• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Chassis Ground / Signal Ground / Ground Loops

For my '61 Fisher, I replaced the power cord and kept it 2-prong, wired hot to the switch and replaced the death cap with a modern XY one. I don't think there's a need to ground the chassis and create new issues.

For my R8, I did the Skunkie mods which introduced a star ground and it's dead quiet. I also use a star ground on my Pass DIY builds. Use nice heavy wire all to a central point.
 
If something upstream had a "death cap", in theory, the described behavior could happen.

I have also heard of a situation, where stray capacitance inside a power transformer, somehow managed to induce AC voltage on an ungrounded preamp chassis.
I don't see how either of those scenarios would be applicable to a modern build with a three-wire AC cord and a grounded chassis.

capacitive coupling is mostly what I was thinking, but failing insulation could also do it.
Can you be more specific? As far as I know, the only component that connects to both the internal circuitry and the AC line is the power transformer.

Just to be clear. I only asked this question because it was stated in another thread that the loop breaker has a secondary benefit for safety. So far, I can't conceive of any real-world scenario where that would be the case.

Jack
 
It's enough to break the ground loop and still provide a measure of safety beyond simply floating signal.
Just to be clear. I only asked this question because it was stated in another thread that the loop breaker has a secondary benefit for safety. So far, I can't conceive of any real-world scenario where that would be the case.
Then it doesn't?

I know in my case, if B+ were to somehow short to the ground plane on my mono block's VTA driver, it would pretty well light up my RCA input's negative terminal and also the RCA connector shell. Speaker common too, as it's also connected to the ground plane on the VTA driver right next to the feedback connection. The VTA driver ground connects back to the first cap in my cap bank, same spot the PT center tap is connected too as well as all other signal path grounds. Power tube cathodes, ground reference for bias test point, filament center tap.

I'm not bothered that I have a path to chassis ground for my signal ground consisting of a few dollars in parts. What could possible be wrong with that? Maybe I need to wire it differently? The amps have never been as quiet as they are now, A bit of tube his and that's it.

VTA.jpg
 
I know in my case, if B+ were to somehow short to the ground plane on my mono block's VTA driver, it would pretty well light up my RCA input's negative terminal and also the RCA connector shell.
A short from B+ to signal ground (which is also power supply ground) should blow the fuse. And that should happen regardless of the presence of the loop breaker.

Again, I'm only concerned with the issue of safety that you mentioned, not noise. Based on all the responses so far (including yours), I think maybe you misspoke.

Jack
 
I should take it out then? You're fine with trusting a fuse will open? I've seen a GC amp, (very high current mag amp) connection in an MRI scanner turn cherry red because it was not torqued to spec and it didn't trip the breaker. I know why it didn't trip the breaker, but...

Does it make it unsafe?

If it makes you happy, I'll take it back.

What did you win?
 
Can you be more specific? As far as I know, the only component that connects to both the internal circuitry and the AC line is the power transformer.
If you have primary to secondary leakage or even a dead short from primary to secondary you'd have mains current of some amount at your signal ground. If there was no connection from signal ground to chassis ground, now your input jacks have a potential difference and there might be a shock hazard there. Having the signal ground end up at the chassis keeps that difference in potential from existing, or at least keeps it to such a low level as to not be a concern.

Also consider your sources. I've never seen a turntable that had a 3 wire cord, and mine are all old and use motors that are just 120vac. A failure in one would put the chassis live to the mains and thats going to end up right at the amp.

a short from mains to signal ground wouldn't neccesarily blow the fuse. Actually I doubt it would if there was no connection from signal ground to mains ground. Its the equivalent of a paperclip stuck in an outlet. Dangerous, but with no path to ground there is no current flow so its not going to kick a breaker or blow a fuse.
 
I'm only concerned with the issue of safety that you mentioned, not noise. Based on all the responses so far (including yours), I think maybe you misspoke.
I think you're correct - a grounded chassis and floating circuit is perfectly safe, and a grounded chassis plus loop-breaker is no safer (but it does add shielding from RF interference).
Btw, did you read the link in post #7?
 
I should take it out then? You're fine with trusting a fuse will open? I've seen a GC amp, (very high current mag amp) connection in an MRI scanner turn cherry red because it was not torqued to spec and it didn't trip the breaker. I know why it didn't trip the breaker, but...

Does it make it unsafe?

If it makes you happy, I'll take it back.

What did you win?
It's not a contest. I wanted to know why someone would claim that it's safer, maybe something I hadn't thought of. And yes, I trust a fuse to open if the B+ supply shorts to ground.

Jack
 
I think you're correct - a grounded chassis and floating circuit is perfectly safe, and a grounded chassis plus loop-breaker is no safer (but it does add shielding from RF interference).
Btw, did you read the link in post #7?
I admit to only scanning it quickly. Much of the material relates to noise reduction, and I didn't see anything really new regarding safety. I will give it a more serious look though when time permits.

Jack
 
Last edited:
If the circuit is totally isolated from chassis, there is no shock hazard if B+ shorts to chassis (B- would go to the negative equivalent). But it can't be totally isolated, would be a noise nightmare. Single-ended inputs are referenced to B-, differential inputs have common-mode limits. Transformer isolation of inputs and outputs could work, feedback local or otherwise isolated. Common mode noise can only be filtered by components connected to ground, affecting the isolation. Not practical in most cases. So circuit ground must be connected to chassis ground with a low impedance. A 10 Ohm resistor, 0.1 uF capacitor and back-to-back diodes, all in parallel can do that, though a single-point direct connection may do as well.

Our AC power lines have the neutral grounded for a very good reason - lightning and similarly, power cross - the chance of a high-voltage line dropping onto a lower-voltage line. Grounding one conductor shunts most of the common-mode voltage. Some special cases do use isolated power - military vessels do, so that a ground fault won't shut everything down. Some industrial process equipment - imagine losing power to a furnace full of molten glass or aluminum or a plastic molder - much better to have time to detect the fault and have the time to shut it down properly. These would run on their own dedicated isolated transformer.
 
If there is no connection between B- / signal ground and chassis/earth ground, current isn't going to flow B+ to chassis ground and the fuse will not blow.

The earlier reference was to a short from B+ to power supply ground (to the "ground plane," or signal ground). That will blow the fuse:
I know in my case, if B+ were to somehow short to the ground plane on my mono block's VTA driver, it would pretty well light up my RCA input's negative terminal and also the RCA connector shell.

You're correct that a short from B+ to the earthed chassis won't necessarily blow a fuse, but it also doesn't present a safety risk. Maintining a safe external environment is the reason for connecting the chassis to earth ground.

Jack
 
Last edited:
If the circuit is totally isolated from chassis, there is no shock hazard if B+ shorts to chassis (B- would go to the negative equivalent). But it can't be totally isolated, would be a noise nightmare.
Fully isolating the active circuitry from the earthed enclosure does not induce noise. The enclosure simply acts as an earthed shield around the circuitry. It's only when this configuration is mixed with units where signal ground and chassis are common that noise sometimes occurs due to ground loops.

Jack
 
Can't be totally isolated from ground unless it's running on batteries. Capacitve coupling to the power line will occur, through the chassis and elsewhere..

The map is not the territory.
The schematic is not the circuit.
 
Can't be totally isolated from ground unless it's running on batteries. Capacitve coupling to the power line will occur, through the chassis and elsewhere..
I believe we are now discussing peripheral concepts that have no impact on safety if the chassis is earthed.

Jack
 
Last edited:
It would be very much appreciated if we could focus on the topic of safety that prompted this thread. Someone please start a separate discussion if you want to discuss noise or noise vs. safety or some such. My system is isolated, earthed, and dead quiet, so noise is not an issue I'm curious about or care to discuss here. :)

Jack
 
It would be very much appreciated if we could focus on the topic of safety that prompted this thread. Someone please start a separate discussion if you want to discuss noise or noise vs. safety or some such. My system is isolated, earthed, and dead quiet, so noise is not an issue I'm curious about or care to discuss here. :)

Jack
Sorry, posts deleted!

:thumbsup:
 
@TriodeLuvr What about what @gadget73 said in post #27? I could be wrong, but what if the power transformer primary winding shorted to the part of the secondary such that the secondary/signal circuit ground was floating on top of some portion of the mains voltage. That would be a potential shock hazard at the RCA shell?
 
I learned the hard way about the ground connection on an AM tube radio with a non- polarized plug. It seemed logical to me to ground it, see what happens. After all, why else would it say ground.

The house was built in 1952. It looked like the neutral wiring was grounded, but there where no grounded outlets in the house.

Anyways, I connected the ground to some furnace ducting on a gravity furnace, where the only source of electricity was from one of those thermal mili- volt gas valve.as soon as I turned it on it nearly exploded looking more like someone arc welding.
 
Back
Top Bottom