John, I have indeed, and in everything from chamber ensembles to large forces (I play saxophone, clarinet, flutes, although I haven't done much recently) and I also have written a lot of music, especially back in high school days. And that's why I told the composer I suspected he would revise the piece, because the place-losing was so adroitly concealed to my ears at least. The music of this composer is generally quite "accessible"; had it been a typical academic atonal piece it might well have been more difficult to notice, but such was not the case. That last movement alternated between the dance rhythms of the waltz and the tango, and Mark said in greater detail today via email that it was a portion of the tango that got messed up.
I may go to hear it done again on the 13th, in hopes that by then the issues have been corrected. I am sure the players were quite embarrassed. The piece did get a quite enthusiastic reception nonetheless, and there were a good many composers and musicians in the audience.
Brad
Last edited:
Or you can do like in guitar amps and use the point where the input
ground (ring) ties to the chasis (no insulators) as the center of your 'star' grounding system.
Help, I'm a Klugscheisser
Aluminium cases have to be (approximately) 11 times as thick as iron (steel) cases
to reach a comparable magnetic shielding between 50 Hz and 100kHz.
The shielding effect of iron increases fast above 1kHz
(and is alomst constant below 1kHz).
(Data from: P. Skritek: Schirmungen in der Audio-Schaltungstechnik)
Aluminium cases have to be (approximately) 11 times as thick as iron (steel) cases
to reach a comparable magnetic shielding between 50 Hz and 100kHz.
The shielding effect of iron increases fast above 1kHz
(and is alomst constant below 1kHz).
(Data from: P. Skritek: Schirmungen in der Audio-Schaltungstechnik)
Last edited:
You know any cut or opening made in the shield worsens HF shielding effect, and a mere cap connected between a connector body and a box does not help much. You either need to use RF connectors with body firmly connected to the case or, in case of insulated connectors, to use multiple shield design, like 2 or 3 shielded boxes inside the main shielded box.
Yes, critical listening is always best done when using a cellphone.
I have a computer in the vicinity of my stereo. In fact, it's my primary music source. My equipment is designed with sensible and basic rather than heroic RF protection. The computer, a cheap Dell, seems to cause no untoward problems. Nor do the switching amps used on the woofers.
The situation for thousands of feet of wiring in large venue sound reinforcement is a different story (anticipating a stadium anecdote from Ed).
I have a computer in the vicinity of my stereo. In fact, it's my primary music source. My equipment is designed with sensible and basic rather than heroic RF protection. The computer, a cheap Dell, seems to cause no untoward problems. Nor do the switching amps used on the woofers.
The situation for thousands of feet of wiring in large venue sound reinforcement is a different story (anticipating a stadium anecdote from Ed).
I think a sense of proportion is necessary. We are not talking about making a totally RF-tight enclosure, merely attenuating interference sufficiently that it can't do much harm. Microwaves (e.g. from cellphones) can sneak through slots but it doesn't take much low pass filtering in the circuit to stop them from getting any further. MF/HF broadcasts are less affected by circuit filters, but can't get through a metal screen and see all but the biggest holes as a continuation of the screen so a single cap with short leads will help.
If you want multiple screens and continuous shielding then you probably should be designing professional RF sig gens rather than audio equipment.
If you want multiple screens and continuous shielding then you probably should be designing professional RF sig gens rather than audio equipment.
Well since called upon....
Doing my noise induced issue thingie...
Vacuum tubes have the least problem with out of band noise causing audio band artifacts even at high levels.
Fets are almost as good but don't like really high levels (1 volt or so of HF noise really shows up.)
Bipolar transistors can start having problems at as little as 30 millivolts of noise and maybe much lower depending on circuit topology.
As to using a ceramic bypass capacitor SY, I will use a short wire story. In the late 60's the ARRL (Handbook maybe?) as I recall had someone do self resonant experiments on standard leaded capacitors. .001 uf caps with full length leads were self resonant below the 2 meter band.
Doing a quick search yields http://www.home.agilent.com/upload/cmc_upload/All/exp79.pdf?&cc=US&lc=eng
where they report 2 mm leads on a 1 nF ceramic capacitor have a 94 Mhz resonance! That would limit my use of such a filter to under 10 Mhz using a conservative design value.
Now guitar amplifiers have used bypass capacitors to keep RFI out and it does work. Hearing someone else's music in the background on your head can be a bit of a challenge. But that is because most of the issue is with low level detection of local strong AM stations. So small ceramics will work fine in this application.
Now the issue in far more demanding listening applications is not if you hear Rush Limbaugh, but if RFI degrades the reproduction quality. In simple A/B tests it is very easy to detect deliberately induced AC line noise while playing CD's. Now while it is not practical to do source, path, receive analysis to be a definitive study on the issue, it is possible to produce some guidelines.
5% harmonic distortion is allowed on AC lines as within normal power quality standards. RF noise does not couple well through AC power distribution transformers. Home electronics have FCC RFI (EMI for younger folks) limits. So it is reasonable to expect that audio equipment should not be affected by any of these sources at a reasonable distance. (6")
Now in quick playing with some common circuit topologies there are issues that show up even within those limits.
Then there is the issue of digital audio sources. It is clear from listening that bipolar circuits do not perform as well as FET ones inside the equipment case. The question becomes what measurements will show this? (This degradation is in some cases not the I can sort of hear it but rather the "Is it broken" level.)
ES
Doing my noise induced issue thingie...
Vacuum tubes have the least problem with out of band noise causing audio band artifacts even at high levels.
Fets are almost as good but don't like really high levels (1 volt or so of HF noise really shows up.)
Bipolar transistors can start having problems at as little as 30 millivolts of noise and maybe much lower depending on circuit topology.
As to using a ceramic bypass capacitor SY, I will use a short wire story. In the late 60's the ARRL (Handbook maybe?) as I recall had someone do self resonant experiments on standard leaded capacitors. .001 uf caps with full length leads were self resonant below the 2 meter band.
Doing a quick search yields http://www.home.agilent.com/upload/cmc_upload/All/exp79.pdf?&cc=US&lc=eng
where they report 2 mm leads on a 1 nF ceramic capacitor have a 94 Mhz resonance! That would limit my use of such a filter to under 10 Mhz using a conservative design value.
Now guitar amplifiers have used bypass capacitors to keep RFI out and it does work. Hearing someone else's music in the background on your head can be a bit of a challenge. But that is because most of the issue is with low level detection of local strong AM stations. So small ceramics will work fine in this application.
Now the issue in far more demanding listening applications is not if you hear Rush Limbaugh, but if RFI degrades the reproduction quality. In simple A/B tests it is very easy to detect deliberately induced AC line noise while playing CD's. Now while it is not practical to do source, path, receive analysis to be a definitive study on the issue, it is possible to produce some guidelines.
5% harmonic distortion is allowed on AC lines as within normal power quality standards. RF noise does not couple well through AC power distribution transformers. Home electronics have FCC RFI (EMI for younger folks) limits. So it is reasonable to expect that audio equipment should not be affected by any of these sources at a reasonable distance. (6")
Now in quick playing with some common circuit topologies there are issues that show up even within those limits.
Then there is the issue of digital audio sources. It is clear from listening that bipolar circuits do not perform as well as FET ones inside the equipment case. The question becomes what measurements will show this? (This degradation is in some cases not the I can sort of hear it but rather the "Is it broken" level.)
ES
Last edited:
I am glad to hear you finally grasp the real power behind magic stones and boxes!
Yes Scott, I loose sleep over it every night!
Please take back the results!
Ed, I can always count on you to make a simple issue more complicated than it needs to be.
We have a series resonance, not parallel. What's the impedance at 10 MHz for the cap which worries you? What's the impedance at the self-resonant frequency? Now, what's the impedance at 150Hz? What's the input impedance of the amp (including input capacitance)? How do these scale? How does all this compare to the far more common X7R or Z5U dielectrics?
We have a series resonance, not parallel. What's the impedance at 10 MHz for the cap which worries you? What's the impedance at the self-resonant frequency? Now, what's the impedance at 150Hz? What's the input impedance of the amp (including input capacitance)? How do these scale? How does all this compare to the far more common X7R or Z5U dielectrics?
SY,
Bypassing a shield with a .01 uf ceramic Z5U capacitor to reduce RF noise using 1/2" leads is well below 1 Ohm at AM radio frequencies. This can produce RFI losses of 50 db or so. It is almost useless at FM radio frequencies, as it will increase the loss by less than 3 db.
Is that simplified enough?
ES
- Status
- Not open for further replies.