I don't understand the objections against single pole switches. 😕
From a manufacturers point of view, amateurs are not allowed to open the case and poke around the equipment while the mains plug is still inserted.
Service techs have to use their isolation transformer fed workbench. Otherwise they would have no backup by their insurance.
People always hold others responsible for their own errors or foolishness.
I wonder to whom these guys will complaint. 😀
http://www.woltlab.com/volcano/inde...6b59b11b8a9d25ea1dbd91ae4442a156086b&1b2f4c18
regards
From a manufacturers point of view, amateurs are not allowed to open the case and poke around the equipment while the mains plug is still inserted.
Service techs have to use their isolation transformer fed workbench. Otherwise they would have no backup by their insurance.
our mains electricity system is not designed to be idiot proof. It offers reasonable safety against 'normal' levels of abuse.KSTR said:
People always hold others responsible for their own errors or foolishness.
I wonder to whom these guys will complaint. 😀
http://www.woltlab.com/volcano/inde...6b59b11b8a9d25ea1dbd91ae4442a156086b&1b2f4c18
regards
Sure Juergen, this is a valid standpoint. Still I think the cost increase for dual pole vs. single pole is so marginal compared to the effect which is: less zapped people!
- Klaus
- Klaus
Thanks, Demian, for your notes. Pretty much what are my takes on this, also.
FWIW: I now know why individual secondarys/bridges (in power amps, especially) can induce more common mode noise in typical systems, the diodes actually switch the xformer's coupling capacitance on and off all the time! With a continuous galvanic or AC connection to the secondaries (like a center tap at GND) the coupling capacitance (be it to live/neutral or to safety-gnd in case of a shielded primary) is present all the time (lumped analysis, in fact some inbalanced parts are still switched on and off by the diodes), still not nice, but much better. And common mode chokes, at the right points, do help a lot, in both cases. Best: xfomers with very low coupling plus shields.
- Klaus
FWIW: I now know why individual secondarys/bridges (in power amps, especially) can induce more common mode noise in typical systems, the diodes actually switch the xformer's coupling capacitance on and off all the time! With a continuous galvanic or AC connection to the secondaries (like a center tap at GND) the coupling capacitance (be it to live/neutral or to safety-gnd in case of a shielded primary) is present all the time (lumped analysis, in fact some inbalanced parts are still switched on and off by the diodes), still not nice, but much better. And common mode chokes, at the right points, do help a lot, in both cases. Best: xfomers with very low coupling plus shields.
- Klaus
@SY 😀
We have a saying here, among (motor)cyclists:
I had right of way, and St.Peter confirmed!
- Klaus
We have a saying here, among (motor)cyclists:
I had right of way, and St.Peter confirmed!
- Klaus
Wavebourn said:
Yes, although I may have seen you on a USENET group, not sure... but I have not read your resume or bio, so I don't know what you know or have done. 😀
___----------_____________----------------_____
Perhaps someone might comment on what the "two diode trick" is, and provide an example? I'm not sure right now what it is, although I think someone mentioned it here once before??
Also, what is this bit about "active ground reference"?? Could we see an example?
Many readers are following along, and not all (myself included) are up on all these esoteric techniques.
_-_-bear
Bottleneck, eventually, for low power nano transistors. regarding the frequency of the gene.rated noise.
Theory from larger transistors predict wrongly.
Intesresting and good report.
Shows that there are still new problems and benefits to discover.
When electronics goes into a new 'shrinking age' .. to save some power.
John Curl
what do you think of the future nano-technique ...
Maybe not the clever guys hyave the curage to comment
to Jan Didden link
😀 😀
/Lineup 😎 has got one Copy of JC on Audio Matters
Theory from larger transistors predict wrongly.
Intesresting and good report.
Shows that there are still new problems and benefits to discover.
When electronics goes into a new 'shrinking age' .. to save some power.
John Curl
what do you think of the future nano-technique ...
Maybe not the clever guys hyave the curage to comment
to Jan Didden link
😀 😀
/Lineup 😎 has got one Copy of JC on Audio Matters
janneman said:
lineup said:
Preamp grounding
This is what I am using as a grounding layout for single ended, dual mono, preamps for ages now, and never had a shred of a problem.
I realize that this may not be ideal from a RFI perspective, but then I am not in the business of selling equipment. OTOH it always passes the "cell phone test" and, as I already mentioned, allowed HPS 3.1 to have the equivalent of 5nV hum at the input (50uV at the output).
To my experience, switching to a RFI immune layout (with all signal connectors grounded to the case, etc... will never allow the same level of LF harmonic noise (mains hum & harmonics, etc...).
Another interesting observation is the case material. While for power amps steel is a big nono (because of the high currents in the presence of magnetic walls you would be guaranteed a high level of additional 2nd harmonic distortions), for preamps steel is definitely the best. I have never being able to measure any adverse magnetic effects in low signal circuits. So much for "components with non-magnetic terminals" boogeyman stories.
This is what I am using as a grounding layout for single ended, dual mono, preamps for ages now, and never had a shred of a problem.
I realize that this may not be ideal from a RFI perspective, but then I am not in the business of selling equipment. OTOH it always passes the "cell phone test" and, as I already mentioned, allowed HPS 3.1 to have the equivalent of 5nV hum at the input (50uV at the output).
To my experience, switching to a RFI immune layout (with all signal connectors grounded to the case, etc... will never allow the same level of LF harmonic noise (mains hum & harmonics, etc...).
Another interesting observation is the case material. While for power amps steel is a big nono (because of the high currents in the presence of magnetic walls you would be guaranteed a high level of additional 2nd harmonic distortions), for preamps steel is definitely the best. I have never being able to measure any adverse magnetic effects in low signal circuits. So much for "components with non-magnetic terminals" boogeyman stories.
My guess is (Demain needs to confirm) : The Diodes in Syn08's drawing, right?bear said:
- Klaus
bear said:
Active ground, I as know it, is when you return your ground lines to the output of an aux amp that functions as a virtual ground. If you then trace the ground return currents, they end up in the supplies of the aux amp ('dirty supplies') and not in the ground system or supply of the main amp.
Jan Didden
Re: Preamp grounding
Shouldn't D103/D104 be anti-parallel like D101/102?
syn08 said:This is what I am using as a grounding layout for single ended, dual mono, preamps for ages now, and never had a shred of a problem.
Shouldn't D103/D104 be anti-parallel like D101/102?
J2 output cable braid and J3 output cable braid will be connected together on receiver side, will it be a ground loop?
I don't like bridges connected that way. See what Klaus wrote:
I don't like bridges connected that way. See what Klaus wrote:
dimitri said:
That's only the left-right channels grounding connection, it's not a ground loop. Nothing returns through that connection.
I was not expecting you or John to like it, after all it's not an absurdly expensive carving in a solid aluminum billet. Though, it served me well and I'll continue to use it until I'll find a better idea. I'm sick of criticism without offering any concrete alternatives, beyond "i don't like it" or bombshell words. Schematics please.
Re: Preamp grounding
- Klaus
I completely agree, for a design to achieve (sub-)ppm distortion levels I don't think it would be easy to have it optimized for best RFI-perfomance at the same time. Which then depends on how prone to demodulation the circuit is in first place, if it isn't, no problem then.syn08 said:
I think pwr-amps could also be designed to work perfectly with steel chassis... if we can truly find out and seperate any currents in 100% precise "send" and "return" paths which then could be routed as twisted pairs or likewise technique to eliminate field pickup/emission. It's certainly hard to achieve this with PCBs, on short standoffs, parallel to a steel base plate. To use other materials than steel is a wise practial engineering shortcut to the problem.
- Klaus
If find the time, I'll try to illustrate the concept. Currently I'm tied up in an experiment we conduct on a german forum on audibility of group delay distortion (both overall and inter-channel).janneman said:
- Klaus
KSTR - I am interested in the experiment on group delay... link?
Presumably Gurgle Translate will do the trick for me... 😀
Re: the diode trick - what is the mechanism by which this is preferable to the resistor to chassis ground (assuming you do that)? And how so does it improve matters?
Re: the active ground - are there any schematic examples of how this works? How large does the active ground have to be compared to the (power level) circuit that it "grounds"? And, do not the grounds of the "active ground" and the main power ground have to be connected for there to be a "return path"? Or are we going to return power to the B+/- rails?
Is this a sucessful method - if so, why is it not typically used more often?
And, if it provides a "high impedance" (infinite?) ground point why is this different than a (finite impedance) choke in the ground path?? Or why not take the AC ground as the point between two caps?
_-_-bear
Presumably Gurgle Translate will do the trick for me... 😀
Re: the diode trick - what is the mechanism by which this is preferable to the resistor to chassis ground (assuming you do that)? And how so does it improve matters?
Re: the active ground - are there any schematic examples of how this works? How large does the active ground have to be compared to the (power level) circuit that it "grounds"? And, do not the grounds of the "active ground" and the main power ground have to be connected for there to be a "return path"? Or are we going to return power to the B+/- rails?
Is this a sucessful method - if so, why is it not typically used more often?
And, if it provides a "high impedance" (infinite?) ground point why is this different than a (finite impedance) choke in the ground path?? Or why not take the AC ground as the point between two caps?
_-_-bear
Hi Ovidiu,
thanks for the schematic!
Always interesting to see well working wiring schemes.
So if I see correctly you connect the enclosure to safety ground and disconnect (from chassis) power and signal ground using a bridge. Ever used a thermistor in this place?
Interestingly you connect left and right at their inputs; that reminds me a little bit of D. Selfs scheme, where the input grounds meet also at their chassis input points (they are, however, also connected to the enclosure at this point in contrast to your scheme).
You certainly know also Self's way of grounding, what do you think about that?
I can post a scheme if needed, but to summarize, he connects the input signal grounds to chassis at their entry-points, and connects these to a sequence of internal star grounds (where power and whatever meet). The chassis is connected to safety ground at a separate point. No bridges or thermistors.
His argument is that ground loops can form only from signal ground at the chassis entrypoint to the safety ground and cannot interact with the internal signal ground.
Have fun, Hannes
thanks for the schematic!
Always interesting to see well working wiring schemes.
So if I see correctly you connect the enclosure to safety ground and disconnect (from chassis) power and signal ground using a bridge. Ever used a thermistor in this place?
Interestingly you connect left and right at their inputs; that reminds me a little bit of D. Selfs scheme, where the input grounds meet also at their chassis input points (they are, however, also connected to the enclosure at this point in contrast to your scheme).
You certainly know also Self's way of grounding, what do you think about that?
I can post a scheme if needed, but to summarize, he connects the input signal grounds to chassis at their entry-points, and connects these to a sequence of internal star grounds (where power and whatever meet). The chassis is connected to safety ground at a separate point. No bridges or thermistors.
His argument is that ground loops can form only from signal ground at the chassis entrypoint to the safety ground and cannot interact with the internal signal ground.
Have fun, Hannes
bear said:
Bear,
Don't have a circuit right now, but just picture a gain-of-one opamp circuit with the input being a clean ref ground. The output is the virtual ground for the main system ground returns - a zero volt, (ideally) zero impedance point. It may not be that ideal, BUT the return currents 'disappear' into that opamp output and do no longer give rise to ground currents in the amp system that might lead to contamination. Instead, those ground currents disappear into the opamp output and end up getting back to the opamp supplies - and these can be separate - must be separate - from the main amp supplies.
There's art least one very successfull outfit that uses this technique to get -160dB S/N in low level circuits with smps supplies - Funk in Germany.
http://www.funk-tonstudiotechnik.de/ - sorry, only in German.
Jan Didden
- Status
- Not open for further replies.