board is causing the issue, with some emphasis on the center legs of each MJE5731A.  The other option is that there's a shielded twisted pair of wires that connect to pins 1 and 9 on each 9 pin socket.  If that has come loose on the offending side, you'd end up with voltages like these.  You'd also end up with a 5670 that doesn't glow visibly.

 

[bombay317]

I will take a look at these suggestions in the next couple of days.

The way that this started was gradually decreasing output on that channel.  It was working fine, then i noticed that i would have to keep increasing the output of that channel.

 

[bombay317]

I haven't had the time to look at this so I am sending it in for the repair service!

 

[bombay317]

Interesting update.  I did send this for repair and got it back and everything is working (the repair service is pretty great if you get strapped for time).  However, i just did the BeeQuiet upgrade, and then the right channel appeared to be out again.

Some thoughts / things I did:
1. The BeeQuiet upgrade removes the balance pot, so there is a good chance that I had been attenuating the right channel a lot on the original BeePre I had set up.
2. After doing the upgrade, I switched the output L/R channels, and the "quiet" side stayed on the right channel, indicating that this issue is with the Kaiju not with the BeePre with BeeQuiet upgrade
3. I rotated the 100k balance pot on the Kaiju one full rotation, and then everything appears to be back to normal balance.
4. if the levels drift again, such as the volume continuing to drop on the right channel, something would be off and need more troubleshooting.

 

[Paul Birkeland]

2. After doing the upgrade, I switched the output L/R channels, and the "quiet" side stayed on the right channel, indicating that this issue is with the Kaiju not with the BeePre with BeeQuiet upgrade

Check your cables.  This could be a bad cable.

3. I rotated the 100k balance pot on the Kaiju one full rotation, and then everything appears to be back to normal balance.

These are not balance pots, they are trim pots for minor level differences.

What happens if you change the interconnect cables at the Kaiju?

 

[bombay317]

I changed the RCA cables going into the input of the Kaiju, and it still gives what would seem to be normal balance and power sent to both speakers.

Could there have been an error in my assembly of the BeePre in the first place that was fixed with the changing over to the BeeQuiet?

 

[Paul Birkeland]

The easiest way to check this is to play a 60Hz tone through the BeePre with the level controls all the way up.  Measure the AC voltage from the center pin of each RCA jack to ground and let us know what you get there.

 

[bombay317]

I will gather my tools and check this.

 

[bombay317]

With a 60Hz tone playing into the input of the BeePre, i read out 0.98VAC Left channel output and 0.99VAC right channel output with the attenuation all the way at 0dB.

 

[Paul Birkeland]

Alright, that's about perfect.  Try measuring the AC voltage at the end of each interconnect connected to the output of the BeePre (use both the center pin and shell of the RCA plug for this measurement).

-PB

 

[bombay317]

I did that for both the L/R channels.  I measured at the end of the RCA cables where they would be plugging into my Kaiju.  They both read the same voltage of 0.98VAC.  I also added the Kaiju into the mix, but this time with some attenuation.  With the output of the BeePres reading the same voltage, the output of both Kaiju outputs were the same.  I even rotated the Kaiju knobs to make sure I was starting from the same point, and they ended up roughly at the same location on the dial.

Is it possible that I had a bad solder join in the BeePre balance dial that was adding in more resistance and attenuation on the right channel?  I think that is the best explanation I can come up with at this point.

Everything seems to be working super fine, and the sound is amazing.