| Tyimo |
Hi!
After a serious grunding noise problem with a tubepreamp and the SEWA mosfet follower amp I was re-reading all my articles about earthing and I found something mistakable thing with the diode bridge coupling.
Mr. Pass wrote in the ZV4 docs:
| quote: | Note in Figure 6 that we have chosen to isolate
the two channels through a rectifier bridge to ground, with each
channel’s ground appearing on one of the AC legs of the bridge. |
As it is in the attached image.
Mr. Rod Elliot wrote:
| quote: | | Note the way the bridge is wired, with the two AC terminals shorted, and the two DC terminals shorted. Other connection possibilities are dangerous, and must be avoided. |
as it is in the next attached image.
Now I have only one question:
How should I connect the diode bridge to the circuit grund?
DC legs to Earth - AC legs to the circuit? Nelson Pass
AC legs to Earth - DC legs to the circuit? Rod Elliot
So, Who has true?????????????????????????????
Greets:
Tyimo |
|
|
| Tyimo |
| Here is Rod Elliots version: |
|
|
| Babowana |
| quote: | Originally posted by Tyimo
DC legs to Earth - AC legs to the circuit? Nelson Pass
AC legs to Earth - DC legs to the circuit? Rod Elliot
So, Who has true?????????????????????????????
|
Both.
:darkside: |
|
|
| Tyimo |
Hi Babowana!
Than why Rod Elliot wrote:
| quote: | | Other connection possibilities are dangerous, and must be avoided. |
?????
Tyimo |
|
|
| Babowana |
| quote: | Originally posted by Tyimo
Than why Rod Elliot wrote:
?????
|
When he wrote that, he was unsure of himself . . . ?
I hope the attached sketches show the understanding.
The horizontal flip is always direction-free symmetry.
Either end of (+/-) or (~/~) could go the gnd.
:darkside: |
|
|
| Nixie |
| For a ground loop breaker, besides the diodes you should put a 5 to 20 ohm resistor in parallel, as well as a small cap as HF bypass (0.1-1 uf). Make sure the diodes are rated at sufficient current so that in a fault situation the fuse on the hot mains line blows! |
|
|
| Babowana |
| quote: | Originally posted by Nixie
For a ground loop breaker, besides the diodes you should put a 5 to 20 ohm resistor in parallel, as well as a small cap as HF bypass (0.1-1 uf).
|
Why?
:darkside: |
|
|
| Nixie |
| In most ground loop breakers I've seen, the diodes are there as protection in case of fault where the resistor would just burn out. It's the resistor that's doing the ground lifting--you just need sufficiently different resistances on each of the legs of the ground loop. If you just have diodes, the circuit ground is floating within the one diode drop from earth, with no connection to earth; with a low value resistor, I find it still breaks ground loops, but the circuit tends to stay closer to eart. The bypass capacitor allows earth to sink HF interference. With the scope I see less RF on the circuit ground when the cap is added. |
|
|
| Tyimo |
Hi Nixie!
| quote: | | For a ground loop breaker, besides the diodes you should put a 5 to 20 ohm resistor in parallel, as well as a small cap as HF bypass (0.1-1 uf). Make sure the diodes are rated at sufficient current so that in a fault situation the fuse on the hot mains line blows! |
Thanks! Yes, I know. Rod Elliot wrote the same. Most of time a simple diode bridge (35A) or a thermistor (10R/5W) was enough for me.
Greets:
Tyimo |
|
|
| Nixie |
| The diodes are there for safety reasons. Do not omit them, and again, make sure their current rating exceeds the main fuse sufficiently. |
|
|
| AndrewT |
Hi,
both ESP and Pass circuit are correct. But they are different and for different purposes.
The Pass circuit is a for a 2channel amplifier and isolates the two channels from each other thus breaking the ground loop and avoiding hum. The disadvantage is the 35A rating of the safety connection from audio ground to safety earth.
The ESP circuit is for a single channel amplifier and has the advantage of 35*2 =70A rating of the safety connection from audio ground to safety earth.
The additional capacitor and resistor can be added to either the ESP or Pass version of the circuit. They are effectively optional and may work better for some amplifiers/user's ancilliaries/house earthing than others. I tend to use the dual Power diode //cap//Power resistor for EACH channel. i.e. the ESP version. The Pass version is basically a money saving exercise. |
|
|
| Babowana |
| quote: | Originally posted by AndrewT
The Pass version is basically a money saving exercise.
|
I don't 100% agree.
Otherwise, he is no different from my another side, isn't he.
:darkside: |
|
|
| Tyimo |
Thanks Nixie and AndrewT!
Now it is clear.
Greets:
Tyimo |
|
|
| AndrewT |
| quote: | Originally posted by Babowana
I don't 100% agree.
Otherwise, he is no different from my another side, isn't he. |
which part of omitting a full dual diode with //resistor//cap for each channel do you not 100% agree is money saving? |
|
|
| Babowana |
| quote: | Originally posted by AndrewT
The additional capacitor and resistor can be added to either the ESP or Pass version of the circuit. They are effectively optional and may work better for some amplifiers/user's ancilliaries/house earthing than others. I tend to use the dual Power diode //cap//Power resistor for EACH channel. i.e. the ESP version. The Pass version is basically a money saving exercise.
|
The paralleled diodes in the opposite direction each other are to break current between the signal star ground and the protective earth (connected to the chassis) – to break possible ground loop. Then, the signal star ground would not see the small signal noises coming through the ground loop. But, this works only when the potential difference between the signal star ground and the protective earth is less than about 650mV. When the potential is greater than this value, the diodes are shorted for current flow with no resistance, and the current will blow off the fuse installed at the entrance of the live (active) line or switch off the house wall breaker – for safety.
Enough is enough.
Your comments are just words to talk words, saying nothing about clear technical explanation. And, the last sentence is just a kinda bad mouthing.
:darkside: |
|
|
| AndrewT |
Hi,
I see a far more onerous demand from those inverse parallel diodes.
The fault current that can pass prior to blowing a fuse is enormous.
A 25A or 35A power diode has a peak current rating in the hundreds of Amperes. A dual 25A or dual 35A has roughly twice as much capacity at peak rating. This should help the diodes to survive the many kA that can pass to earth in the few uS that it takes for the fuse to rupture and the arc to extinguish.
A single diode without a parallel cap and/or parallel power resistor for each channel is a cheaper option than the dual diode//Power resistor//cap. I believe that if us amateurs cannot extensively test what we assemble then when it comes to safety we should not take the risk of cutting corners nor costs.
The Pass system as shown is the cut price system. Whether it is adequate is arguable. The fact is,| quote: | | The Pass version is basically a money saving exercise. |
|
|
|
| Babowana |
| quote: | Originally posted by AndrewT
I believe that if us amateurs cannot extensively test what we assemble then when it comes to safety we should not take the risk of cutting corners nor costs.
|
US amatures . . . LOL ~
<AndrewT>
The ESP circuit is for a single channel amplifier and has the advantage of 35*2 =70A rating of the safety connection from audio ground to safety earth.
A 25A or 35A power diode has a peak current rating in the hundreds of Amperes. A dual 25A or dual 35A has roughly twice as much capacity at peak rating. This should help the diodes to survive the many kA that can pass to earth in the few uS that it takes for the fuse to rupture and the arc to extinguish.
<AndrewT>
Two words from one mouth! LOL ~
I hope you are able to look into the attached concept.
:darkside: |
|
|
| stefanobilliani |
Hello All.
In Zen V4 for example , where is this simple diode circuit best applyed ,
at the output , or close to the main dc filtering caps ?
What do you guys think ? |
|
|
| Nixie |
| Same place you would normally earth your circuit. |
|
|
| stefanobilliani |
| quote: | Originally posted by Nixie
Same place you would normally earth your circuit. |
Nice , considering most of the consumer electronics doesnt have an earth connection at all .
From the Pass V4 article it seems to me clear enough that a good role is to connect the earth bridge near the voltage rectifier bridge .
Usually, if I consider a phono preamp it is common sense to tie the 2 grounds of both channels together at the output and then from there to the chassis . ( they use to do that also in cd players ) .
The difference is in the potential of the chassis . Without an earth connection it can be at quite high-ac values .
:att'n: |
|
|
| Babowana |
I've re-drawn the above sketch in the post #17.
I think that the inserting of the ground loop breaker between the
PE on the chassis and the circuit ground is just a connection --
with respect to both the ground loop breaking and fault circuit
safety.
The important thing is whether the circuit ground is located on
the right position without hum and/or buzz. If there is hum
and/or buzz, the position of the circuit ground might be wrong.
Or, re-layout of the signal ground loop should be considered.
Or, other altenative solution might be introduced as shown in
Zen V4 with the 22ohms attached between the input ground and
the chassis.
Any comment?
:darkside: |
|
|
| AndrewT |
Hi,
your grounding and safety earth look OK.
The thermistor and/or power diode and RF capacitor can all be fitted in parallel rather than exclusively one or other.
I think any voltage diode will do, 50V or 100V will work just as well as 1000V. |
|
|
| Babowana |
| quote: | Originally posted by AndrewT
your grounding and safety earth look OK.
The thermistor and/or power diode and RF capacitor can all be fitted in parallel rather than exclusively one or other.
I think any voltage diode will do, 50V or 100V will work just as well as 1000V.
|
AndrewT,
I didn't ask your approval for the already proven standard.
If you don't have any real technical contribution, just sit back
on the highland quietly.
And, don't try to be a big man, just putting your name on the
top of experienced guys. I hope you will kindly undrstand
what I mean!
:darkside: |
|
|
| Magura |
| quote: | Originally posted by AndrewT
Hi,
your grounding and safety earth look OK.
The thermistor and/or power diode and RF capacitor can all be fitted in parallel rather than exclusively one or other.
I think any voltage diode will do, 50V or 100V will work just as well as 1000V. |
What's the deal about the RF cap?
Magura :) |
|
|
| Nixie |
| I hook it up near the transformer center tap, or if there isn't one, at the rectifier or first filter capacitor grounds. Don't just use the diodes; add the resistor and capacitor as described above. |
|
|
| Babowana |
| quote: | Originally posted by Nixie
Don't just use the diodes; add the resistor and capacitor as described above.
|
If you want to say, "don't do this or that,"
it should be backed up by reasons.
Otherwise, I see it as an ignorant BS.
Thanks.
:darkside: |
|
|
| Nixie |
| AndrewT and myself have already backed it up. Of course, we cannot compensate for your lack of reading comprehension. |
|
|
| Magura |
| quote: | Originally posted by Nixie
AndrewT and myself have already backed it up. Of course, we cannot compensate for your lack of reading comprehension. |
Hmm, what could such a circuit be called (I need something to feed google) ?
Cheers
Magura :) |
|
|
| Babowana |
| quote: | Originally posted by Nixie
AndrewT and myself have already backed it up. Of course, we cannot compensate for your lack of reading comprehension.
|
What about your lack of comprehension about tech?
LOL ~ ^^
:darkside: |
|
|
| SY |
| :cop: C'mon, guys, keep this to the topic, OK? |
|
|
| GRollins |
While I can arguably see the resistor as part of the safety scheme, I don't see that the cap fits in as a safety item; it's just there to drop RF past the 10 Ohms of resistance and/or the diode drop represented by the bridge. This is a sonic issue, not a safety one.
If you assume that there's going to be sufficient RF voltage present to be physically harmful, you need to get the amp on the bench and determine why it's oscillating rather than simply try to dump the RF to ground.
Face it, the rails on a fairly large amp will run towards 50-75V. Nervous nannies notwithstanding, this is just not an instant-death scenario. I stick my hands in that sort of voltage all the time. 50Vdc you don't even feel. 75Vdc will give you a mild tingling sensation. Above 100Vdc, I turn the thing off before sticking my hands in, as it's enough to be unpleasant.
Derate those figures by half if it's AC you've got your hands in, but that's another question entirely.
If you're really all that concerned with safety--as opposed to sound quality--then wire signal ground straight to chassis ground and be done with it. Ninety jillion pieces of audio equipment have done this over the years and it works okay.
Note that a 10 Ohm 5W resistor will reach its rated wattage when presented with 7.07V. In a worst-case fault scenario, meaning rail voltage is running rampant in your amp, it's toast anyway. It's just along for the ride.
Or you could side step all this fussing and use a varistor to ground...I'm not so sure I'd want to be polluting my ground with rectified RF hash, anyway.
Grey |
|
|
| Babowana |
OK, SY. Your black hairs and long nose looking fresh.
I understand that the small cap could be a RF path to the chassis
from the lifeted circuit common ground point. It could be an
effective means when there is a RF signal at the lifted circuit
ground point. But, I would not force to always have it. It could
be a path to the opposite way.
:darkside: |
|
|
| stefanobilliani |
| quote: | Originally posted by GRollins
...I'm not so sure I'd want to be polluting my ground with rectified RF hash, anyway.
Grey |
In some service manuals I have seen the use of 2 AC rail with center tap( signal ground ) to produce just the positive DC rail , through 2 diodes .
Is that what you mean Grey ?
I tried it for a Zen amp and indeed the sound quality was different ;
there was little more hum however , but the test wasnt probably under an optimal layout or enough filtering . |
|
|
| GRollins |
Assume first that you have RF in the audio portion of your circuit--not an outlandish assumption.
--If circuit ground = chassis ground = earth ground, then it goes to ground and all is well
--If you have a resistor between the grounds, then the presumption is that it has "flat" response out to RF and again the RF goes to ground.
--If you have a bridge to ground, then the RF gets rectified, which gets complicated because the rectification process can itself yield RF (you did remember to use a fast/soft bridge, didn't you?). Depending on whether the RF is AM or FM you branch off into other concerns.
There's also that pesky diode drop to deal with if you're trying to get rid of DC.
--A cap between the grounds does nothing whatsoever for DC. Whether it does anything for RF depends on the value and construction of the cap. Too small a value and you miss the lower frequencies you're trying to get rid of. Use something like an electrolytic and it ceases to be a cap at all before it reaches RF.
So you see ESP combining a bridge with a cap and a resistor in an effort to cover the weaknesses of each individually. To my way of thinking, this verges on violating the KISS principle. I'd rather just use a resistor (or varistor) and be done with it.
Grey |
|
|
| richie00boy |
I can't believe such a simple circuit is causing so much confusion!
The resistor is to insert higher resistance between the mains earth and signal ground to limit the amplitude of earth loop currents.
The cap is to overcome the inductance of the resistor at higher frequencies, shunting RF to earth rather than letting it rampage inside the equipment, i.e. allowing the system to provide effective RF screening.
The diodes play no part in the earth breaker, they merely provide a safety backup in the event of high current needing to be dumped to earth, as would occur with a catastrophic fault. The resistor and cap would simply pop and let you die.
If you use only a diode to attempt to lift ground it is basically preventing any screening effect of the case and mains earth until a small voltage is present between the two - in which case you have a pretty severe fault/issue elsewhere in the system anyway. |
|
|
| Nixie |
| quote: | Originally posted by richie00boy
If you use only a diode to attempt to lift ground it is basically preventing any screening effect of the case and mains earth until a small voltage is present between the two - in which case you have a pretty severe fault/issue elsewhere in the system anyway. |
Don't bother. Apparently Babowana has put his fingers in his ears and is going singing "lalalalala I can't hear yoooouu!" |
|
|
| GRollins |
Granted, I'm not a/the moderator anymore, but can you kinda tone down the attitude?
Sometimes explanations don't work the first time around and a reworded version is required.
Given Babowana's personality and history of posting, I have no reason to believe that insults are called for. I'm sure he isn't trying to cause problems.
Grey |
|
|
| Babowana |
| quote: | Originally posted by Nixie
Don't bother. Apparently Babowana has put his fingers in his ears and is going singing "lalalalala I can't hear yoooouu!"
|
I think you need to go and ask milk further,
before your bones are grown up in terrible way.
:darkside: |
|
|
| Nixie |
?
Try again, in English this time. |
|
|
| Babowana |
| quote: | Originally posted by Nixie
Try again, in English this time.
|
Okay, your English is good.
You are better than me.
But, don't think that you know where my fingers are put in.
Enjoy!
:darkside: |
|
|
| GRollins |
Babowana, you're not helping any.
Grey |
|
|
| Babowana |
| Understood, Grey. Thanks. |
|
|
|
