Thanks ...
Thank you , Apex and Joni.
Those examples of Wilson mirror based amps sure helped. Further refinement
to my amp .... I'm now into the "golden realm" of ppm. 🙂
I even retained the "heart" of lazy cat's input stage. I was going to cascode
the VAS , but I already was at (rail - 7V) due to my wilson /triple EF.
Apex's marantz (what model ?) uses the wilson as a "current multiplier"
instead of a typical mirror. Full power THD dropped from .002% to .0007% ,
nearly the same as the current ratio.
Yeah , 🤐 garbage ! "walmart" Class D / various IC driven darlingtons ...
and of course , just barely enough power supply to avoid embarrassment.
I like older vintage Japanese amps , some over 20 years old. Recap them
with double the uF , they outperform modern offerings.
They are mostly VFA/EF3 with sanken OP's - I just sell the new stuff to
the "sheeple". 😀
Our DIYA creations outshine even the best OEM's , I've yet to see a
manufactured amp under 3K$ with 2- 600VA toriods + 80Kuf (PS).
OS
OS
Thank you , Apex and Joni.
Those examples of Wilson mirror based amps sure helped. Further refinement
to my amp .... I'm now into the "golden realm" of ppm. 🙂
I even retained the "heart" of lazy cat's input stage. I was going to cascode
the VAS , but I already was at (rail - 7V) due to my wilson /triple EF.
Apex's marantz (what model ?) uses the wilson as a "current multiplier"
instead of a typical mirror. Full power THD dropped from .002% to .0007% ,
nearly the same as the current ratio.
Yeah , 🤐 garbage ! "walmart" Class D / various IC driven darlingtons ...
and of course , just barely enough power supply to avoid embarrassment.
I like older vintage Japanese amps , some over 20 years old. Recap them
with double the uF , they outperform modern offerings.
They are mostly VFA/EF3 with sanken OP's - I just sell the new stuff to
the "sheeple". 😀
Our DIYA creations outshine even the best OEM's , I've yet to see a
manufactured amp under 3K$ with 2- 600VA toriods + 80Kuf (PS).
OS
OS
The second one looks like the H bridge CFA config. Great because you get hi-z inputs on both + and - Inputs. I simmed this a few months ago. Very low distortion and extremely high SR. Looks similar to the structure discussed in the ADI app note that I based my sims on.
Buf634 is, dual bias mode, protections, and bandwidth adjustment. I am using it at my last schematic.
Interesting. Can-you said some words about their different sonic signature, if any ?
laterals are softer than the BJT's, some would say more airy and delicate, but also to a degree somewhat restrained and less expansive in the dynamics than the BJT's.
I fail to hear much difference between dual and triple EF. (but this may also be a little cluttered as there also was some alteration to the PS)
I would like to work a little on the OPS that RNMarsh showed An interesting Diamond.
I fail to hear much difference between dual and triple EF. (but this may also be a little cluttered as there also was some alteration to the PS)
I would like to work a little on the OPS that RNMarsh showed An interesting Diamond.
Just saw this.
The AD846 is now obsolete, but shows what can be accomplished in audio using knowledge gained from utilising Video techniques 🙂 I'm only posting this as a talking point, & for futher insights etc.
The AD846 is now obsolete, but shows what can be accomplished in audio using knowledge gained from utilising Video techniques 🙂 I'm only posting this as a talking point, & for futher insights etc.
Thanks Zero D. - here's a link to the Jung article:
waltjung.org/PDFs/High_Performance_Audio_Stages_Using_TransZ_Amps.pdf
waltjung.org/PDFs/High_Performance_Audio_Stages_Using_TransZ_Amps.pdf
AD 811 is a sub for the AD846, Ad 844 is also very very good for audio applications, It almost begs to be used for a non feedback IV converter.
This may be true but what technical reasons could there be for this? Always like to try and relate listening experience to something technical if at all possible.
I believe that it's due to the different tempo of the device families, Bjt's are expansive in nature where the others are the opposite. I am not sure if those assumptions has any merit.
One thing I am very very sure of is that the most important may not be the circuit, sure the circuit holds the ultimate potential, but the execution of the amplifier, PCB Grounding, PSU, cabling, is of equal if not larger importance. The Box and the mechanics of the box also plays a huge role. I have made an amplifier in a super expensive milled solid enclosure, just only to find it simply sounded terrible, compared to the same sample playing on a wooden plate. Aluminum/hysteresis and output currents somehow does not match.🙁
One thing I am very very sure of is that the most important may not be the circuit, sure the circuit holds the ultimate potential, but the execution of the amplifier, PCB Grounding, PSU, cabling, is of equal if not larger importance. The Box and the mechanics of the box also plays a huge role. I have made an amplifier in a super expensive milled solid enclosure, just only to find it simply sounded terrible, compared to the same sample playing on a wooden plate. Aluminum/hysteresis and output currents somehow does not match.🙁
it is 2 decades later
811 is a really poor sub for 846 - the 846 specs were unique, tried to establish a new "precision" CFA category - guess not enough demand
846 has 50-100x the gain of the other 2 mentioned, 1/10 the Vos of the 811
and few here seem to give the high current noise spec of the -inputs of these CFA the deserved attention
you can get better performance in other CFA op amps now, but without the Tz/compensation/hi Z output pin - which the 811 never had
the only excuse I see for 811 is today is if you are a slavish follower of Gurus and rumors and can't read datasheets - I doubt Walt Jung would choose it today for new design
811 is a really poor sub for 846 - the 846 specs were unique, tried to establish a new "precision" CFA category - guess not enough demand
846 has 50-100x the gain of the other 2 mentioned, 1/10 the Vos of the 811
and few here seem to give the high current noise spec of the -inputs of these CFA the deserved attention
you can get better performance in other CFA op amps now, but without the Tz/compensation/hi Z output pin - which the 811 never had
the only excuse I see for 811 is today is if you are a slavish follower of Gurus and rumors and can't read datasheets - I doubt Walt Jung would choose it today for new design
Last edited:
I agree -- the pos vs neg tempco of the OPS power devices affect device parameters when under high current operation -- expan - pos and compressive - neg tempco. A subject to study in more detail.
And, yes, layout, especially for fast amps/wide BW amps.... layout is critical to match the SIM and work as predicted. But those are for other forums.
Thx-RNMarsh
It would be good to add vertical MOSFETs to the discussion. Have found that with the same front end that the character of the amp changed with the output devices. The verticals somehow sounded brighter when compared to Laterals. (Don't like making subjective comments). Not tried BJTs yet with said front end.
Edit: Have to wait till the next prototype is functional. The one above met a quick end due to violent oscillation and no gate protection on the MOSFETs. Won't take that short cut again.
I have never used vertical fets for OPS, I don't find them close enough in N&P channel matching, and then it also bugs me that the larger die has huge capacitance. But that's maybe just my excuse not to open another chapter. I find enough challenge in the devices I try to thame/master.
Last edited:
I know what you mean about N and P channel mismatch. That's why I decided to go with error correction on the OPS. The main benefit to verticals was that they are relatively cheap. This was a big positive for me having spent loads on my first amp design only to end in failure.
Don't think the higher input capacitance is a major issue as long as the drivers are up to the job.
I think my struggle with MOSFETs is part of the reason I haven't got round to BJTs yet. Not enough time. Wish I could spend all day every day on this stuff. 🙂
I believe, yes, dynamic tempo of devices are very important on the listening result. One more reason why we whould like to get other measurement methods than continuous signals.
About CFOPAs, i regret my favorite for audio: OP260. Very transparent, and using a higher feedback impedance (ok for most of audio circuits) than their equivalent devices designed for low z video. It is at work everywhere in my hifi system, and i used thousand of them in mixing desks.
Last edited:
This here with the dynamic tempco of the the devices is a little bit in the direction of memory distortion. I have a sense it's really important, but not really sure how to quantify it. The same goes for GND impedances. I have noticed a great improvement in dynamics and resolution when GND connections are really good. but it's not something I can see in my measurements. These issues are really important for the amplifiers ability to convey the emotions of the music.
Last edited: