Big difference between early 70's and 70's!
Looking back, it seemed like a pivotal era: Janis Joplin, Jimi Hendrix, Jim Morrison et al died, taking with them the essence of R&R, to be replaced by disco, which got me into cowboy music (country & western).
The IBM PC AT was still a few years away, introduced in 1983, only to become affordable in the late 80's, after IBM "open sourced" their internals and so created the clone business. I was attending the technical university of Geneva, 1971-75, and the only computing resource available was a CDC 3800 with punched cards in FORTRAN. By day, the CDC was fully booked to run the circuit analysis programs we were required to write and debug, and by night, it was reserved by one of the math teachers to play chess with. I did find a way to do a .1dB RIAA: I lifted it from a professor.
I was asking about using the balance control to compensate for sound pressure imbalances due to the room with both speakers equidistant, sorry I don't know how to make it any clearer. I thought you were getting equal sound level by sitting where you are? If not, it's even worse than I thought
Hm, not sure what you are getting at with this post as your first sentence is not very clear to me. Since you think it to be quite explicitly clear, there is an obvious impasse.
As to your second sentence, well yes I get pretty equal sound pressure where I sit, which is the whole objective of positioning the speakers as I do.
Last edited:
I did find a way to do a .1dB RIAA: I lifted it from a professor.
An easy way for highly accurate RIAA eq is to use a passive topology with ideal values (and accurate parts), taking care to get the series resistor just right, considering the driving circuit's source impedance. Like PS Audio did.
You almost always sacrifice noise performance or overload with this approach. I’ve simmed a lot of these circuits and Self is correct. (See ‘Small Signal Audio Design’).
Baxandalls’s response to Lipshitz’s paper makes interesting reading. He used a split active-passive design which is very easy to get right and calls out the ‘complexity’ of Lipshitz’s equations. I suspect most of the reasonable RIAA’s of the day used this approach. But, the problem with this is you sacrifice HF overload margin to the tune of 18 dB.
The issue was with all-active - which was a major focus of Lipshtiz’s Paper. On solid state phono amps, you really want to go all-active since this brings noise and overload benefits.
(I don’t quite read the same things you do into Lipshitz’s remarks at the start of the paper Mark).
Baxandalls’s response to Lipshitz’s paper makes interesting reading. He used a split active-passive design which is very easy to get right and calls out the ‘complexity’ of Lipshitz’s equations. I suspect most of the reasonable RIAA’s of the day used this approach. But, the problem with this is you sacrifice HF overload margin to the tune of 18 dB.
The issue was with all-active - which was a major focus of Lipshtiz’s Paper. On solid state phono amps, you really want to go all-active since this brings noise and overload benefits.
(I don’t quite read the same things you do into Lipshitz’s remarks at the start of the paper Mark).
I've had no problem at all with this approach. The noise is moot with the LP surface in the equation and this will never get better.
In practice, passive is not a problem at all. Two gain blocks with RIAA in between for mm, taking into consideration the first stage's output impedance, and the second stage's input impedance. Sounds great, and there's no eq error from the open loop gain being too low/non-constant/varying unit-to unit, as in active eq. Of course, an active non-inverting RIAA has that hf problem, and an output RC filter would need buffering.
Last edited:
A simple fix is a JFET/tube cascode. For now, it sims OK, I'll tell you more when I finish building it.
They had the tools in 1964. But not everybody used the tools.
"Some of the most expensive and highly regarded disc preamplifiers on the market deviate audibly from correct RIAA equalization".
Not all of them. Some of them.
The erroneous designs ("some of" the high end preamps) are the target of Lipschitz's paper. The non-erroneous designs didn't need it.
"Some of the most expensive and highly regarded disc preamplifiers on the market deviate audibly from correct RIAA equalization".
Not all of them. Some of them.
The erroneous designs ("some of" the high end preamps) are the target of Lipschitz's paper. The non-erroneous designs didn't need it.
Come on guys, especially Mark J, it was not that easy in the early '70's. First, most of us used SLIDE RULES (I did until mid 73). Not calculators. Computers were reserved for larger companies. I ran an IBM 7094 in 1963,(note my avatar) BUT I was not allowed to run programs on it. In 1966, I became in charge of an IBM 1620 that had ECAP (electronic circuit analysis program) that was way ahead of its time. AND I could have sneaked on an RIAA program at the time, if I thought it useful. Heck, my tube Dyna preamp worked OK. But I would have gotten the WRONG answers, and I would have not have easily known that they were wrong, because there were no detailed tables showing the RIAA to easily compare to, at least to me at the time. That is where Tom Holman came in, (early 70's) yes, 6 years or more later, and I was on my own without a big computer.
It just so happens that the STANDARD RIAA used by tubes and transistors alike is NEVER perfectly accurate. It is one string of 2 caps and 2 or more precision resistors around one gain amplifier. Because of the interaction, (lack of separation) of the time constants, they smear the frequency response, and ONLY cut and try can get it acceptable for today. Back then, it did not seem very important compared to other things. Of course 2 stage phono circuits, like the Vendetta Research SCP-2 can be very accurate, because they separate 2 of the 3 RC time constants. With accurate resistors (easy even then) and 1% caps, we can get down to 0.1 dB or so. BUT the argument for a two stage design back then was not very effective. Why complicate an obvious circuit? Dyna, Marantz, etc did it with one stage. It took me years to change, and THAT is the fundamental problem with the Levinson JC-2 preamp phono stage, it was just one gain block. Just one gain block, unfortunately, compromises the entire phono section, even if the RIAA is adjusted to be approximately correct. Of course, many here will not believe this, which is OK with me. I have too many phono stages out there to worry over your opinions on this.
It just so happens that the STANDARD RIAA used by tubes and transistors alike is NEVER perfectly accurate. It is one string of 2 caps and 2 or more precision resistors around one gain amplifier. Because of the interaction, (lack of separation) of the time constants, they smear the frequency response, and ONLY cut and try can get it acceptable for today. Back then, it did not seem very important compared to other things. Of course 2 stage phono circuits, like the Vendetta Research SCP-2 can be very accurate, because they separate 2 of the 3 RC time constants. With accurate resistors (easy even then) and 1% caps, we can get down to 0.1 dB or so. BUT the argument for a two stage design back then was not very effective. Why complicate an obvious circuit? Dyna, Marantz, etc did it with one stage. It took me years to change, and THAT is the fundamental problem with the Levinson JC-2 preamp phono stage, it was just one gain block. Just one gain block, unfortunately, compromises the entire phono section, even if the RIAA is adjusted to be approximately correct. Of course, many here will not believe this, which is OK with me. I have too many phono stages out there to worry over your opinions on this.
Last edited:
This is one point I disagree with. While obviously the vinyl surface noise is order of magnitudes larger than any low noise electronics design, many users (me included) would hate a large standby (before the needle hits the vinyl) hiss. It will be useless to argue “the vinyl surface noise will hide this hiss”, they will not consider this as “low noise” by any subjective metric - and hence they won’t buy it.
Obviously you have not read "Electronics for Vinyl" by Initial LastName. The key idea is to build active electronics which monitors the 1-4 Hz Ups And Downs behavior of the cartridge ... it is a zero cost and zero effort side-effect of the infrasonic filter (whether switched IN or OUT). If there are no Ups And Downs, that means the needle is not in the groove, therefore it is safe, desirable, and wise to mute the output. Presto, zero noise.
Is this not clear, straightforward, immediately apparent? No proof of "needle in groove" therefore no output. Duh?
_
Is this not clear, straightforward, immediately apparent? No proof of "needle in groove" therefore no output. Duh?
_
Last edited:
They've been promising digital LPs for decades now ...Flat phono amps and digital RIAA. It's the future doncha know
No, I did not, and this is the dumbest idea I’ve heard lately. As dumb as self driving cabs, mind you, I wouldn’t step into one for the price of the car.
Last edited:
It also reminds us of what real questions are about. Hawking provided us a peek forward beyond Einstein (who ignored entropy) showing that gravity and entropy do something special at a black hole's event horizon. Our current universe began with a lot of gravity and not much entropy, and the arrow of time has been modifying both ever since.
The next Einstein will need to show how the two are related, but it'll take an Einstein to ask the right question.
All good fortune,
Chris
Yes agree, passive or active passive works well with tube circuits. The high impedances involved don’t really suit all active.
Last edited:
Why bother to mute the output when you can just design low noise circuits? You have a great choice of discrete devices (JFET’s etc) and really good opamps for the job.
I don’t buy the ‘why worry about the noise’ thing. Agree, it’s not as critical as a LIGO instrumentation amplifier, but just because it’s audio does not mean you should not make some effort - kind of like the whole cart resonance thing. Just sayin
(I’ve used the single resistor around the EQ network a la Linear Tech LT1115 phono amp design to provide subsonic attenuation - not as steep as a dedicated rumble filter but does a great job given its simplicity.)
I don’t buy the ‘why worry about the noise’ thing. Agree, it’s not as critical as a LIGO instrumentation amplifier, but just because it’s audio does not mean you should not make some effort - kind of like the whole cart resonance thing. Just sayin
(I’ve used the single resistor around the EQ network a la Linear Tech LT1115 phono amp design to provide subsonic attenuation - not as steep as a dedicated rumble filter but does a great job given its simplicity.)
Last edited:
- Status
- Not open for further replies.