Rega Elicit MK2

I redid all the welds on the left side and the breakdown is still the same...
I don't know what else to do to identify the culprit.
 

Attachments

  • 20230306_004840.jpg
    20230306_004840.jpg
    467.4 KB · Views: 80
  • 20230306_005009.jpg
    20230306_005009.jpg
    510 KB · Views: 78
May I suggest, if you're tracing an AC signal and it "disappears", a DC bias fault may have the amp stuck on a supply rail. You may have better luck tracing DC voltages. Compare with the working channel for guidance.
 
Hello,

@rayma , I tried to press on the different zones but that does not change anything ...

Do you have an idea of the characteristics of the signal that I can inject on the test pin of the channel after having switched the jumpers to test ?

@BSST , Thank you for your help. I don't quite understand how I could do this.
 
Last edited:
Hello,

@rayma , I tried to press on the different zones but that does not change anything ...

Do you have an idea of the characteristics of the signal that I can inject on the test pin of the channel after having switched the jumpers to test ?

@BSST , Thank you for your help. I don't quite understand how I could do this.
I believe the preamp output bias should be nearly 0VDC. What DCV do you observe at the junction of R85, R85? Assuming it's dramatically not 0V, also measure some key points within the amp stage. I suggest the gates of U404-A, and U404-B, and U4A pin 1. I also suggest measuring the same points in the other channel (which I understanding is working?) for insight to what's expected. Depending what you report, we'll poke around at additional circuit nodes.
 
Good morning,
at the R84 / R85 junction I observe 6.9mV which is close to the R86 / R87 junction which is at 8.6mV

At U4A pin 1 i see 0.7V DC, i use the side of C33 to read the value.

At the junction of R4 and R57 i read 6.7 mv DC and on the other side (the good) at the junction of R7 ans R60 i have 3.3mV DC

The base of JF1B on the bad side (read at the junction of C47) is about 6.6mV and the good side is at 20mV (junction of C49) !

On the blue points i read the same value 13.28V DC
 

Attachments

  • 16786519073697184320705774004161.jpg
    16786519073697184320705774004161.jpg
    523.5 KB · Views: 81
  • Screenshot_20230312-212909_Samsung Notes.jpg
    Screenshot_20230312-212909_Samsung Notes.jpg
    284.6 KB · Views: 84
Last edited:
The those voltages look very reasonable.

Help me refresh my memory--- you were having trouble getting a line input signal to pass though to the preamp output? Did I surmise that it was unclear if the WM8816 or the associated discrete amplifier was at fault?

My observation is that a couple of post jumpers can be moved to allow the amp to be tested independently of the volume control IC. Specifically: at JP2, move the jumper from 1-2 to 2-3 and at JP3, move the jumper from 1-2 to 2-3. This configures the amp for a gain of -10 (i.e. -R25/R34). You can inject test signals at "Left Test I/P" to confirm amplifier operation. This configuration also helps debug if there is a problem. Given your reported measurements, you'll probably confirm the amp is OK. If so, problem is probably U6. You can probe digital signals at U6 pins 6,9,10 to confirm control signals reach the IC. My understanding is the IC is discontinued. Good luck!
 
The part after the XWM8816G works!
I observe what I heard before it broke down, the left channel now down is slightly weaker than the right.

Input 220mV p-p
Output R 2.2V
Output L 2.04V

Perhaps a PR1 setting will correct this?