Noisy channel [resolved]

[JulienG]

I finally finished building my Eros last week, and all seemed fine through the early tests, but in use the right channel has a high amount of background noise.

I switched the first stage tubes just to be sure it wasn't that, but the problem remained with the right channel.

If I put a scope on OB I see the noise, so it's probably on the first stage.

The other odd thing is that D1-B isn't lit (I've tested the LED and it does work), but D2-B is (fairly certain this wasn't the case in early tests, but is now).

[Paul Birkeland]

Can you post the voltages on that channel?  Specifically:

IA/IB
OA
OB
Kreg
Breg

Almost all the time when you have a noise like this, it's a loose wire or a flaky solder joint.

[JulienG]

Bad channel:

IA/IB 221.0v
OA 141.3v
OB 132.0v
Kreg 6.9mv
Breg 132.2v

Good channel:

IA/IB 222.2v
OA 155.0
OB 103.4v
Kreg 1.6v
Breg 103.9v

Line voltage here is about 245v RMS (low for urban Australia)

I'd already pulled the board and touched up any joints that didn't look perfect.

[Paul Birkeland]

Well, you have a channel with bad voltages.  This preamp is not yet usable.

Your Kreg voltage on the bad channel is 0V (6.9mV is 0.0069V).  I would focus on that PC board first.  Is the 2N2222 transistor properly installed? Are all the jumpers present and where they should be?

You have one working channel to compare against, and that will be very helpful in diagnosing the non-working channel. 

[JulienG]

Reflowed the joints around that area, which succeeded in returning the voltages to those expected, but noise still remains, and the LED is still out.

Bad channel:

IA/IB 221.0v
OA 170.4v
OB 98.9v
Kreg 1.83v
Breg 100.3v

Good channel:

IA/IB 222.1v
OA 158.7v
OB 101.4v
Kreg 1.56v
Breg 102.4v

[Paul Birkeland]

Can you post some photos of your build?

[JulienG]

Sure, will grab some photos.

I just swapped the boards, and noise moved with the board, the 2N2222 appears to be installed correctly, and Jumpers are identical on both boards.

[JulienG]

Photos attached, in case the names don't come through, the channel with the electros on top & left is good, the one with electros on bottom & right is bad.

[Paul Birkeland]

I would look under the board as well for solder joints that aren't well flowed. 

[JulienG]

I'd already done that when I first removed the board.

I just went through and made sure all joints look good on top too, adding solder in cases where it hadn't flowed through from the bottom when I first soldered it. No change.

Went through with the scope again, noise is present on OA, OB & Breg, but not kreg or any components further off (and I think I probed everything on the board a hop away from those points).

[Paul Birkeland]

Is there noise on IA or IB?

[JulienG]

Nope, just went and double checked, IA/B on both boards is clean.

[JulienG]

My guess is, absent some sort of soldering issue or board error I've somehow missed that it has to be one of Q2 (the MJE350), the 100uF cap, or outside chance the 2N2222.

I'll pop over to the local electronics shop sometime this week and pick up a replacement set of caps for testing (if I find they're the problem I can always order some quality replacements from Digikey next time I drop in an order), and I've ordered replacement sets of transistors for testing.

[Paul Birkeland]

it has to be one of Q2 (the MJE350), the 100uF cap, or outside chance the 2N2222.

If you disconnect the 100uF cap, gain goes way down in the circuit and you'll actually lose a lot of low frequency response.  A bad MJE350 or 2N2222 will cause DC voltage issues. 

[JulienG]

I tried replacing the caps, transistors, and even one of the resistors, but ultimately, replacing that unlit LED did it.

Out of circuit the LED still shows the same test voltage as good ones, but doesn't light.

Now to give it a proper test.