If you use SET input topology, thats true. But then the distortion is higher also.
If you use compliementary push-pull input the base currents cancel and no coupling cap is needed.
THx-RNMarsh
This has been my experience.
In fact in most cases I have found BJT's to sound and measure superior to
Fets but there is a LOT in the subtleties of how to use them properly -
especially if you are using no / low feedback.
T
Bipolars only really excel when running in CURRENT MODE. This is because the non-linear exponential characteristic of the bipolar is CANCELLED by the 'diode' distortion input.
Richard, you must read a book on this, or an equivalent lecture to understand that fets are INHERENTLY more linear than bipolar, BUT less linear than tubes. It is in the fundamental equations and the distortion predictions from these equations. IF you drive your original circuit hard enough, you will see higher order distortions.
Richard, you must read a book on this, or an equivalent lecture to understand that fets are INHERENTLY more linear than bipolar, BUT less linear than tubes. It is in the fundamental equations and the distortion predictions from these equations. IF you drive your original circuit hard enough, you will see higher order distortions.
You told me this 40 years ago. I just dont see it's inherent superiority as a practical matter though.
Yes, I am sure as I get closer to clipping that i will see odd harmonics start to show up. How much harder? I am running original +/- 24vdc.
I run output to 2.5 v rms and see still no 5,7,9 11 etc. Which is my point.
Do I need to go higher output than DAC normal output levels or PA normal max input drive?
THx-Richard
Last edited:
This is not my cascoded version et al. But increasing from 1v rms to 3 v rms and 20KHz (no filters applied) the harmonics are still below -100 un-trimmed/ 20% matched version.
How hot should I be driving it ?
Its late. I will see all the harmonics when I get closer to clipping. But for all practical purposes at normal max output levels, well below clipping, the 2H and 3H from all bipolar only dominate here..
Later.
-Richard
Last edited:
Kulish Circuit
https://www.diyaudio.com/forums/solid-state/74861-single-darlington-line-preamp-3.html#post857018
I am producing that evidence now in the attachment.Tektronix used bipolar transistors to cancel tube distortions...
For every poison there's a cure.
Attachments
Last edited:
Mark Levinson Audio System ML-2
American Renaissance
The little-known and underrated Hadley Laboratories 622C aside, the John Curl designed ML-2 was the first solid-state amp to put the transistor on an equal footing with the tube. It featured Curl’s JFET complementary differential input stage, a circuit so remarkable that it would soon become a staple of virtually all high-end amplifiers (and remains so to this day). Introduced in 1977, these 25W Class A monoblocks (50W into 4 ohms, 100W into 2 ohms) from Mark Levinson Audio Systems set a new standard of clarity, liquidity, timbral beauty, and three-dimensional imaging and soundstaging for solid-state. The polar opposite of the grainy, piercing, high-in-odd-order-harmonics sound that many of us then associated with solid-state designs, the ML-2 proved that the transistor could make music as readily as it could generate power. The ML-2 gave new meaning to the word “transparency,” and set a benchmark against which all subsequent solid-state amplifiers were judged. Perhaps more than any other single product, the ML-2 exemplified, in performance and build-quality, the vanguard of the American high-end renaissance. Today, the ML-2 is the most collectable of all solid-state amplifiers.
I have to think that FET's played a big part of this amps sound. This quote is from The Absolute Sound magazine.
American Renaissance
The little-known and underrated Hadley Laboratories 622C aside, the John Curl designed ML-2 was the first solid-state amp to put the transistor on an equal footing with the tube. It featured Curl’s JFET complementary differential input stage, a circuit so remarkable that it would soon become a staple of virtually all high-end amplifiers (and remains so to this day). Introduced in 1977, these 25W Class A monoblocks (50W into 4 ohms, 100W into 2 ohms) from Mark Levinson Audio Systems set a new standard of clarity, liquidity, timbral beauty, and three-dimensional imaging and soundstaging for solid-state. The polar opposite of the grainy, piercing, high-in-odd-order-harmonics sound that many of us then associated with solid-state designs, the ML-2 proved that the transistor could make music as readily as it could generate power. The ML-2 gave new meaning to the word “transparency,” and set a benchmark against which all subsequent solid-state amplifiers were judged. Perhaps more than any other single product, the ML-2 exemplified, in performance and build-quality, the vanguard of the American high-end renaissance. Today, the ML-2 is the most collectable of all solid-state amplifiers.
I have to think that FET's played a big part of this amps sound. This quote is from The Absolute Sound magazine.
I would like to have a sit down talk with him about his responses to questions at the end. First of all 'expectation' as shown in the slide is the wrong word, it should probably be 'primed.' Second, people don't miss the gorilla in the movie 'because they weren't expecting it', they missed it because they had been given a cognitive loading task to do while watching the basketball game. That is to say, they missed the gorilla because of distraction caused by cognitive loading (same type of thing that makes ABX an insensitive test for showing detection of small differences, IMHO).
Last edited:
Yes, no argument about the sound of early SS amps producing a lot of odd harmonics are NFG for accuracy. Just FETs are not the only one's any more. Fets are still doing a great job.
However, many newer designs do as well using bipolar. Esp some of the CMA circuits and also DIYAudio's DADod's circuit for example. All bipolar except OPS.
THx-RNMarsh
Last edited:
..that is a theory, not a fact. We don't know enough to really now why the gorilla is missed. Cog Psych is still very much a work in progress.
- Status
- Not open for further replies.