What are the pro's and con's of each method?
A. 2 diodes and bypass capacitor
B. Resistor and bypass capacitor
A. 2 diodes and bypass capacitor
B. Resistor and bypass capacitor
The diodes (usually a 35A diode block) will handle much more fault current
than a resistor, and will decouple any ground loop hum better also.
than a resistor, and will decouple any ground loop hum better also.
The B+ fuse will not protect from an AC line short to the chassis. The AC line fuse is for that.
Most fuses are not rated for adequate DC levels (32VDC) to be safe for the B+ fuse.
Most fuses are not rated for adequate DC levels (32VDC) to be safe for the B+ fuse.
Last edited:
Question 1:
Can you list the equipment that will be connected together.
Example 1:
CD player, preamp, power amp, loudspeaker
Example2:
Guitar, Microphone, Guitar Amp, loudspeaker.
Example3:
Please fill in the blanks of your system ______________________
Question 2:
What are you trying to accomplish?
Safety?
Minimum Hum Ground loops?
Something else, fill in the blanks _____________________
Can you list the equipment that will be connected together.
Example 1:
CD player, preamp, power amp, loudspeaker
Example2:
Guitar, Microphone, Guitar Amp, loudspeaker.
Example3:
Please fill in the blanks of your system ______________________
Question 2:
What are you trying to accomplish?
Safety?
Minimum Hum Ground loops?
Something else, fill in the blanks _____________________
1. dac->preamp->poweramp-> loudspeaker.
This is for a SE amp construction. Wiring up the PS now. Earth ground from mains to chassis. Currently B+ is tied to ground bus 'bar" (12ga copper wire).
Looking for best method to tie DC ground to earth ground.
Looking to accomplish a best practice.
This is for a SE amp construction. Wiring up the PS now. Earth ground from mains to chassis. Currently B+ is tied to ground bus 'bar" (12ga copper wire).
Looking for best method to tie DC ground to earth ground.
Looking to accomplish a best practice.
I do not use a separate DAC.
I do not use a line level Preamp.
But I live in the US, and this works for me:
Sources:
CD Player
Tuner
Turntable, with the turntable ground wire to the phono preamp ground screw.
Wall switcher supply and Phono Preamp
All of the above have 2 wire plugs, not 3 wire plugs.
Most of the time, I only connect one pair of RCA cable from one signal source to the power amplifier at one time (no multiple RCA cable hum loops).
Amplifiers:
Vacuum Tube Power amplifiers, with IEC 3 wire power cord (the only 3 wire power connection of all my audio system devices).
I use RCA to RCA shielded cables to connect the units together.
Power amp:
The IEC socket ground ties directly to the chassis.
The vacuum tube B+ negative *, RCA input connector return *, and output transformer Common *, are eventually tied to the chassis ground.
Use plenty of capacitance in the B+ supply.
Generally, LCRCRC, or CLCRC.
Using an 8 Ohm load resistor, and testing the power amp by itself:
The amplifier has less than 100uV hum.
Works for Single Ended, works for Push Pull.
I have never heard any hum from my loudspeakers, using these methods.
Now, about the *.
* The center tap of the B+ secondary connects to the negative of the first filter cap. The negative of the first filter cap ties to the negative of the second filter cap. The negative of the second filter cap, ties to the Chassis Ground (in that "serial" order). That reduces the B+ hum ground loops.
* The RCA input connector Return ties to the bottom of the input tube's self bias parallel RC network. The bottom of the self bias parallel RC network connects to the Chassis Ground (serially in that order). And, the input tube g1 Rg ties to the bottom of the input tube self bias RC network. That reduces the Input circuit hum ground loop.
* The output transformer secondary Common connects directly to the Chassis ground.
* Finally, the output tube g1 grid resistor ties to the bottom of the output tube self bias parallel RC network. Then the bottom of the RC network then connects to the chassis ground (serially in that order).
How do I reduce hum?
1. Follow the above connection orders.
2. Pay attention to the spacing and orientation of the output transformer, versus the placement and orientation of the power transformer and B+ choke.
3. Do not use a Magnetic Chassis. That ruins the good planning of # 2. above.
All of the signal sources are designed to be safe with 2 wire power cords (UL, CSA, etc.).
The power amps all originally used 2 wire power cords, but now use 3 wire IEC power cords, and plug into approved 3 wire grounded power mains.
Safety first!
No dead person ever enjoyed their stereo, perhaps their surviving spouse enjoyed it.
I do not use a line level Preamp.
But I live in the US, and this works for me:
Sources:
CD Player
Tuner
Turntable, with the turntable ground wire to the phono preamp ground screw.
Wall switcher supply and Phono Preamp
All of the above have 2 wire plugs, not 3 wire plugs.
Most of the time, I only connect one pair of RCA cable from one signal source to the power amplifier at one time (no multiple RCA cable hum loops).
Amplifiers:
Vacuum Tube Power amplifiers, with IEC 3 wire power cord (the only 3 wire power connection of all my audio system devices).
I use RCA to RCA shielded cables to connect the units together.
Power amp:
The IEC socket ground ties directly to the chassis.
The vacuum tube B+ negative *, RCA input connector return *, and output transformer Common *, are eventually tied to the chassis ground.
Use plenty of capacitance in the B+ supply.
Generally, LCRCRC, or CLCRC.
Using an 8 Ohm load resistor, and testing the power amp by itself:
The amplifier has less than 100uV hum.
Works for Single Ended, works for Push Pull.
I have never heard any hum from my loudspeakers, using these methods.
Now, about the *.
* The center tap of the B+ secondary connects to the negative of the first filter cap. The negative of the first filter cap ties to the negative of the second filter cap. The negative of the second filter cap, ties to the Chassis Ground (in that "serial" order). That reduces the B+ hum ground loops.
* The RCA input connector Return ties to the bottom of the input tube's self bias parallel RC network. The bottom of the self bias parallel RC network connects to the Chassis Ground (serially in that order). And, the input tube g1 Rg ties to the bottom of the input tube self bias RC network. That reduces the Input circuit hum ground loop.
* The output transformer secondary Common connects directly to the Chassis ground.
* Finally, the output tube g1 grid resistor ties to the bottom of the output tube self bias parallel RC network. Then the bottom of the RC network then connects to the chassis ground (serially in that order).
How do I reduce hum?
1. Follow the above connection orders.
2. Pay attention to the spacing and orientation of the output transformer, versus the placement and orientation of the power transformer and B+ choke.
3. Do not use a Magnetic Chassis. That ruins the good planning of # 2. above.
All of the signal sources are designed to be safe with 2 wire power cords (UL, CSA, etc.).
The power amps all originally used 2 wire power cords, but now use 3 wire IEC power cords, and plug into approved 3 wire grounded power mains.
Safety first!
No dead person ever enjoyed their stereo, perhaps their surviving spouse enjoyed it.
Last edited:
Any thoughts on using a MOV instead of diodes? They are typically UL listed and have well defined energy handling specs. Something like a V18ZA40P?
After all I said above in Post # 7 . . .
Why isolate the B+ ground from the chassis?
Why isolate the chassis from earth ground?
What is the purpose or function of isolating?
Anything other than a straight wire is at least partially isolated.
Using a UL listed part incorrectly does not make a circuit UL approved.
I have worked with UL many times.
Define the problem first.
Then solve the problem.
I have not heard what the problem is yet.
Eventually, without more information of the problem, tie the Mains ground directly to the chassis. Tie the B+ return directly to the chassis.
Directly, means use a wire that will not open up, if there is any short that violates safety, no matter how high the current.
Prevent the "Surviving Spouse Syndrome".
No Joke.
Why isolate the B+ ground from the chassis?
Why isolate the chassis from earth ground?
What is the purpose or function of isolating?
Anything other than a straight wire is at least partially isolated.
Using a UL listed part incorrectly does not make a circuit UL approved.
I have worked with UL many times.
Define the problem first.
Then solve the problem.
I have not heard what the problem is yet.
Eventually, without more information of the problem, tie the Mains ground directly to the chassis. Tie the B+ return directly to the chassis.
Directly, means use a wire that will not open up, if there is any short that violates safety, no matter how high the current.
Prevent the "Surviving Spouse Syndrome".
No Joke.
Last edited:
Any thoughts on using a MOV instead of diodes? They are typically UL listed and have
well defined energy handling specs. Something like a V18ZA40P?
MOVs are for clamping voltage spikes, for a brief duration only.
After all I said above in Post # 7 . . .
Why isolate the B+ ground from the chassis?
No Joke.
My question exactly. It seems lik e a common practice. PMillett does it in his Engineers Amplifier. Why? What problem could that solve?
If you choose to isolate signal ground from PE (safety ground) you'll need to be sure that whatever you do can still trip the panel breaker. That's where the protection occurs. Double diodes (or whatever else) need to be able to survive the fault long enough to get the breaker tripped.
In many home setups, heroic measures to break ground loops are unnecessary. If all equipment is plugged into a single AC power strip, including connecting the shield of any coax coming in from the outside world to PE at the strip, then all safety ground loops are small and local enough to be a non-issue. Usually.
But, if you absolutely can't do that - a remote subwoofer or connection to stuff in another room, or whatever - then you want something to isolate signal ground from safety ground, but only at small potential differences. Above a Volt or so this something needs to crowbar fast and survive long enough to trip the breaker. Double diodes in both polarities can do this.
YOS,
Chris
In many home setups, heroic measures to break ground loops are unnecessary. If all equipment is plugged into a single AC power strip, including connecting the shield of any coax coming in from the outside world to PE at the strip, then all safety ground loops are small and local enough to be a non-issue. Usually.
But, if you absolutely can't do that - a remote subwoofer or connection to stuff in another room, or whatever - then you want something to isolate signal ground from safety ground, but only at small potential differences. Above a Volt or so this something needs to crowbar fast and survive long enough to trip the breaker. Double diodes in both polarities can do this.
YOS,
Chris
Why? What problem could that solve?
A "ground loop" is an unexpected path that can inductively pick up noise from surroundings and mix it into the signal path. An example might be a preamp connected to an amp, both with third wire safety connections, plugged into different AC outlets.
This has a "hidden" path: from preamp chassis to signal ground, through the signal interconnect wires to the amp's signal ground to its chassis. From chassis, PE (the safety ground) runs back to panel, connects to the PE running to the preamp's chassis and finally to its signal ground. This long loop acts like a turn on a very large but poor quality transformer, injecting some noise into signal.
YOS,
Chris
"Grounds are Commonly Misunderstood"
A wire is a resistor
A wire is an inductor
Current through a wire = voltage
A voltage that varies with time is a signal.
Signals can be heard (even if the signal is undesired hum).
Ground Loops are Commonly Misunderstood too.
They are caused by resistors, inductors, and current (wires with current through them).
A wire is a resistor
A wire is an inductor
Current through a wire = voltage
A voltage that varies with time is a signal.
Signals can be heard (even if the signal is undesired hum).
Ground Loops are Commonly Misunderstood too.
They are caused by resistors, inductors, and current (wires with current through them).
If you choose to isolate signal ground from PE (safety ground) you'll need to be sure that whatever you do can still trip the panel breaker. That's where the protection occurs.
Completely agree.
What about elevating heater CT? Is that a different thread? Should connect the bottom of the voltage divider to DC ground or at the point where DC ground is tied to chassis ground?
What about elevating heater CT? Is that a different thread? Should connect the bottom of the voltage divider to DC ground or at the point where DC ground is tied to chassis ground?
Using Pete Millett's schematic tie the bottom of the divider to wherever you tie the B+ ground.

There is a 1956 application note from The Application Engineers of The Advisory Group on Electron Tubes distributed by RCA called "Heater-Cathode Leakage" and it has a good explanation of how elevating the heaters helps reduce hum and why >10VDC is preferred.
I have a respect for a belt & suspenders attitude, as in building to avoid problems that may never even turn out to be an issue.
However, I find your suggestions for a loop breaker too lightweight. As has been stated, you need to ensure a fault will trip the breaker. Have a look at Rod Elliot's design, the redundancy and parts are there for a reason: Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
However, I find your suggestions for a loop breaker too lightweight. As has been stated, you need to ensure a fault will trip the breaker. Have a look at Rod Elliot's design, the redundancy and parts are there for a reason: Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
Do Not count on a short circuit to trip the power mains circuit breaker, consider this . . .
A short that now connects to an un-grounded floating chassis that you touch, and another part of you touches something else that IS grounded (like another chassis that uses the IEC 3rd wire) . . .
your body will not trip the breaker.
Instead, the current from the still closed breaker will be "breaking you".
Shocking!
Check All possible scenarios, before you say 'it is safe' . . . famous last words.
As I said, Safety First!
Please remember what can happen to a floating chassis that has some "safety" parts from ground to the chassis, which are intended as a safety shunt.
If those parts can only shunt 7 amps from the chassis to ground, that will not overcome a 10, 15, 20, or 25 Amp circuit breaker. Those "safety parts" will break open.
A short that now connects to an un-grounded floating chassis that you touch, and another part of you touches something else that IS grounded (like another chassis that uses the IEC 3rd wire) . . .
your body will not trip the breaker.
Instead, the current from the still closed breaker will be "breaking you".
Shocking!
Check All possible scenarios, before you say 'it is safe' . . . famous last words.
As I said, Safety First!
Please remember what can happen to a floating chassis that has some "safety" parts from ground to the chassis, which are intended as a safety shunt.
If those parts can only shunt 7 amps from the chassis to ground, that will not overcome a 10, 15, 20, or 25 Amp circuit breaker. Those "safety parts" will break open.
Last edited:
- Home
- Amplifiers
- Tubes / Valves
- DC to earth ground isolation