Is there any interest in the community to design a RIAA LCR network yourself with the freedom of choosing your favorite impedances, the complete theoretical background, the design calculation and the simulation?
I’m working on a white paper and will publish this article in a couple of weeks in the forum.
I’m working on a white paper and will publish this article in a couple of weeks in the forum.
What would be the advantage over the RC EQ? Coils have lower Q due to DC resistance, they can pick up stray magnetic field (hum), are difficult to manufacture to tight tolerance, big and nasty... Non-air core inductors have some more problems, like hysteresis, nonlinearity, etc. I assume they are more sensitive to vibration (acoustic feedback) than capacitors.
I like your comment on my question, and that is one of the mythos. What you posted is all correct and everybody knows it and it is published and copied several thousand times in the internet.
But what about if you want design your own LCR network, e.g. with an impedance of 2.2kOhm. Do you know how to calculate all the components? What is about the answer to add the 3.18us high frequency cut-off time constant. Do you have a reference for the design?
There are so many questions in the internet not answered. My intention is to clarify some of them and to help DIY people to design their own LCR network.
But what about if you want design your own LCR network, e.g. with an impedance of 2.2kOhm. Do you know how to calculate all the components? What is about the answer to add the 3.18us high frequency cut-off time constant. Do you have a reference for the design?
There are so many questions in the internet not answered. My intention is to clarify some of them and to help DIY people to design their own LCR network.
Most of what one needs to know about LCRs is covered in the phono section of Dave Slagles :: Index. The subject is covered from almost every angle. May take a while go through it though.
In addition to the virtues Thomas mentiones I´d like to add the abscense off large series resistances found in all other passive networks.
I look forward to see your paper.
In addition to the virtues Thomas mentiones I´d like to add the abscense off large series resistances found in all other passive networks.
I look forward to see your paper.
Last edited:
I really thank you very much for the link. I think Dave has a lot of experience in designing tube amplifiers and gives a lot of contribution to the DIY community (also for me). Mostly, he's working with LTSpice which also is my favorite simulation solution (free of charge and unlimited).
I answer you spontaneously without reviewing all of Dave's details but it seems that his explanations relies on the “Tango solution” and does not consider the general generic design of an RIAA LCR filter.
My approach will be independent of any solution, where of course I also will discuss the accuracy of the “Tango Solution” but with the opportunity to choose the individual filter components based on the required (or chosen) impedance and the inductance limits without any relation to the “Tang Solution”.
Additionally, I try to explain the theoretical background, the evaluation of the formulas for calculation the component values, an Excel sheet for calculation, a LTSpice file for verification of my approach and a generic LRC RIAA tube preamplifier in LTSpice.
I answer you spontaneously without reviewing all of Dave's details but it seems that his explanations relies on the “Tango solution” and does not consider the general generic design of an RIAA LCR filter.
My approach will be independent of any solution, where of course I also will discuss the accuracy of the “Tango Solution” but with the opportunity to choose the individual filter components based on the required (or chosen) impedance and the inductance limits without any relation to the “Tang Solution”.
Additionally, I try to explain the theoretical background, the evaluation of the formulas for calculation the component values, an Excel sheet for calculation, a LTSpice file for verification of my approach and a generic LRC RIAA tube preamplifier in LTSpice.
Hi oshifis
You are completely right regarding the LCR challenge design. My interest is not to judge the advantages or disadvantages of a LCR design. On the other hand there is a large community for exactly this solution and I want to contribute.
I'm working a couple of time on a symmetrical DC-coupled tube phone preamplifier where I really have no solution for an LCR filter (I use a traditional RC filter) but I saw so many many internet links, discussions, statements, wrong solutions and... unverified solutions. My decision was to analyze the theoretical background and write this paper give my contribution to the community.
You are completely right regarding the LCR challenge design. My interest is not to judge the advantages or disadvantages of a LCR design. On the other hand there is a large community for exactly this solution and I want to contribute.
I'm working a couple of time on a symmetrical DC-coupled tube phone preamplifier where I really have no solution for an LCR filter (I use a traditional RC filter) but I saw so many many internet links, discussions, statements, wrong solutions and... unverified solutions. My decision was to analyze the theoretical background and write this paper give my contribution to the community.
Hi!
Again, I don't see any mythology about LCR. It is one possible way to do filter design.
Changing the impedance is easy. It is more difficult to calculate a new filter for different corner frequencies or to add other constants. A lot of the theory is covered for example in the book 'Motion Picture Sound Engineering'.
I have used various LCR type filters over the years and calculated many duifferent ones for example also for crossover use. These are simple to calculate. I have written about that here:
VinylSavor: Making of a Line Level Crossover
Please do share your knowledge. I am curious what you have to say and can possibly learn something new!
Best regards
Thomas
Again, I don't see any mythology about LCR. It is one possible way to do filter design.
Changing the impedance is easy. It is more difficult to calculate a new filter for different corner frequencies or to add other constants. A lot of the theory is covered for example in the book 'Motion Picture Sound Engineering'.
I have used various LCR type filters over the years and calculated many duifferent ones for example also for crossover use. These are simple to calculate. I have written about that here:
VinylSavor: Making of a Line Level Crossover
Please do share your knowledge. I am curious what you have to say and can possibly learn something new!
Best regards
Thomas
Hey Claus,
As Thomas points out, classical constant impedance filters are to be used. These are also covered in "the Audio Cyclopedia". What you call "Tango solution" is not ;-). Pultec was the first company to adapt this technique to RIAA and Tango copied it. I guess Pultec did this in the 40s or 50s using a heavily feedbacked 12AX7(gain)/12AU7(CF)-circuit. http://forums.melaudia.net/attachment.php?aid=594
This is a great thread: http://forums.melaudia.net/showthread.php?tid=230
The 3.18uS you mention is easily adopted(for those who want to) to this network with a single resistor as I showed out in Daves forum. Although it will be an approximation as Neumann used a second order filter at 50kHz.
But please share your findings, this is a great subject.
As Thomas points out, classical constant impedance filters are to be used. These are also covered in "the Audio Cyclopedia". What you call "Tango solution" is not ;-). Pultec was the first company to adapt this technique to RIAA and Tango copied it. I guess Pultec did this in the 40s or 50s using a heavily feedbacked 12AX7(gain)/12AU7(CF)-circuit. http://forums.melaudia.net/attachment.php?aid=594
This is a great thread: http://forums.melaudia.net/showthread.php?tid=230
The 3.18uS you mention is easily adopted(for those who want to) to this network with a single resistor as I showed out in Daves forum. Although it will be an approximation as Neumann used a second order filter at 50kHz.
But please share your findings, this is a great subject.
Last edited:
Yes. the PDF of your paper is 3 megs, too big to attach here. Probably the graphs and the schematics near the end that cause the big file size.
I could try to squeeze it down on Acrobat, but I don't know if it will squeeze enough to post here.
