Ah, seems like I was thinking analytical listening and critical listening were synonymous, and what's more, critical listening was not what I thought it was. Haha, so the confusion created a perfect storm of misunderstanding on my part. Thankfully I was right about my art analogy and it being analytical, but critical listening (including listening for distortions) isn't going to ruin my pleasure, what a relief, thanks. 😀
T, I feel like doing the same thing, but I know that is what my 'critics' want. Why the moderators allow this is certainly beyond me! Hang in there!
I'm surprised the mods didn't at least comment on the "hot neutral" version of the power-line filter.
jneutron, I'm wondering about estimating coupling between loops on a PCB.
If you want to estimate the coupling of an irregularly shaped loop to another irregularly shaped loop, can you decompose the loops into say a combination of squares and triangles, and calculate the combined coupling of the smaller shapes?
For instance to simulate the coupling of a loop to a signal pair, would it be valid to decompose the signal pair into a number of squares and combine the coupling calculated to each square?
If you want to estimate the coupling of an irregularly shaped loop to another irregularly shaped loop, can you decompose the loops into say a combination of squares and triangles, and calculate the combined coupling of the smaller shapes?
For instance to simulate the coupling of a loop to a signal pair, would it be valid to decompose the signal pair into a number of squares and combine the coupling calculated to each square?
Last edited:
You most certainly can. Or as I normally do, try to figure out the worst case scenario. Experience tends to show that whatever worst case scenario you come up with....its worse..😉
In general, it's usually not difficult to measure coupling. Couple of amperes in the aggressor, measure the victim. I've setup my audio amp to drive the aggressor, usually with a 4 ohm load in series, and use a scope to see the resultant.
If you are careful in setup so that the measurement circuit is not also coupling, you can get decent results. It is very important to sweep frequencies. I like to do the 2/5/10 thing...20 hz, 50 hz, 100 hz, 200hz etc, it plots reall well on log.
Jn
Edit: for the diagrams I posted showing ground loops, I always drive the aggressor with a wide sweep to view the coupling. It's always interesting what you see. Oh, always watch for conductive objects during these tests. They tend to really muck up inductance and coupling because of eddies and flux exclusion. Very important that you measure the actual site without confounders.
In general, it's usually not difficult to measure coupling. Couple of amperes in the aggressor, measure the victim. I've setup my audio amp to drive the aggressor, usually with a 4 ohm load in series, and use a scope to see the resultant.
If you are careful in setup so that the measurement circuit is not also coupling, you can get decent results. It is very important to sweep frequencies. I like to do the 2/5/10 thing...20 hz, 50 hz, 100 hz, 200hz etc, it plots reall well on log.
Jn
Edit: for the diagrams I posted showing ground loops, I always drive the aggressor with a wide sweep to view the coupling. It's always interesting what you see. Oh, always watch for conductive objects during these tests. They tend to really muck up inductance and coupling because of eddies and flux exclusion. Very important that you measure the actual site without confounders.
Last edited:
In your diagrams you should the power cord running next to the signal cable. Wouldn't there be significant magnetic and/or E-field coupling into the signal line depending on the power cord and cable geometry?
You are right, there can be significant coupling depending on geometries.
I relied on the fact that the power cord and my line cables had different twist pitches. The net result was on average, zero coupling. This is how cat 5e cable works, they have four different twist pitches by specification.
By twisting a 100 foot plus single ended signal cable around a line cord feeding my mix rack, I eliminated any coupling from either the line cord, or from external sources. By twisting the two together, I averaged out external influence..
The simplest scenario for that solution is a turntable to preamp connection. If you twist the table ground wire tightly around the two rca's to the pre in, you remove all external influence at cart level.
Jn
I relied on the fact that the power cord and my line cables had different twist pitches. The net result was on average, zero coupling. This is how cat 5e cable works, they have four different twist pitches by specification.
By twisting a 100 foot plus single ended signal cable around a line cord feeding my mix rack, I eliminated any coupling from either the line cord, or from external sources. By twisting the two together, I averaged out external influence..
The simplest scenario for that solution is a turntable to preamp connection. If you twist the table ground wire tightly around the two rca's to the pre in, you remove all external influence at cart level.
Jn
Did it sound good at line level?
I do that with my TT because of convenience. There is no audible difference.
I do that with my TT because of convenience. There is no audible difference.
What if one cable wraps around the other more? Wouldn't that situation be sort of like a current transformer where you have a wire passing through a coil?
Yes, the skill in mastering is in achieving a release that translates well to pretty much any PB system.
The first thing for the mastering engineer to get right is his mastering acoustic level.
Once the mix sounds 'right' at this 'correct' acoustic level (SPL), then it will also sound 'right' at pretty much any SPL on any system.....turn the sound down and the mix does not fall apart, turn the sound up and the track/album just gets more fun, turn it way up and it becomes spectacular.
The albums and tracks that have stood the test of time satisfy the above criteria and this is part of what has made them (unconsciously) popular.
When a mix 'ain't right' it gets worse the louder it is played, and when turned down improves, with mute being the best.
Mids and highs are usually pretty right, the problems are mostly to do with the bass end.
I find that some tracks/albums require turning SW level up or down (usually only a couple or few DB's) and then the whole mix sounds right as it can be, and counter intuitively perhaps, graininess/harshness in vocals/mids/highs disappears this also......somebody here (DestroyerOS ?) mentioned this recently and yes I also have always found this to be true.
Yeah, but.
Those defects particularly distortion defects once heard on the good system will still be audible on a lesser system, usually more noticeable thereafter.
The key to a really good systems is resolution....distortions serve to exaggerate defects in recordings
Yes.
One that works very well for me and is economical and easily AB comparable is flocked air mattress placed directly behind and between the speakers.
IME the first key to achieving good sound is to get rid of cabinet early side and early rear emissions.
To deal with loudspeaker front panel early reflections acoustic pinboard is economical and works very well particularly when surrounding the tweeter Duntech style.
When I get around to it I will try gluing carpet square inside my cabinets...this will increase mass, increase internal acoustic damping and cabinet mechanical damping and should work very well IMHO.
Dan.
Hi Scott,
We try our best to catch things like that, but we are not charged with any sort of technical review. In my experience our members are normally good at catching those things, and sometimes they will report them as dangerous. At that point we will react.
John,
You are respected. Some of your comments aren't. Please don't confuse the two.
-Chris
'cause none of us noticed it.
We try our best to catch things like that, but we are not charged with any sort of technical review. In my experience our members are normally good at catching those things, and sometimes they will report them as dangerous. At that point we will react.
John,
If you can dish it out or handle yourself well, we generally try to use a light finger moderating this thread.
You generally run into trouble when you post certain things that you know for sure will incite some members to remark. The same thing happens to other people who post similar things. This site is mostly self moderated by it's membership and I would rather not interfere in the natural processes of a healthy technical forum. Other moderators may have their own judgment in this respect. I speak for myself on this.
You are respected. Some of your comments aren't. Please don't confuse the two.
-Chris
Ok, so my question on the ground loops.
Attached is one of the diagrams linked earlier from the 2013 discussion on loop areas. The bit that raised my eyebrow on this was 'can be differential'. This just got a bit heavy as it is saying that AES48 is not optimal and there is a better way than pin 1 to chassis, only to chassis, do not pass go, do not collect $200. For a diff and balanced input into an INA this is causing some head hurting.
I then realised that my soon to be (at least as soon as they are boxed) amplifiers use a very similar setup. In this case because it uses the Whitlock THAT input receivers with the full data sheet EMC diodes, so the shield link was necessary.
It still feels wrong though to increase the impedance from shield to ground in a balanced connection. So why is that good and how does that scale when you have >1 input and an output?
Attached is one of the diagrams linked earlier from the 2013 discussion on loop areas. The bit that raised my eyebrow on this was 'can be differential'. This just got a bit heavy as it is saying that AES48 is not optimal and there is a better way than pin 1 to chassis, only to chassis, do not pass go, do not collect $200. For a diff and balanced input into an INA this is causing some head hurting.
I then realised that my soon to be (at least as soon as they are boxed) amplifiers use a very similar setup. In this case because it uses the Whitlock THAT input receivers with the full data sheet EMC diodes, so the shield link was necessary.
It still feels wrong though to increase the impedance from shield to ground in a balanced connection. So why is that good and how does that scale when you have >1 input and an output?
All of them that convert to balanced power in the "Felicia" back to back transformer setup. They need a GFCI on them as per code (and a wise code in this situation).
Ok, thanks.
GFCI is required for grounded center tap output, what about floating output ?.
Dan.
GFCI is required for grounded center tap output, what about floating output ?.
Dan.
Safety ground is bonded in both situations, just to be clear, bypassing the transformer.
Single ended transformers do not require GFCI.
Single ended transformers do not require GFCI.
- Status
- Not open for further replies.