5B3 deluxe clone

[goldstache]

Wrenching on what appears to be a 5B3 Deluxe, but with 6sl7 subbed for 6sc7.


Amp had a fried cone.


But, something caught my eye, as the HT CT was fused.


Then I realized that the whole B- and audio grounds are lifted from chassis (mains ground) by a lift network.
-.047 600v for ac ground.
-100R 2W lift from mains gnd
-back to back 1n4007’s


All the above in parallel.


I’ve tried lifting the B- on fender type reverbs for noise that ya battle halfwave.


But never seen it on a power amp type setup.
 (Yeah I know the spring driver has power tube/bias)


So now I’m wondering.
Take me to school here.


Wondering if there’s any difference to voltage calcs.


Lifted B- shows 0v to chassis. But 100R of lift.


So it’s just potentials, and their disparity helping to mitigate current path noise?


I think I’m spooking myself thinking that in combo bias say, the bias needs calculated for cathode/fixed cumulative.


So I’m thinking everything is off that mains 0v. But doesn’t measure any potential difference in DC.


Trying, poorly here, to wrap my brains around it.
Thanks
-joe

[HotBluePlates]

... I realized that the whole B- and audio grounds are lifted from chassis (mains ground) by a lift network.
-.047 600v for ac ground.
-100R 2W lift from mains gnd
-back to back 1n4007’s

All the above in parallel.

... never seen it on a power amp type setup.
 ...
So now I’m wondering.
Take me to school here.
...
So I’m thinking everything is off that mains 0v. But doesn’t measure any potential difference in DC.
...

Scroll to "Page 275" for the Ground Lift section Merlin wrote.

He shows a typical setup of 10Ω, 0.1µF and at least a pair of anti-parallel diodes (the drawing shows a bridge with the "middle terminals connected" probably because it is a readily-available part).

All you have is someone adding a ground-lift to a vintage(-style) amp to kill hum when more than one thing is connected together, and each has its own chassis-ground connected to the power cord.

   - If this is a vintage amp, they could achieve the same using a Humdinger or P-Split that galvanically-isolate the grounds of the 2 pieces of equipment using a transformer in the audio cabling prior to the gear's Input Jack.  This is also useful for any amp one might buy already assembled, because adding a ground-lift to an already-built amp is a major undertaking.

   - If this is a new build or clone, it is a reasonable that they re-thought grounding for the gear and incorporated a ground-lift when building.

Lifted B- shows 0v to chassis. But 100R of lift.

If you're not actually measuring "100Ω" then the builder goofed somewhere and has an item grounded to the chassis.

If you are measuring 100Ω to chassis, then the lift is functioning (though I've found this is noisy in one of my amps, if I don't have the 2nd piece of gear making the direct chassis-to-ground-wire connection).

You shouldn't measure any voltage unless there is leakage current from B+ ground to chassis.  And this "leakage current" is more likely to happen when multiple pieces of gear are cabled together to create a large ground-loop.  That's because they're more likely to have very slightly different potentials from their chassis to the actual Ground at the outlet(s).

The point of the resistor is that "different chassis potential" (a voltage) is dropped across the resistor, and the resistance reduces the "resulting ground current" to very low levels.  If you didn't measure, and the resistor's color code is Brown-Black-Black, then that's 10Ω (or maybe the builder only had 100Ω handy from artificial center-tap use).

The cap bypasses the resistor for high-AC-frequency, to nix RF noise.

The anti-parallel diodes ensure that the voltage-difference between the 2 chassis cannot be more than about 0.6v, as a safety precaution against gross faults.

[goldstache]

This is so concise. Thank you.
Appreciate it.

[bnwitt]

Here is a Visio layout of that amplifier and a PDF