If the shield is carrying the return current it will cancel. But the sectional current through the cable must be exactly zero.
A shield that is not carrying the return current will not zero out the external field.
At that, the twisted pair will cancel any rail to rail currents, which should not be there anyway other than bias, the shield to each rail will cancel most out, but each wire in a twisted pair in a shield will not have an exact cancel outside the shield as the centroids do not exactly match.
A triax cable would however. core as one rail, first shield as other rail, outer shield as ground and current return. I've made triax for audio supply runs before, quite easy.
jn
A shield that is not carrying the return current will not zero out the external field.
At that, the twisted pair will cancel any rail to rail currents, which should not be there anyway other than bias, the shield to each rail will cancel most out, but each wire in a twisted pair in a shield will not have an exact cancel outside the shield as the centroids do not exactly match.
A triax cable would however. core as one rail, first shield as other rail, outer shield as ground and current return. I've made triax for audio supply runs before, quite easy.
jn
yes, the high voltage and high current and grounds and return... wiring. PCB is fine.
-RNM
Nothing to do with power --- There is the use of an enclosure within an enclosure isolated from each other. Grounds and shielding. Very good for sensitive, low noise circuitry.
-RNM
Last edited:
I'm not sure what you mean by "power".
I am talking about the plus and minus supply rails in a class B amplifier.
jn
I am talking about the plus and minus supply rails in a class B amplifier.
jn
🙂🙂🙂 That's power. As in power supply wiring/rails?
I just indicated my next comment wasn't about that. Your use of the word Triax cable reminded me of double and triple shielded enclosures for some reason. Care to describe what goes where in a triple shielded enclosure?
-RNM
Last edited:
Here was the result as measured by SL (It includes the null cases for
crossover analysis). Clearly it was not damaged for lack of EQ on the top.
😛
Attachments
Some small amount of improvement.
I don't know what dsp was used for EQ but this isn't very good job. Should be flat(er) by a long shot when used for EQ. These are really large peak variations in both cases. 10 and 14dB ! wow. Those two HF peaks are bad news. The decay there must be long due to what are probably resonances.
THx-RNMarsh
I don't know what dsp was used for EQ but this isn't very good job. Should be flat(er) by a long shot when used for EQ. These are really large peak variations in both cases. 10 and 14dB ! wow. Those two HF peaks are bad news. The decay there must be long due to what are probably resonances.
THx-RNMarsh
Last edited:
Yes, as in power supply wiring/rails.
No.
My speaking of triax cable was directly associated with wire. Not enclosures.
My only mention of enclosure was the fact that the circuitry is in one.
jn
John if you think of the first order integrator model of an op-amp (for instance} the unity gain crossover is determined by gm and Ccomp. If Ccomp varies with voltage (as in large junctions loading the Vas) or gm varies with input (as in an un-degenerated LTP). The open-loop BW will vary with level giving an effective PM or FM. It's pretty easy to see how small this effect is with modern devices.
Trying to improve the "fiddly bits" is fraught with problems. Just ask Slartibartfast.
I must have missed the Ccomp dependence with voltage in the Otala equations(even with the corrections).😉
My question remains.. Has anyone actually measured FM distortion?
And the true question is... I do not care what class of distortion it actually is, has anybody even tried to look for temporal distortion at the 2 to 5 uSec level interchannel with actual music. Kinda like...what humans can hear??
So far, I've seen no evidence that anyone has tried.
Jn
Ron Quan has measured a lot of amplifiers but this does not relate directly to the ITD stuff.
I haven't had that experience when I have done it (successfully) using dsp. If those huge peaks shown are called 'fiddly bits'. One could also put a passive LRC notch filter on each to flatten the response.
However, I would never have chosen speakers with that poor behavior in the first place. Sure would be interesting to know what he was thinking when he chose that driver.
THx-RNMarsh
Last edited:
I know that. It is quit plain to read that. Next subject... how to use double or triple shielded enclosures? I only see it done with test equipment. Do you have any direct knowledge about that?
I have used double shielded enclosures when I need to measure extremely low signal levels (eg <-145dBv) without extraneous noise/freq.
THx-RNMarsh
Last edited:
It's interesting that no one from the subjectivists has jumped on that - it's probably the first time in here that someone has suggested something sensible (ie not down at -100-and-lots dB or somesuch) that we can both hear, and measure, that might point to why some amps sound different.
So, how would one go about measuring this signal induced 'jitter' ....what kind of signal and what kind of sensing ?.
Dan.
Ron's first paper has a complicated analog demodulator system but 1 usec at 10k is ~4 degrees, IIRC Bob was measuring nsec's on his amps.
Ok, this is for SOTA amplifier, what about more typical MidFi/HiFi amplifiers, power and line level.
Dan.
I would probably stick to sub 10kHz, maybe 5. Did Ron look at a complex stereo signal?
We should try to spot the ITD on a specific musical source buried in both channels while each channel is carrying complex and changing content.
Jn
- Status
- Not open for further replies.