They've got eight tubes, four per channel, in the phono stage of the SP-10, plus another four in the high-level stage! What are they doing with all these tubes? The only other phono stage I've seen that uses eight tubes is the Aikido design (http://glass-ware.store.turbify.net/ph2.html), which says in the notes that it uses one gain stage before the RIAA network, and another after, to account for the insertion loss of the RIAA network. Is Audio Research doing something similar? How do other designs with less tubes handle this?
SP-10 reference: https://www.arcdb.ws/model/SP10
SP-10 reference: https://www.arcdb.ws/model/SP10
Last edited by a moderator:
Each dual triode has its two sections in parallel.
The second stage is a cascode, followed by a cathode follower.
It's the usual ARC topology, with the cascode added.
The second stage is a cascode, followed by a cathode follower.
It's the usual ARC topology, with the cascode added.
Not offhand. Most audio guys, including those who have written books, seem to be "one trick ponies" in that respect.
Maybe Self has something in one of his books, though he seems to only use one "blameless" amplifier circuit as well.
Of course, this excepts NP, who totally runs the gamut. But he's willing to accept imperfect measurements.
Maybe Self has something in one of his books, though he seems to only use one "blameless" amplifier circuit as well.
Of course, this excepts NP, who totally runs the gamut. But he's willing to accept imperfect measurements.
Last edited:
You mentioned this phono design was "the usual ARC topology". Combined with what you just said, is it generally the case that engineers have one design that they like and don't really fool around trying "this" and "that"?
Yes, experienced engineers often prefer to stay in their lane.
Less chance of unforeseen problems, faster design cycle, etc.
There isn't time to play around like back in school, you gotta get it done.
And it has to work.
But with tube only audio circuits, there are a lot fewer options than with other devices.
Of course, that does leave the power supply to play around with.
Bill Johnson started out by playing with the Dyna PAS3, and added the phono CF stage.
Then added a regulated power supply with higher voltages, and reworked the quirky line stage.
That topology stayed around for some time, until the cascoded input fet was added.
Their power amplifiers followed a similar pattern. Cross coupled driver borrowed from scopes,
partial cathode coupling output stage, completely balanced circuit with added "phase inverter" to keep
the external operation single ended for the market, until balanced preamps were generally accepted.
The D50, D51, D75, D76, D100, D150 were all rather similar circuits, scaled up or down.
Later on, more changes were made to the basic approach.
Less chance of unforeseen problems, faster design cycle, etc.
There isn't time to play around like back in school, you gotta get it done.
And it has to work.
But with tube only audio circuits, there are a lot fewer options than with other devices.
Of course, that does leave the power supply to play around with.
Bill Johnson started out by playing with the Dyna PAS3, and added the phono CF stage.
Then added a regulated power supply with higher voltages, and reworked the quirky line stage.
That topology stayed around for some time, until the cascoded input fet was added.
Their power amplifiers followed a similar pattern. Cross coupled driver borrowed from scopes,
partial cathode coupling output stage, completely balanced circuit with added "phase inverter" to keep
the external operation single ended for the market, until balanced preamps were generally accepted.
The D50, D51, D75, D76, D100, D150 were all rather similar circuits, scaled up or down.
Later on, more changes were made to the basic approach.
Last edited:
Morgan Jones. "Valve Amplifiers" and "Building Valve Amplifiers". Awesome books. I learned everything from those.
I usually do this while seeing a movie or listening to music, so it doesn't really take up any extra time.
Fair enough to keep an open mind. That said, I don't regret reading Morgan Jones' two books.
It's a fantastic beast but requires tight tube selection at a minimum. I sprung for the original SP-8 which is considerably less drama. Even the SP-6 is quite good and it set the bar at that time. The SP-11 was also beautiful with less drama. The 10's require a special owner relationship. And I'm not sure if ARC is still up to the required commitment.
DD
DD
Morgan Jones.
https://www.amazon.com/Valve-Amplifiers-Morgan-Jones/dp/0080966403
and Merlin Blencowe Designing High Fidelity Tube Pre-amps
https://www.amazon.com/Designing-High-Fidelity-Preamps-Merlin-Blencowe/dp/0956154530
Between these two you will find a lot of the answers you seek in the realm of tube audio. Both are/have been active here in recent times.
One trick pony.. Not me... 😀
https://www.amazon.com/Valve-Amplifiers-Morgan-Jones/dp/0080966403
and Merlin Blencowe Designing High Fidelity Tube Pre-amps
https://www.amazon.com/Designing-High-Fidelity-Preamps-Merlin-Blencowe/dp/0956154530
Between these two you will find a lot of the answers you seek in the realm of tube audio. Both are/have been active here in recent times.
One trick pony.. Not me... 😀
For some reason they use two triodes in parallel everywhere. It might make sense in the input stage to reduce the equivalent input noise voltage, and maybe in the output stage to increase the maximum output current, but why they do it everywhere is unclear to me.
Otherwise it's just a configuration that's used all over the place in solid-state circuits: input stage, common-cathode/-emitter stage (which Douglas Self would call the VAS), voltage follower stage. You don't see it made with valves very often, but there is no reason why that wouldn't work. The only thing that's missing is the usual Miller compensation, apparently they got it stable without that.
I also use eight valves in my phono preamplifier, by the way, but that's including power supply rectification and output relay control.
Last edited:
It depends. As an integrated circuit designer, at work, I frequently have to come up with new circuits because the old ones don't work in the latest IC process or don't meet the latest customer requirements. Fortunately there is this thing called competition that forces engineers to try different things every now and then.
Two gain stages and a follower, classic since the days of McIntosh and Marantz, with long loop RIAA feedback. Worked well then, still works. Today, ideal RC values (with bogey valve assumptions) are easy enough for the motivated. I'd argue that the topo, with practical valves, even ECC83s, is close enough to ideal that it cannot be improved upon. Many would disagree, but it would be contentious.
The really interesting discussion is about an RIAA amplifier of an ECC83 per channel without follower, as in so many classic era (H.H. Scott, Dyna, Eico, Fisher, etc.) amplifiers. IMHO, of course.
All good fortune,
Chris
The really interesting discussion is about an RIAA amplifier of an ECC83 per channel without follower, as in so many classic era (H.H. Scott, Dyna, Eico, Fisher, etc.) amplifiers. IMHO, of course.
All good fortune,
Chris
Ha ha..... Take a look at the Aesthetix Io..... It has 6 x 12AX7 + 1 x 6SN7GT + 1 x 6922.......... pr. channel.
On top of that the PSU has a sxxtload of SS regulators + 6 x 12AX7 + 2 x EL34.
Gives "Gonna change tubes in my riaa" a whole new meaning + an unexpected grab in your purse 🤣 🤣 🤣
