Greetings, Friends. Hope the turkey was good. I was looking at schematics over at DiyAudioProjects.com and found Bruce Heran's DIY ECC802S (12AU7 / ECC82) Tube SRPP Preamp with a nicely drawn Power Supply schematic with an interesting detail I've been missing:
Where the Earth wire is clearly connected directly to the chassis for safety, the b+ ground is connected to the chassis through a resistor/capacitor in parallel, a 0.1uF Type X2 cap designed for this purpose and a 150R 1/2W Metal Film Resistor. In a linked text, Heran goes on to elaborate on his grounding scheme and directs the reader to connect both the power supply's ground and the signal returns to the chassis, at a single point, through this filter network. Connecting those grounds directly to the chassis would create ground loops, but the filter network blocks noise from entering the circuit while allowing the safety features of the Earth wire to function. Up til now I've been connecting all grounds directly to the chassis and now must determine how best to re-wire some projects to eliminate this direct connection and hopefully excise some noisy demons that have been vexing me.
Which leads me to the phono preamp. How should the ground wire from a 1200 mk2 turntable be connected to the phono preamp? Does it connect to the Chassis/Earth or should it connect to the lifted/isolated ground network in the preamp?
What about the heaters? If heaters should be referenced to Ground, should it be the Chassis/Earth or the isolated Ground? I usually reference the heater circuit to a power tube's cathode anyway.
I've also seen talk about other methods of isolating the ground, including diode bridges and 5W wirewound resistors; any thoughts on those?
thanks for taking a look!
Where the Earth wire is clearly connected directly to the chassis for safety, the b+ ground is connected to the chassis through a resistor/capacitor in parallel, a 0.1uF Type X2 cap designed for this purpose and a 150R 1/2W Metal Film Resistor. In a linked text, Heran goes on to elaborate on his grounding scheme and directs the reader to connect both the power supply's ground and the signal returns to the chassis, at a single point, through this filter network. Connecting those grounds directly to the chassis would create ground loops, but the filter network blocks noise from entering the circuit while allowing the safety features of the Earth wire to function. Up til now I've been connecting all grounds directly to the chassis and now must determine how best to re-wire some projects to eliminate this direct connection and hopefully excise some noisy demons that have been vexing me.
Which leads me to the phono preamp. How should the ground wire from a 1200 mk2 turntable be connected to the phono preamp? Does it connect to the Chassis/Earth or should it connect to the lifted/isolated ground network in the preamp?
What about the heaters? If heaters should be referenced to Ground, should it be the Chassis/Earth or the isolated Ground? I usually reference the heater circuit to a power tube's cathode anyway.
I've also seen talk about other methods of isolating the ground, including diode bridges and 5W wirewound resistors; any thoughts on those?
thanks for taking a look!
The turntable ground wire goes directly to the preamp chassis.
Heaters are referenced to the audio ground, not the chassis.
High current diodes (one in each direction) can go in parallel with the resistor/capacitor to the chassis.
This shows a 35A diode block being used.
Heaters are referenced to the audio ground, not the chassis.
High current diodes (one in each direction) can go in parallel with the resistor/capacitor to the chassis.
This shows a 35A diode block being used.
It's important to remember that any isolating elements between signal ground and chassis/PE ground must be able to survive a worst-case fault, likely B+ level to chassis, long enough to open the fuse, and be low enough impedance to ensure the fuse opens reliably.
The proposed 150R of unknown wattage rating may not do this, and shouldn't be trusted. Large back-to-back diodes may not be code or UL, but are our best choice. The small (typically 10R) resistor in parallel is just there to overcome small residual leakages, keeping the normal, non-fault, voltage across the diodes near zero. It can be small wattage without danger.
All good fortune,
Chris
The proposed 150R of unknown wattage rating may not do this, and shouldn't be trusted. Large back-to-back diodes may not be code or UL, but are our best choice. The small (typically 10R) resistor in parallel is just there to overcome small residual leakages, keeping the normal, non-fault, voltage across the diodes near zero. It can be small wattage without danger.
All good fortune,
Chris
Last edited:
Heran identifies the 150R resistor as being a common 1/2W Metal Film type. I certainly want that fuse doing its job in a timely manner.
Therefor normally 10 Ohm 2W is chosen (when the Graetz bridge is omitted). It also helps to use the correct terms and to have some carved in stone rules.
A metal casing/chassis is always connected directly to PE. Audio GND (secondary side of PSU) can be referenced to PE by means of the resistor/cap combination. An error from the past is to use chassis as GND but the casing/chassis should not be part of the audio circuit.
The idea is to have both the standard safety measures and no hum or ground loops.
A metal casing/chassis is always connected directly to PE. Audio GND (secondary side of PSU) can be referenced to PE by means of the resistor/cap combination. An error from the past is to use chassis as GND but the casing/chassis should not be part of the audio circuit.
The idea is to have both the standard safety measures and no hum or ground loops.
Last edited:
Also, Best Practice requires that the Earth/PE ground to chassis have its own bolt, nut and lockwasher, with nothing else attached there. Might seem excessive, but we should all do it anyway.
Goes without saying that transformer cases and (if present) separate chassis pieces in fancy chassis all need to be bonded to the chassis piece with mains inlet (and PE through mains), anodization scaped off and bolt, nut, lockwasher.
All good fortune,
Chris
Goes without saying that transformer cases and (if present) separate chassis pieces in fancy chassis all need to be bonded to the chassis piece with mains inlet (and PE through mains), anodization scaped off and bolt, nut, lockwasher.
All good fortune,
Chris
Very good addition. Normally chassis connection with M4 bolt/nut/washers/serrated washer combo suffices but I have had situations where paint or anodizing made the front and back panels isolated from chassis.
Bold statement but wrong grounding and wrong PE connections are mostly done in the tube world, also in new devices that miraculously escape a technical approvement eye when imported. This is your hum issue folks, check your devices and do it right and your stuff is safe & silent.
Bold statement but wrong grounding and wrong PE connections are mostly done in the tube world, also in new devices that miraculously escape a technical approvement eye when imported. This is your hum issue folks, check your devices and do it right and your stuff is safe & silent.
Last edited:
Ya, dedicated bonded connection. Star washer. I even got a counter-bore bit to scrape the paint off in a nice neat circle.
What is house in this image? As in house earth? If so, it is wrong as chassis must always be directly connected to earth. Every other gnd connecting to chassis is lifted as in the picture.
It goes wrong when not using the right terms. As proven. L, N, PE, GND.
Calling both GND and PE ground is where the misunderstandings (and the very often occurring hum) come from. Plus stubbornly using terms and practices from the twenties/thirties of last century. It works yeah but please do not expect it to be HiFi.
Calling both GND and PE ground is where the misunderstandings (and the very often occurring hum) come from. Plus stubbornly using terms and practices from the twenties/thirties of last century. It works yeah but please do not expect it to be HiFi.
Last edited:
The various ground symbols also add to the confusion:
https://www.eeweb.com/do-you-know-your-ground-symbols/
It is also common to mark the signal ground as a thick horizontal line.
https://www.eeweb.com/do-you-know-your-ground-symbols/
It is also common to mark the signal ground as a thick horizontal line.
It could be easy but the habit of certain countries to ignore agreed upon terminology and practices don't make matters any better. Then mixing up applications as a consequence of wrong naming. I solved a lot of such issues in the past as I desperately wanted to understand why my DIY stuff always hummed. Decades later and I still see even professionals mixing up stuff. Then a lot of myth forming and clever gurus making matters more complicated.
It is not. Just connect stuff with metal chassis/casing to PE and reference Audio GND to PE. The latter definitely with higher supply voltages. In for instance 5V 5VA devices stuff may be floating. ALWAYS connect a metal chassis/casing to PE though. And using wooden casings with unshielded contents in todays RF ridden aether is well, eh .....
It is not. Just connect stuff with metal chassis/casing to PE and reference Audio GND to PE. The latter definitely with higher supply voltages. In for instance 5V 5VA devices stuff may be floating. ALWAYS connect a metal chassis/casing to PE though. And using wooden casings with unshielded contents in todays RF ridden aether is well, eh .....
Last edited:
Many tube pilots use a steel chassis as a conductor for their audio circuits. That is what you get when ground and earth are used as names, safety is seen as a futile detail and twenties/thirties practices are not abandoned.
Is it HiFi? Maybe but there are clear reasons why the tube section has many (if not absolute most) hum/grounding threads or mentioning and accepting hum in circuits. In fact most DIY tube amplifiers and tube preamps I get to see hum (and also usually have extreme high gain and plain wrong signal levels and impedances often not even adapted to each other). Amplifying, then severe attenuating to then amplifying again is common.
Is it HiFi? Maybe but there are clear reasons why the tube section has many (if not absolute most) hum/grounding threads or mentioning and accepting hum in circuits. In fact most DIY tube amplifiers and tube preamps I get to see hum (and also usually have extreme high gain and plain wrong signal levels and impedances often not even adapted to each other). Amplifying, then severe attenuating to then amplifying again is common.
Last edited:
General mistake to "use" CT of secondary or first capacitor negative pole as "ground". Here is the worst noise level.
Read this:
Read this:
Check the schematic you posted again: the author uses a voltage divider from B+ to elevate the heaters above cathode. That way no noise from the heaters will modulate the cathode potential and show up on the anode.