I recently upgraded from an old Hafler (DH110) preamp to a new Linn. The new pre uses a three prong grounded AC plug...the Hafler was a two prong ungrounded unit. The new pre has a very faint humm ( down 70db? ) when plugged into a grounded power strip. When I Ungrounded the preamp, the Linn was dead quiet ! I mean it was like it was not even turned on ....SUPER quiet.
Is there any risk running a preamp "ungrounded" ? My Halfer was happy being ungrounded for 20 years. I know grounded equipment has less risk of shock hazzard, but other than that is there any other consideration?
Is there any risk running a preamp "ungrounded" ? My Halfer was happy being ungrounded for 20 years. I know grounded equipment has less risk of shock hazzard, but other than that is there any other consideration?
Use a back to back diode safety floating scheme. What you do is you employ two 5A diodes in parallel with their strap marked ends joined in opposite. You put those at the grounding point you lifted the cable from, and you attach that grounding cable on the new end. Now you got a ground lift but in case of a problem, current can run in the safety ground since the diodes are going to conduct in any 0.7V and over incident. Else, if any chassis develops a potential due to a fault in any system, the house safety trip relay that senses ground will know nothing about it and its against any safety rule. Be generally cautious.
Does this actually work? Have you tried it?
By the way, the GFI/GFCI should still trip even if ground is isolated at an end point. Yes, it stands for Ground Fault Interrupter, but it actually trips when current out through hot doesn't match current returned through neutral. Ground doesn't enter into it.
A common source of ground loops I've seen is the difference between the ground of AC power and the ground provided by a cable TV cable. These come together in AV systems. One solution for this is to get a 75 Ohm isolation transformer, or make one by using a 75-300 Ohm balun connected to a 300-75 Ohm F-Plug adaptor. Adding this in-line with the incoming cable separates the ground but couples the signal.
Yes, it does work. I install them in every pro amp I build. But use a heavy duty full wave bridge. Strap + to -, then use the AC as the loop breaker. A high current bridge will handle very large fault currents without failing open, and will provide two diode drops instead of only one. Voltage rating does not matter so a 50V part that you can't use for an amp power supply will work.
Yes I have it in some of my DIY builds. It works. Its fairly known from some power amps.
GFI/GFCI works as other schemes globally? Is it totally safe without ground?
More Info
Yes I do have a SAT TV ( sound ) in the loop so to speak. But with EVERYTHING ( 3 power amps; 1 electronic crossover; 1 CD player) all turned OFF the results are the same.
It just seems so much easier to run the preamp ungrounded and have black hole silence like it was with my ungrounded Hafler for the past eons. I'll keep reading your posts on safety and possible fixes. THANKS !
Yes I do have a SAT TV ( sound ) in the loop so to speak. But with EVERYTHING ( 3 power amps; 1 electronic crossover; 1 CD player) all turned OFF the results are the same.
It just seems so much easier to run the preamp ungrounded and have black hole silence like it was with my ungrounded Hafler for the past eons. I'll keep reading your posts on safety and possible fixes. THANKS !
Yes. For example, the ABB GFCI modules (in the breaker box) I installed when upgrading the electrical supply in my apartment don't connect to ground. The main one was three phases hot plus neutral. The bathroom one was one phase hot and one neutral. Further, in that apartment, in Belgium, we actually didn't have neutral; all 220V power was two hot phases. So, in that case, the correct wiring was to jumper the neutral on the main GFCI to one of the hot phases, while the bathroom GFCI gets hot/hot.
I added ground to all the outlets (original wiring was from 1966 and didn't have ground), but ground still had nothing to do with GFCI.
The basic idea is that if you have a device that's leaking current to some non-standard return path, the GFCI will trip. E.g., a hair dryer dropped in the bathtub, where current will be flowing from hot to the ground of the tub/plumbing, rather than coming back through neutral.
Other than potential death to you or some innocent person? No, not really.
NEVER NEVER NEVER operate equipment with the safety ground disconnected. NEVER!
Death ?
I agree safety is an issue....but so too is sound quality. The Hafler company was not sued out of business by a shocked customer...and they had no ground. If I recall properly...The Hafler DH200 power amp manual came with a suggestion to defeat the ground if pick up was an issue. Seems with the ground attached the super high impedance of the FETS would act as an antenna. That was my experience with the Hafler power amp too.
I agree safety is an issue....but so too is sound quality. The Hafler company was not sued out of business by a shocked customer...and they had no ground. If I recall properly...The Hafler DH200 power amp manual came with a suggestion to defeat the ground if pick up was an issue. Seems with the ground attached the super high impedance of the FETS would act as an antenna. That was my experience with the Hafler power amp too.
Devices without a grounded chassis are supposed to be double insulated. All of the commercial devices you are talking about were UL certified, so they got some sort of testing of their voltage isolation strategies. I'm skeptical that you can convert most devices that were designed for grounded chassis to non-grounded without violating safety parameters.
For this diode isolation thing, I'd be interested in seeing a schematic for a commercial, UL rated, amp that uses it. Anyone know one?
Also, the diode isolation eliminating noise implies that the source of ground loop noise is much greater than 1.4V p-p (or 2.8V for wg_ski's setup). Considering that most amps have full output at less than 2V RMS input, that surprises me. (Edit: No - much less than 1.4V... since that would keep it from forward biasing the diodes. Hmmm...)
For this diode isolation thing, I'd be interested in seeing a schematic for a commercial, UL rated, amp that uses it. Anyone know one?
Also, the diode isolation eliminating noise implies that the source of ground loop noise is much greater than 1.4V p-p (or 2.8V for wg_ski's setup). Considering that most amps have full output at less than 2V RMS input, that surprises me. (Edit: No - much less than 1.4V... since that would keep it from forward biasing the diodes. Hmmm...)
Last edited:
A ground loop that induces a couple tens of *microvolts* of hum is annoying. You can get several hundred *millivolts* on a long cable - which is why I don't trust one diode to be enough.
This is fine if you assume that the fault originates inside the trafo or SMPS inside said piece of line level equipment (with a 20ga cord). Murphy's law says it will come from somewhere else, with a lot more oomph behind it.
I find that it depends on what the amp is connected to.
For a floating source I connect the zero volts to earth.
For an earthed source I disconnect the zero volts from earth.
A switch is the best if you use many sources.
Not even in the same league.
My attitude is this: if after being warned by several experienced people that what you're proposing is DEADLY DANGEROUS, you want to proceed, if you fry yourself, well, that's Darwin in action. Problem is that you're just as likely to fry some innocent person, and that seems to concern you not a bit. After all, it's easier to put people at risk of death than to actually fix your grounding problem.
Thread closed. For anyone reading who is not insanely foolish, NEVER TRY TO DEFEAT SAFETY GROUNDS. If you have a grounding issue, fix the grounding issue. DON'T PUT INNOCENT PEOPLE AT RISK.
- Status
- Not open for further replies.