Chris,
The discussion was about AC line problems designers need to be aware of. The AC line variation requires a regulated reference to avoid the problem. Just filtering doesn't get it. More importantly power supplies actually have resonant frequencies, so designers need to pay attention.
If you want to test for many of the AC line noise problems connect the secondary of a 6 volt transformer through a 25 ohm power resistor to the safety ground of your bench AC supply. The other end goes to the equipment chassis or RCA connector shell. A well designed piece of equipment will not have any issues. Not so good will hum and the bad guys will step outside for a smoke!
Hags,
I've used a bit more than a dozen of the Sony Tuners, all of them seem to work the same.
ES
Spoke to Dick Sequerra and Mitch Cotter. Both recommended initially to Saul Marantz that FM tuners were rather expensive to develop, but commercial pressures from audio dealers got Saul to take a chance in making one to compete with McIntosh, to complete the line.
I understand this pressure, as I was VP Engineering of Lineage with Saul Marantz as President, and I was under pressure to make a receiver, with FM, even before we got our amp and preamp beyond prototypes. Wasted a lot of effort, but that is the 'dealer pressure' for you.
I understand this pressure, as I was VP Engineering of Lineage with Saul Marantz as President, and I was under pressure to make a receiver, with FM, even before we got our amp and preamp beyond prototypes. Wasted a lot of effort, but that is the 'dealer pressure' for you.
Last edited:
Hi ES,
Filtering does help immensely with the one issue that electronic regulation has trouble with. High frequency noise. You have to consider all possible problems, within reason of course!
Right now I check leakage with a resistor from chassis to bench safety ground and measure the voltage across that. I'm using an HP 34401A, so no problems with resolution. Most of the troubles I am seeing are not transformer leakage though. The test you are suggesting might give some of my test equipment a rough time unless the current is limited!
The only types of equipment I see with supply resonance problems are some tube products in the high voltage section. However, there was a Bedini that I ran into as a much younger fellow that may have suffered from this. The two filter caps were massive and ran from the front to the rear of the chassis. There was a substantial 120 sawtooth riding on top of a sine wave (!). The construction of that thing was shoddy, and I'm being nice about it. I just put it down to a poorly designed amplifier that the Hi-Fi fraternity happened to love. The electronic design was unremarkable, although the removable cards were a nice touch. The physical construction was truly rough and sloppy. Even to the poor output transistor heat spreader ("L" bracket) to heat sink contact. That thing was a complete rip-off as far as I was concerned, so I didn't worry about it too much. I did advise the owner to sell it on down the road. He did.
-Chris
Very true, but the short comings go much deeper than that. Regulation was one of the very first things I drew attention to.
Filtering does help immensely with the one issue that electronic regulation has trouble with. High frequency noise. You have to consider all possible problems, within reason of course!
Right now I check leakage with a resistor from chassis to bench safety ground and measure the voltage across that. I'm using an HP 34401A, so no problems with resolution. Most of the troubles I am seeing are not transformer leakage though. The test you are suggesting might give some of my test equipment a rough time unless the current is limited!
The only types of equipment I see with supply resonance problems are some tube products in the high voltage section. However, there was a Bedini that I ran into as a much younger fellow that may have suffered from this. The two filter caps were massive and ran from the front to the rear of the chassis. There was a substantial 120 sawtooth riding on top of a sine wave (!). The construction of that thing was shoddy, and I'm being nice about it. I just put it down to a poorly designed amplifier that the Hi-Fi fraternity happened to love. The electronic design was unremarkable, although the removable cards were a nice touch. The physical construction was truly rough and sloppy. Even to the poor output transistor heat spreader ("L" bracket) to heat sink contact. That thing was a complete rip-off as far as I was concerned, so I didn't worry about it too much. I did advise the owner to sell it on down the road. He did.
-Chris
When talking to Dick Sequerra today, I mentioned the ceramic coupling caps in the Sony tuner. He started off to tell me how awful CERAMIC CAPS are, and I had to remind him that I have criticized their use in audio circuits for decades. He did agree that they can be a big problem, and that I should try to eliminate them, if possible.
John, the ceramic coupling caps in the Sony tuner, are these in the RF, IF or the (demodulated) audio section?
Being that you are an physicist by training and know so much about electron flow that you can explain why wires sound different according to the Fermi velicity of electrons in them, surely you can explain the physics of why ceramic capacitors are inferior to other types. I presume that is for use as interstange coupling and DC blocking capacitors. Or is that your conclusion for other applications as well?
Ceramic Caps
Hello Soundminded
Apart from NPO and COG types ceramic caps have measurable harmonic distortion. Does this partly answer your question or are in interested in the distortion mechanism.
Hello Soundminded
Apart from NPO and COG types ceramic caps have measurable harmonic distortion. Does this partly answer your question or are in interested in the distortion mechanism.
the high dielectric ceramic capacitors have a significant voltage coefficient of capacitance, and a high dielectric absorption coefficient, but to my mind and ears it is not relevant for npo types.
Due to the small sizes available ceramics (X7R, X5R) are used almost exclusively for decoupling on digital boards, with even moderate speeds. The X7R are not to bad, but I wouldn't use for interstage coupling. But due to the physical sizes available they are the best choice for digital decoupling capacitors.
NPO and COG are used a lot in RF circuitry and are not bad caps. Due to the construction they are physicaly much larger than the X7R type for any paticular value.
I cannot understand this attitude towards ceramics, where in some instances they are the cap of choice.
NPO and COG are used a lot in RF circuitry and are not bad caps. Due to the construction they are physicaly much larger than the X7R type for any paticular value.
I cannot understand this attitude towards ceramics, where in some instances they are the cap of choice.
I was hoping for an explanation of how and why. I don't simply accept blanket statements without some reason to support them. Since JC is the physicist, I thought he'd have something to say.
While this test is ok, it is not nearly enough, as it only looks at IR drop at 60 hz.If I remember correctly, Bill Whitlock used a diode in series to force half sine into pin 1, giving a richer harmonic spread.
Here is an even better test, one I posted back in 2008. It MUST be performed through the audio bandwidth. It will find IR drop sensitivities, it will find primary coupling sensitivities, and it will find secondary coupling sensitivities.
Be careful...it also finds errors in the source equipment output stage, something that is not even thought of.
Cheers, John
Attachments
Ceramic cap distortion: You can easily google it, this is well documented. There is no need to re-describe well known issues and measurements already performed and published.
I was going to say that it's so easy to google it, but Pavel was quicker.
There, for the not so versed ones in the crowd, of whom I am first:
http://www.atceramics.com/pdf/technotes/piezoelectric_effect.pdf
There, for the not so versed ones in the crowd, of whom I am first:
http://www.atceramics.com/pdf/technotes/piezoelectric_effect.pdf
Here is another diagram. It details a bit more what is going on.
It describes the mechanism which allows line cords, outlets, and IC's to alter the sound of a system.
Cheers, John
It describes the mechanism which allows line cords, outlets, and IC's to alter the sound of a system.
Cheers, John
Pin 1 problem
Hi all, good discussion about conducted interference.
I have found Jim Brown's papers at:
Audio Systems Group, Inc. Publications
dealing with reducing interference to be good refreshers, and worthwhile addition to my reference on the subject:
Amazon.com: Grounding and Shielding Techniques (9780471245186): Ralph Morrison: Books
Just a thought for those reading this thread who may not have heard of them.
Howard Hoyt
CE WXYC-FM 89.3
Chapel Hill, NC
Hi all, good discussion about conducted interference.
I have found Jim Brown's papers at:
Audio Systems Group, Inc. Publications
dealing with reducing interference to be good refreshers, and worthwhile addition to my reference on the subject:
Amazon.com: Grounding and Shielding Techniques (9780471245186): Ralph Morrison: Books
Just a thought for those reading this thread who may not have heard of them.
Howard Hoyt
CE WXYC-FM 89.3
Chapel Hill, NC
Thanks John,
Your test should show more problems.
One question, Have you ever measured typical stray magnetic fields and how much current they actually contribute.
ES
Not within a typical home environment.
Unless of course you use 30 kA supplies, 5 tesla magnets, and measure nano and microvolt signals in the .1 hz range..at home.. As it is, we generally have to replace all the CRT's within 20 feet or so of the power conductors with flat panels, as the display content is usually color coded. The colors change as the power supply current ramps up.
Of more importance is the coupling mechanism. Once everybody understands fully the couplings and measurements, it is trivial to get to how to fix it..
My test setup measures the ground loop current forced by the driven magnetic field. It can also be used to measure the haversine amp draw ground loop current as well...although that is a further complexity most are not ready to undertake.(sources include line cord, transformer primary and line cord give odd harmonics and audio, secondary gives even harmonics and audio, supply caps couple...etc..)
For the non-geeks, my diagrams will be useful simply because they demonstrate where the physical problem lies. For example, most do not suspect that twisting the IC's around the line cord of the pre to the outlet, then up the amp line cord to the amp would actually reduce hum pickup..not something that can be done for large professional systems of course..
For the geeks, like Jim Brown and Bill Whitlock and yourself, it provides a more complete worldview of the systems they work on, with previously unsuspected coupling mechanisms. I expect these geeks to understand the engineering aspects of my design, put this kind of hardware together, measure, write, and publish papers. From there, it filters to everybody else in either knowledge, or hardware...everybody wins.
Cheers, John
Last edited:
The ceramic caps that cause a problem are in the AUDIO section, of course. C101 and C102, for example.
OF COURSE, for RF decoupling, ceramic caps work OK.
I first published the distortion in a typical large value ceramic in my IEEE paper 'Omitted factors in Audio Design' in 1978. B&K beat me to it in an app note, after I showed the B&K engineer about 1 year earlier. However, I was first shown this distortion mechanism in ceramics at Tektronix in Feb. 1974, at their facility near Portland, OR on a modified 577 curve tracer.
If anyone knows how they modified the 577 to measure caps, please give me the info. I had it once, many decades ago, and tried it on my own with a 577, but that info was lost in the firestorm in 1991. It had to do with bringing out a drive wire and adding a function generator triangle wave through this wire to make the 577 measure cap values. Some sort of dv/dt= I/C. Anyone from Tektronix remember this project?
OF COURSE, for RF decoupling, ceramic caps work OK.
I first published the distortion in a typical large value ceramic in my IEEE paper 'Omitted factors in Audio Design' in 1978. B&K beat me to it in an app note, after I showed the B&K engineer about 1 year earlier. However, I was first shown this distortion mechanism in ceramics at Tektronix in Feb. 1974, at their facility near Portland, OR on a modified 577 curve tracer.
If anyone knows how they modified the 577 to measure caps, please give me the info. I had it once, many decades ago, and tried it on my own with a 577, but that info was lost in the firestorm in 1991. It had to do with bringing out a drive wire and adding a function generator triangle wave through this wire to make the 577 measure cap values. Some sort of dv/dt= I/C. Anyone from Tektronix remember this project?
A competent design will avoid large signals across the capacitor and thus avoid this distortion. A blanket statement like this has no value without data.
jan didden
- Status
- Not open for further replies.