Great Article on Cables

[Resident Loser]

...dumbness...are you saying replace all the PCB "common access" power points with dedicated point-to-point wiring?

jimHJJ(...or am I just a complete boob?...)

[jneutron]

...dumbness...are you saying replace all the PCB "common access" power points with dedicated point-to-point wiring?

jimHJJ(...or am I just a complete boob?...)

Any wire within the amplifier which creates a magnetic field external to that wire, or is capable of receiving an external magnetic field, is a culprit. A star ground, while good practice, can allow the non current carrying grounds to intercept mag field from the current carrying ones..that needs to be fixed..

The most significant ones to broadcast are the output rail/output lines, the supply feeds for that output (all the low z stuff), the primary and secondary xfmr runs, the physical size of the supply caps, the power switch wires.

The significant receivers are the input run, and the feedback path.

Run the amp into a full load at 50% power, sine at 5 to 10 Khz..use a magnetic probe, and look for that signal in the amp box, outside the amp box, and (bet you didn't think of this), inside the source component near the output jacks back to the pc board.

If you can find magnetic signal, tis broken, fix it..

When you cannot find any stray field..you are done..

(good luck around those damn supply caps...)

Cheers, John

Oh, forgot this: It's not the IR drop, so that's not what I meant by replacing the wiring harnesses..it's the fact that the traces generate a dipole field..

What is needed is for the amp (and wire) designers to understand that it is not a case of tossing silver, or large wires, or teflon, or cryo, or single crystal yada yada at the problem, it's about engineering the solution.

[Resident Loser]

...I'd guess all the power wiring should be as segregated as practicable from the signal traces (or wiring) and not run parallel to them at any point...good practice without AND within.

Certainly makes sense..I have an old Fender Bandmaster guitar amp that is somewhat susceptible to hum if you move the interenal wiring a bit too far from it's intended position...and it's all "breadboard" style, point-to-point wiring and all...I'd hazard a guess they arrived at their wiring layout pretty much through trial and error in an effort to optimize s/n and hummmmm. Forty years old and still cookin'!

jimHJJ(...and I do like the idea of the hardwired power cord...)

[jneutron]

...I'd guess all the power wiring should be as segregated as practicable from the signal traces (or wiring) and not run parallel to them at any point...good practice without AND within.

Certainly makes sense..I have an old Fender Bandmaster guitar amp that is somewhat susceptible to hum if you move the interenal wiring a bit too far from it's intended position...and it's all "breadboard" style, point-to-point wiring and all...I'd hazard a guess they arrived at their wiring layout pretty much through trial and error in an effort to optimize s/n and hummmmm. Forty years old and still cookin'!

jimHJJ(...and I do like the idea of the hardwired power cord...)

What is interesting about loop pickup it how it works. For example, take a flat 3 wire extension cord, and push a kilo through it, a space heater load..run the wire flat and straight..

The center wire of the cord is ground.hot and neutral on either side, like romex...

Run a second cord next to it, one inch away, but do not put a load into it..measure the ground to ground potential.

Next, put the unpowered cord 6 inches away, and repeat the measurement..the voltage is MORE, not less..

Put the cord a foot away..the voltage is even MORE..(granted, the voltages for this example are not gonna be big..

What's going on? As you move the cord farther and farther away, the loop formed by the grounds are trapping more of the 60 hz flux within the loop...faraday's law states that the loop voltage is proportional to the rate of change of the flux within..so, in this case, distance is the enemy...the further away, the worse it gets...this is of course, asymptotically approaching a maximum, it doesn't keep getting bigger.

Faraday's law states.....RATE of change of the flux...60 hz is only the primary, most rectifier circuits draw haversines, 180 hz being very large...the flux rate of change is three times larger per amp..keep going up the odd harmonics..

Then, look at the line cord field with 1,5 or 10 Khz power being delivered to the load...with physically large capacitors...

It seems that your amp vendor had to find the minima coupling path by pushin the wires around to null it..hey, worked for them..

Cheers, John