It is the noise leakage, as well as the very non-linear capacitance that makes MOV's less optimum for audio. But you knew that, didn't you? '-)
No, I did not.
- What is noise leakage? I missed this concept in EE, must be specific to the high end audio.
- What is the mechanism through which the MOV nonlinear capacitance is affecting the audio performance?
Same here. I sometimes write articles on things I want to really understand, and
it forces me to prove every little step and even discuss / defend it with a peer
reviewer. That takes a lot of time, but it brings me to the leading edge on that theme.
regards, Gerhard.
That's OK Waly, you are probably right. What do I know?
Just like line cords, how could they possibly make a difference? '-)
Just like line cords, how could they possibly make a difference? '-)
Well, Gerhard, I did independently invent the complementary differential bipolar input stage when I was at Ampex back in 1968. What were you up to then? Later, I designed what appears is the first complementary differential jfet input stage in 1971-2. Then, I incorporated the dual complementary folded cascode input stage in 1981-2. Do you have any better suggestions?
To my surprise, I found myself, by name (associated with one of my audio products) featured in a news article on my I-phone just last night. My reputation appears to be intact, no matter what is said around here. '-)
To my surprise, I found myself, by name (associated with one of my audio products) featured in a news article on my I-phone just last night. My reputation appears to be intact, no matter what is said around here. '-)
A ham friend of mine is able to produce a KW on 144 MHz from his car.
He can toggle the street lights in the Schlangenbader Tunnel (in Berlin)
and approaching the car with a gas discharge lamp in your hand makes you
look like Darth Vader with his light saber.
I'm kind of jealous, but with modern cars we are limited to 10 Watts.
He drives a > 30 year old 730 BMW Diesel, with nearly no electronics, not
even an ignition. Yes, Diesel.
regards, Gerhard
Learning Latin, Cesar & Integrals.
I was talking about the last 40 years. There was a lot of change in your area.
OK, in the audio area.
And, because you ask me so directly: I think I'm quite good at space-bound oscillators,
phase noise, clock comparators and ps time/jitter measurements.
Last edited:
Hope it's not bad manners to quote myself!
But I wanted to also point out that emi or transient control devices in a component or changes to the safety ground circuit or wire can also affect and worsen induced ground loop noise in a system. Like if a MOV cuts in in a preamp box or a clc filter conducts to safety ground, it induces a big spike onto the input ground shield if there is (there almost always is!) a ground loop. The "fix" makes the problem worse, because it is trying to fix the wrong aspect of the problem (the noise, rather than the ground loop)
Last edited:
There is a lot to be said for the old clockwork diesels. The last you could get in UK was an IH unit that started life as the land rover 200Tdi was sold to IH and bored to 2.7. With an engine driven lift pump once started needed no electricity.
Now hands up anyone who was ever brave enough to use a cartridge start engine? I will say I wasn't 🙂
This is a time-honored approach to understanding things. Alternatives include writing an entire book, or teaching a class on the subject.
I set out to write a paper back in 1982, and going in I supposed it would be simple, as the instrument I was describing already existed. It took far more time than I expected, because in some cases I found I didn't really understand certain criteria for optimization, once I tried to justify them. I had hoped as well to include additional explanations of circuit techniques, but I ran out of time and space, and almost jeopardized the indulgence of my boss, the department chairman.
I presented the paper at a conference and had some enthusiasm generated at that moment, but in the long run few people followed anything I suggested. However, about a year or so later I had the satisfaction of a call from someone doing somewhat-similar systems, including clock drivers, in this case for driving CCD imaging detectors. After answering a few questions, he said You mention something on page so-and-so about "details of such isodissipational design techniques will not be discussed further". Where can I read more about this?
I said Well I don't know. As far as I know they are original ideas.
Even more educational is when you start a paper and analysis to make one point.... and then discover in the effort that the point was entirely incorrect! Math and real scientific method can be humbling things when applied.
John:
Please provide some evidence (no peeking at least) that there is any basis for your claim. Clearly you aren't current on power protection or you would not recommend a gas tube across the line. Gas tubes won't quench during 50 Hz or 60 Hz power and become a direct short with about 0 Ohms. If there is no fuse then the problem goes upstream with pretty serious consequences. Otherwise they have a similar characteristic to an MOV, a small capacitance until you get to breakover then conduction. The MOV conducts like a Zener with much higher current capacity. The Gas tube shorts. MOV's have higher capacitance than gas tubes but when I last checked they were no less linear than an equivalent Mylar "X" cap. Also do not use a cap not rated for across the line without a series fuse, otherwise the "self healing" can have catastrophic effects as well.
Its a given that pretty much any domestic or commercial/industrial environment will have some MOV's across the line. Pretty much any modern power supply will have them. I believe its effectively a requirement for EC approval.
Playing with stuff across the line is pretty serious. We test surge protectors with pretty high power, representative of what they need to deal with under fault conditions. A peak surge inside a home is expected to be 6KV @ 3KA (MOV's can handle this pretty easily) and we expect a surge protector to open with 500A at 240V. In 240V countries its more like 480V @ 500A.
If the line voltage is high enough to get MOV's to conduct you will have problems pretty quick. While an MOV can handle 400 KW for 20 uS the continuous dissipation is more like 1 W. There has been a tendency to cut things fine on MOV selection for power to get better let through numbers. If your power is not well controlled and stable you can get into trouble. A 130V MOV (Most common in the US) will start significant conduction around 160V continuous and quickly burn up, often shorting in the process.
Power line networking can use up to 5A of RF. itss all below 30 MHz because the losses on the power distribution are too great above 30 MHz. However a noisy power source (like a cheap wall wart power supply) can radiate as well as conduct down the power line. I would suggest getting a small noise/surge suppressor on every noise generator as close to the source as possible. Once its in the power network its going to radiate. And once radiating its difficult to control and a power line filter at the sensitive device is useless for that.
To get a good handle on the conducted noise you need to look at all three pairs of conductors. Over a distance the energy will be shared among them. Adding a ferrite will alter the noise distribution but you can't be certain how. Some of the noisiest devices I have measured are CD players. Not that you can expect much in a $30 CD or DVD or BluRay player.
A good ultra-isolation transformer with a good low inductance ground return (1/2" braid) for the shield is a reasonable start on clean power. My experience has been the touchy-feeley types don't actually like clean power. The loss of highs, detail and dynamics seems to be seen as a negative.
To get a good handle on the conducted noise you need to look at all three pairs of conductors. Over a distance the energy will be shared among them. Adding a ferrite will alter the noise distribution but you can't be certain how. Some of the noisiest devices I have measured are CD players. Not that you can expect much in a $30 CD or DVD or BluRay player.
A good ultra-isolation transformer with a good low inductance ground return (1/2" braid) for the shield is a reasonable start on clean power. My experience has been the touchy-feeley types don't actually like clean power. The loss of highs, detail and dynamics seems to be seen as a negative.
Damien, loss of highs are from poorly designed equipment, severe mismatches in impedance in signal chain, or.... resonances on the power line that eat even more current.
Cleaner is always better, but it's rarely done right.
Cleaner is always better, but it's rarely done right.
Last edited:
- Status
- Not open for further replies.