on existing board pinout isn't adequate for 2SJ109
yes, 2SJ109 is "factory made" pair of 2SJ74
yes, 2SJ109 is "factory made" pair of 2SJ74
I have a question, the current j-fet/mosfet kit from the DIY audio store that I received comes with IRFP048 for the mosfets and S09014/S09015 in place of ZTX450/550. Are there any changes to the BOM needed for this combination? After building an M2X with the 8 daughter boards with no problem, I figured I would take a stab at the Aleph J. Upon startup with one board, I needed to turn the offset pot quite a lot to get it d to 350-400mV, when I checked the DC offset, it was 18v. I started adjusting the pot and R17 started smoking and took out 3 of 4 mosfets. I was experiencing similar issues with the second board, but didn’t want to push things to far. Lifted each resistor and the values checked out, not sure where to look next?
Last edited:
May have figured out the issue, looks like the 4 LSJ74 may have been manufactured backwards? Not sure if this is an issue with my tester? Is there a better way to test these?
JFets are symmetrical, speaking of D and S
though, ZM Chicken, always taking datasheet as Bible
in short - nothing wrong with these JFets
post pictures of your build, so we can see ......
go back few pages, find schematic I recently posted - whats proper arrangement of various "optional" positions
though, ZM Chicken, always taking datasheet as Bible
in short - nothing wrong with these JFets
post pictures of your build, so we can see ......
go back few pages, find schematic I recently posted - whats proper arrangement of various "optional" positions
The quas-engl-ish instructions that came with my tester say to use a protection diode to test JFETs. Don't ask me how I know this.
Found the problem, and I get the idiot award of the day. Was turning the wrong pot to try setting the offset. Have some parts coming my way to fix the ones I fried. Thanks all for your input.
https://www.diyaudio.com/community/threads/aleph-j-illustrated-build-guide.241729/post-7398484
Referencing @Zen Mod 's point about wiring being far from optimal (CRCRC linear power supply was of my own concoction with larger than necessary can capacitors and power resistors.
The issue was buzz / hum, both channels, pronounced enough on the left channel to prompt resolution. Original PS components weren't going to allow for buzz free wiring, and I did not want to hack at workarounds knowing this is a proven circuit. Make it work, correctly, before resorting to more desperate measures. After much reading here it was obvious to change out the homegrown PS for a proven FirstWatt design.
Went with the store Universal PS, no input or output snubber, 50V Nicochon 15000uF 35mm snap in caps which have an impressive weight and feel about them. Correct port/starboard LEDs. Placed the transformer dead middle of the chassis floor (elevated with spacer nuts per ZM). Screw terminal near the IEC/switch with blue safety cap.
Changed to:
SL15 22102
Ametherm Inrush Current Limiters
But kept the CL60 thermistors at the ceramic block between the PS board ground and star ground near the power switch, as I wanted the 10R value there as the ground lift.
I was careful to run the AC from the switch exactly per the schematic into the two input primaries and verified their continuity first. Twisted AC run under the chassis floor to the front of the amp. The Universal Power Supply board is bolted to the front chassis panel.
Rails are 30.5V both sides super consistent. Offset and bias (400mV) set well.
Noise floor is now acceptably low.
Referencing @Zen Mod 's point about wiring being far from optimal (CRCRC linear power supply was of my own concoction with larger than necessary can capacitors and power resistors.
The issue was buzz / hum, both channels, pronounced enough on the left channel to prompt resolution. Original PS components weren't going to allow for buzz free wiring, and I did not want to hack at workarounds knowing this is a proven circuit. Make it work, correctly, before resorting to more desperate measures. After much reading here it was obvious to change out the homegrown PS for a proven FirstWatt design.
Went with the store Universal PS, no input or output snubber, 50V Nicochon 15000uF 35mm snap in caps which have an impressive weight and feel about them. Correct port/starboard LEDs. Placed the transformer dead middle of the chassis floor (elevated with spacer nuts per ZM). Screw terminal near the IEC/switch with blue safety cap.
Changed to:
SL15 22102
Ametherm Inrush Current Limiters
But kept the CL60 thermistors at the ceramic block between the PS board ground and star ground near the power switch, as I wanted the 10R value there as the ground lift.
I was careful to run the AC from the switch exactly per the schematic into the two input primaries and verified their continuity first. Twisted AC run under the chassis floor to the front of the amp. The Universal Power Supply board is bolted to the front chassis panel.
Rails are 30.5V both sides super consistent. Offset and bias (400mV) set well.
Noise floor is now acceptably low.
Appending to my last post on noise floor reduction. After moving wires around, it seemed apparent getting that last bit of buzz completely eliminated wasn't going to happen. So therefore, try the 10R ground lift between the input ground and speaker ground. I had a nice Ohmite 3W resistor in stock with convenient embedded mounts. Picture shows how done without lifting the PCBs.
Note and question for you all: Since thorough reading indicated the solution here is to eradicate cross channel ground loop, I figured I try the ground lift on one channel only (R). Any issue with doing the ground lift on only one side?
Nearly forgot to report: absolutely no buzz or sound with an ear up to each driver. Remarkable!
Thanks!
Note and question for you all: Since thorough reading indicated the solution here is to eradicate cross channel ground loop, I figured I try the ground lift on one channel only (R). Any issue with doing the ground lift on only one side?
Nearly forgot to report: absolutely no buzz or sound with an ear up to each driver. Remarkable!
Thanks!
@Mulburg - Good question. No negatives as far as I hear. And for the good of the order full interior chassis pic:
Pass DIY Addict
Joined 2000
Paid Member
Adding that resistor is a nice little trick, huh?!? Glad to hear that you've gotten everything straightened out!
No issues with adding a resistor to one channel at a time. Now just enjoy how great this amp sounds!
No issues with adding a resistor to one channel at a time. Now just enjoy how great this amp sounds!
It seems that your speaker return (green wire) is routed back to the PS common (grey wire), via that 10ohm resistor, which is not correct...
All FW / Pass Labs amps (and all amplifiers ever made... in the whole world... a few thousands of which I helped design, make, test, cook, and test again when I worked for Jands Electronics a very long time ago) route speaker negative straight back to the PS PCB common. I suggest you do that first. Then, it will become obvious to you where you made a mistake.
In a nutshell, that grey wire needs to be connected to AMP PCB GND via a 10-ohm resistor, and the speaker negative green wire has to go straight back to PS PCB common...
The best would be to do the mod, as per below... then you can do whatever you want with the speaker negative return...
Here, the speaker negative and the PS PCB negative (two black wires) are tied together which is incorrect....
.....but, the ground lift resistor mod is executed correctly... notice how the speaker negative return and the PS PCB common are still connected together:
The above pictures were taken to show how to test if the ground mod would produce the desired result (no buzz), without removing the AMP PCB whilst having the speaker negative incorrectly routed via the AMP PCB.
All FW / Pass Labs amps (and all amplifiers ever made... in the whole world... a few thousands of which I helped design, make, test, cook, and test again when I worked for Jands Electronics a very long time ago) route speaker negative straight back to the PS PCB common. I suggest you do that first. Then, it will become obvious to you where you made a mistake.
In a nutshell, that grey wire needs to be connected to AMP PCB GND via a 10-ohm resistor, and the speaker negative green wire has to go straight back to PS PCB common...
The best would be to do the mod, as per below... then you can do whatever you want with the speaker negative return...
Here, the speaker negative and the PS PCB negative (two black wires) are tied together which is incorrect....
.....but, the ground lift resistor mod is executed correctly... notice how the speaker negative return and the PS PCB common are still connected together:
The above pictures were taken to show how to test if the ground mod would produce the desired result (no buzz), without removing the AMP PCB whilst having the speaker negative incorrectly routed via the AMP PCB.
Last edited:
Greetings from sunny California everyone. Quick question: anyone ever try swapping out the WIMA C1 from the BOM for something fancy? There's just so many fancy capacitors out there... Just wondering if anyone's had interesting results from trying stuff. Thanks!
