Aleph J illustrated build guide

Jump to Latest
#8,701
Do you have mains earth gnd connected to a single point on the chassis and psu gnd connected to the same point through an NTC or ground loop breaker?
I’m not sure, but it looks as thought the AC input module Earth Gnd tab connects to a fuse socket?
 
Last edited:
#8,702
Vunce said:
Do you have mains earth gnd connected to a single point on the chassis and psu gnd connected to the same point through an NTC or ground loop breaker?
I’m not sure, but it looks as thought the AC input module Earth Gnd tab connects to a fuse socket?
Click to expand...
Yes, it is connected that way. The AC input module I used did not have a fuse for ground as per the schematic, so I added a fuse holder to provide the required fuse. The wire then goes to main chassis ground, where the power supply ground is also connected through an NTC.
 
Last edited:
#8,703
tizman said:
Yes, it is connected that way. The AC input module I used did not have a fuse for ground as per the schematic, so I added a fuse holder to provide the required fuse. The wire then goes to main chassis ground, where the power supply ground is also connected through an NTC.
Click to expand...
Three things:

1: Hassle, but if you twist your DC wires from PSU to amp PCB you can achieve an even lower noise floor.

2: You have connected audio gnd wires between amp boards and PSU at opposite sides of the junction on the PSU. Moving them to the same side of the PCB, and as far away from the junction (connection between the two psu board halves) as possible can achieve a quieter ground.

3: Have you tried rotating the transformer? The noise might be caused partly by magfields from the transformer, and some areas may radiate more than others.

4: I might not remember correctly, but the difference in noise with and without inputs as well ad it moving from one chan to the other, may indicate a sort of ground loop issue, perhaps cross channel ground loop. The hifisonix article you have read provides some tests for this. PS: Ensure good conductivity between all chassis panels. Panels with high impedance to gnd can also increase noise, especially with open inputs

5: Your rectifiers are also close to the amp pcb’s. Moving them either close to the center of the chassis, or radically to the front of the chassis (like in FW designs), may provide an improvement too.

6: Connecting the ac wire shield to the same place as the IEC inlet gnd/safety/chassis gnd may also help.
 
Last edited:
#8,704
tizman said:
When shorting plugs or cables from the source are plugged in, there is a slight hum/buzz in the previously silent channel, and a louder hum buzz in the noisy channel.
Click to expand...
You should do a ground-lif modification, most definitely.
tizman said:
My guess is that there isn't a grounding issue, and that there is a bad component or soldering problem on the one noisy board.
Click to expand...
Most likely an input differential pair needs to be replaced. Did you use an antistatic wrist strap (properly grounded)? You have to use it when handling the JFETs. Once they (the JFETs) are soldered onto the PCB, you can remove the wrist strap.
tizman said:
When I built the amp boards a long time ago, the BOM I used called for a 500 Ohm pot in the the LTP Bias space. This was apparently an error, as that is supposed to sit at 1K Ohm, so I replaced the pots with 1K resistors without removing the boards. I did this by rocking the pots back and forth slowly (they were not installed tightly to the board) until they snapped off, and then I put in a 1K Ohm resistor in their place. Unfortunately, on the board I'm having a problem with, I accidently removed the 2K Ohm pot at R7 instead of the LTP Bias pot. I replaced that pot with another pot, but did it without removing the board, and didn't do a good job. When I removed the R30 jumper and R7 to get to the pot, I tore out the trace on one side of the R30 jumper. All in all a mess that wouldn't have happened had I done the right thing and removed the board and done it properly. I am going to test voltages first as per your post, and then report back. Here are some closeups of the interior....
Click to expand...
This looks burnt to me...

1665471203509.png


You are kind of in a bit of a pickle. You may need a brand-new PCB... painful.. I know....
 
#8,705
Extreme_Boky said:
Just to add to what's already provided in the above posts, there are 3 critical voltages on the schematics that will set the DC operating values of the whole amplifier. Make sure they are all good. Double-check the bias through the MOSFETs and confirm that R16 - R19 are all good (0.47ohms).

Aleph J may require a ground lift resistor when used with a single-ended RCA input, to team the slight hum/buzz. This hum/buzz will be audible on both channels - equally, and will not cause what you are experiencing - which is the loud motorboat noise when turning the amplifier OFF. The most likely root cause for this motorboat noise is the input differential pair... and possibly the active current source transistor Q4.

You could also use a good working side (PCB) as a reference point, i.e. compare the bad side with a good working side.

I'd do the ground lift mod first, replace the Q4 second, and then replace the JFETs last.... just make sure what I mentioned above is all good first. Or, do all 3 in one go... while the PCB is removed.

Post some close-up well-lit photos. Make sure the braided shield mains wire is not resting on top of those filter caps.

Good luck. The build looks great - such a shame you have that issue with one channel.

View attachment 1097859
Click to expand...
Thank you all for your help with this! I measured the voltages according to your annotated schematic, and got the following…
-across R7 was 5V for the good channel and 4.9V for the bad
-across R8 was 8.5 for both channels
-from R10 to R17 was 4.6 for the good channel and 4.5 for the bad
 
#8,707
tizman said:
Thank you all for your help with this! I measured the voltages according to your annotated schematic, and got the following…
-across R7 was 5V for the good channel and 4.9V for the bad
-across R8 was 8.5 for both channels
-from R10 to R17 was 4.6 for the good channel and 4.5 for the bad
Click to expand...
The drop across R7 should be 4.3V; 5V is too much.
Drop across R8 seems okay.
You did not measure the voltage across Q4 - the collector-emitter voltage...? (be careful when measuring this)
 
#8,709
andynor said:
Three things:

1: Hassle, but if you twist your DC wires from PSU to amp PCB you can achieve an even lower noise floor.

2: You have connected audio gnd wires between amp boards and PSU at opposite sides of the junction on the PSU. Moving them to the same side of the PCB, and as far away from the junction (connection between the two psu board halves) as possible can achieve a quieter ground.

3: Have you tried rotating the transformer? The noise might be caused partly by magfields from the transformer, and some areas may radiate more than others.

4: I might not remember correctly, but the difference in noise with and without inputs as well ad it moving from one chan to the other, may indicate a sort of ground loop issue, perhaps cross channel ground loop. The hifisonix article you have read provides some tests for this. PS: Ensure good conductivity between all chassis panels. Panels with high impedance to gnd can also increase noise, especially with open inputs

5: Your rectifiers are also close to the amp pcb’s. Moving them either close to the center of the chassis, or radically to the front of the chassis (like in FW designs), may provide an improvement too.

6: Connecting the ac wire shield to the same place as the IEC inlet gnd/safety/chassis gnd may also help.
Click to expand...
Thanks. To your points…
1. I can and will do this as I have enough slack to do so.
2. I will do this as well.
3. I have not tried rotating the transformer. I have enough slack to do about a quarter turn in one direction (clockwise with respect to my photos)
4. The chassis panels are screwed together in multiple places. There should be good continuity between panels, but I’ll check it with a meter.
5. I can move the rectifiers, but I am inclined to move the entire power supply out of the chassis and put it in another separate chassis and use an umbilical. Would this be a better path to take?
6. I tried listening with and without the shield wire connected to the chassis, and it made no difference, but if it’s best practice, I will connect it to the main chassis ground.

That said, the amp isn’t just humming. It also either makes a loud motorboat sound that starts fast and slows down, or a loud high pitched whine on shut down. My concern is that working on the hum and fixing it seems unlikely to fix the other problem. Incidentally, the amp sounds really great when playing music despite the hum and shut off drama.
 
#8,713
The bad channel:
MOSFETs and 0.47-ohm resistors seem to be okay. The motorboat noise is most likely due to the bad JFETs, but it could be because of the CCS transistor going bad... the CE voltage of 5 V seems too much. Replace all 3 transistors.

The good channel:
The 0.414 seems to be too much. Either the MOSFET is bad/dying, or the 0.47-ohm resistor is shorting... was this measured across the resistor that I pointed out (circled in red) in one of my previous posts?

I am not sure what's the story with the PCB tracks; have you repaired them?

I am assuming that you already checked your soldering... on either amp PCBs...??
 
#8,714
Extreme_Boky said:
You should do a ground-lif modification, most definitely.

Most likely an input differential pair needs to be replaced. Did you use an antistatic wrist strap (properly grounded)? You have to use it when handling the JFETs. Once they (the JFETs) are soldered onto the PCB, you can remove the wrist strap.

This looks burnt to me...

View attachment 1098643

You are kind of in a bit of a pickle. You may need a brand-new PCB... painful.. I know....
Click to expand...
This resistor is R18 on the bad channel. It measures at .409V. The resistor at .414V is R19 on the other good channel. The darkness on the resistor in this photo was just some dirt that wiped off. It’s not hot at all in operation. Here it is rubbed clean…
AAFEF89D-1BC2-4D39-8F21-224FEF1F06EB.jpeg
 
#8,715
Extreme_Boky said:
The bad channel:
MOSFETs and 0.47-ohm resistors seem to be okay. The motorboat noise is most likely due to the bad JFETs, but it could be because of the CCS transistor going bad... the CE voltage of 5 V seems too much. Replace all 3 transistors.

The good channel:
The 0.414 seems to be too much. Either the MOSFET is bad/dying, or the 0.47-ohm resistor is shorting... was this measured across the resistor that I pointed out (circled in red) in one of my previous posts?

I am not sure what's the story with the PCB tracks; have you repaired them?

I am assuming that you already checked your soldering... on either amp PCBs...??
Click to expand...
I adjusted the voltages on the good channel. Please see attached photo.
DE2CC170-066A-4A50-AE51-1B7664EBD459.jpeg

I think the track repair was good. I will remove the PCB and make sure that it is. By replace all three transistors, are you referring to the two jfets and the CCS transistor at Q4 on the bad channel? Q4 C to E measures close to 5V on both channels, and the transistor that was .414V is R19 on the good channel. It is now .409V after adjustment. It seems that with the adjustment, both channels show similar spread between the transistors.
 
You must log in or register to reply here.
Top Bottom