For bass , the power supply current capability and damping factor of the OPS is the major factor.
BJT or FET , all the same. Either one poorly designed will sound ... well - poorly !
Full range is another story , the input stage / feedback method is the key.
My builders did prefer the BJT EF3 (for bass) , but they were just swapping
HEXFET's for the BJT's (in EF2 mode).
For those "crystal clear" engaging mids and highs - CFA feedback or a fast VFA
(like the "leach" front end). OPS type didn't matter.
There is a lateral OPS for my input stages , the BJT EF3 still was preferred over 4:1
over any FET.
PS - I did not design the FET output stages , so who knows ??? I HAVE heard VERY
good lateral output stages. 🙂
OS
BJT or FET , all the same. Either one poorly designed will sound ... well - poorly !
Full range is another story , the input stage / feedback method is the key.
My builders did prefer the BJT EF3 (for bass) , but they were just swapping
HEXFET's for the BJT's (in EF2 mode).
For those "crystal clear" engaging mids and highs - CFA feedback or a fast VFA
(like the "leach" front end). OPS type didn't matter.
There is a lateral OPS for my input stages , the BJT EF3 still was preferred over 4:1
over any FET.
PS - I did not design the FET output stages , so who knows ??? I HAVE heard VERY
good lateral output stages. 🙂
OS
Last edited:
At bass frequencies, there is always lots of feedback available to linearise the output stage, so BJTs and MOSFETs should handle bass exactly the same.
Bruno Putzeys suggested that the bass sound quality of power amps might actually be explained by a high level of crossover distortion that adds a tiny bit of high frequency "fizz" to bass notes, and this makes them sound subjectively tighter and faster.
I think this is a reasonable suggestion, because when someone makes a subjective judgment of bass sound quality, they probably take the whole frequency spectrum into account. Bass taken strictly by itself is not fast, slow, tight, flabby or whatever. A subwoofer listened to alone will never sound tight or fast.
Bruno Putzeys suggested that the bass sound quality of power amps might actually be explained by a high level of crossover distortion that adds a tiny bit of high frequency "fizz" to bass notes, and this makes them sound subjectively tighter and faster.
I think this is a reasonable suggestion, because when someone makes a subjective judgment of bass sound quality, they probably take the whole frequency spectrum into account. Bass taken strictly by itself is not fast, slow, tight, flabby or whatever. A subwoofer listened to alone will never sound tight or fast.
Don't know, I am using it with my Acurus and I like it!!!
Don't know whether it's a BJT or MOSFET. Never open it. I am happy with it and have no intention to replace it.
Very good point, never stop and think about this. I just know what bass sounds good to me. I love my Velodyne, never cross my mind to replace it. It becomes my coffee table and I even custom made a 3/4" thick glass top for it and it looks nice!!!
Without the regular stereo, the sub sounds like "muuu muuu muuu...." anyway!!!!
Quite possibly so. This may explain, at least partially, the great bass of old Krells. So maybe bass isn't that great without some added hi frequency distortion 🙂
I know people get upset with suggestions of an inherent "sound" of active devices, but i fail to see a reason. So caps and resistors are allowed to have "sound", even if differences only exist below -130db level, but output devices, even with wildly different transfer characteristics are not?
Or input devices for this matter. Does anyone believe that all small signal bjts per example have to sound the same in a competent circuit?
I think human ears are more sensitive than nay test instrument can show. In guitar amp, after I build and trouble shoot to make sure everything is working electrically, I put away the scope and generator. Then it's just my ears to work on the amp. So many times I simulate with LTspice to copy the response of a network but it sounded totally different in real circuit.
I have been studying theories and all, I have theories coming out my ears. BUT I learn from experience, sound is a totally different animal and you might as well throw all the knowledge out the window after you get the circuit to work electrically. I once thought I am ahead of the game in audio because I studied a lot of theories and have a lot of experience in electronics. WAS I WRONG!!!! I am learning as any green horn tinkering with a soldering iron!!!!
I believe there is a big disconnect between theory and measured data to the sound. After all, it's the sound that matters, not how much power, distortion and all.
That's the reason I still going to design and build a tube amp no matter what people said it's the "distortion.....the distortion" that matter. There is so much more to sound than distortion, it's an art!!!!
I have been studying theories and all, I have theories coming out my ears. BUT I learn from experience, sound is a totally different animal and you might as well throw all the knowledge out the window after you get the circuit to work electrically. I once thought I am ahead of the game in audio because I studied a lot of theories and have a lot of experience in electronics. WAS I WRONG!!!! I am learning as any green horn tinkering with a soldering iron!!!!
I believe there is a big disconnect between theory and measured data to the sound. After all, it's the sound that matters, not how much power, distortion and all.
That's the reason I still going to design and build a tube amp no matter what people said it's the "distortion.....the distortion" that matter. There is so much more to sound than distortion, it's an art!!!!
Last edited:
Your right and wrong.
After a year of just simulating and giving everyone else good designs , the amp's
"musicality" was predicted by the simulator.
Ones that had low THD , dominant 2nd harmonics , and perfect overload behavior became the most ordered group buy PCB's.
With the right models , very predictable perfection.😀
I'm listening to a leach , cordell's textbook amp , others - all on the same output
stage .... a real close call.
These "uber low THD" amps are different than guitar amps , the lines are blurred
at normal output levels , while each has a subtle "character" @ 200W + .
Edit - where you are right ... output stage interaction with the loudspeakers (and the resulting sound)
is not accurately shown in LT ... even with a simulated speaker network.
OS
Last edited:
An electrodynamic speaker is a linear motor. In LTSpice usually the static speaker is modeled, the counter- EMF of the real linear motor is not modeled.
The mechanical motion of the voice coil as described by motional parameters acceleration ,velocity ,excursion is due to the force B*l*i.
However if the amp is a voltage source, the resulting current i is determined by Vamp-Vemf. A linear motor can be simulated in LTSpice, where Vemf is a source controlled by mechanical velocity of the voice coil. To sum it up, a mode of a real loudspeaker ( preferably in enclosure) is not quite easy. But an however inaccurate model has just a voltage source in the simulated speaker network which is controlled by the current through the voice coil. Then one is a bit closer to "reactive distortion". And hence to "sound" of an amp.
It turns out, that the "damping factor" of an amp as voltage source is not only frequency-dependent, but the amp adds harmonics when fed at the output. These are then reactive distortion.
Insofar it is not surprising that controllers of linear motors in industrial applications ( robots and such) are mostly current amps.
Last edited:
The active devices are certainly more likely to influence the sound than the resistors and capacitors. 🙂 The input devices in particular are outside the feedback loop and any nonlinearity in them is uncorrected by feedback. But "more likely" isn't the same as "likely".
Are you sure the customers didn't cheat by looking at your test results? 🙂
I looked at what parts were easily available and what looked like what was likely not to self destruct when I build from his designs (He technically has no customers. He never sold anything himself, just provided some excellent designs).
I personally think when an amp is designed for BJT outputs and built with BJT outputs it will not sound different than an amp designed for FET outputs and built with FET outputs. Mix the design with a different output device and there may be a difference. There's more likely going to be a stability issue though.
I agree. BJT's and mosfets (even vertical fets, which I prefer😀) are not the same and work differently with different properties, even if they are doing the same job. They are not interchangeable to the corresponding supporting circuitry. But if the output devices are coloring the sound then they are not being used properly regardless of the type.😉
For 400W 8 Ohm, peak voltage is almost 80V, then you need +- 90V DC power supply. VDSS max. is same or greater than 200V.
For vertical mosfet: IRFP240 and IRFP9240
For lateral mosfet: ecx10p20r and ecx10n20r, ecw20p20-z and ecw20n20-z
In my power amp circuit all that requires changing is the bias circuit to support the higher voltage for the bias circuit using FETs. Other than that single change FETs and BJT may be interchanged in my circuit. Sorry, this does not really agree with what you wrote exactly. Some minor change was required. About the same output power also.
I like lateral mosfet amplifiers.
I took an old Hitachi datasheet circuit and modified it for my own uses.
I found if I put 12 volt zeners across the mosfet gate and source I can limit output current to about 11 amps.
The laterals also get higher resistance as they get hotter.
This makes laterals almost bombproof.
An externally hosted image should be here but it was not working when we last tested it.
It's not often I agree with Sreten, but here I agree totally.
The performance of the amplifier is down to the DESIGN implementation and the skill of the Designer.
R.Cordell shows both BJT and FET designs and designing.
D.Self for his own reason/s excludes any significant FET DESIGN from his books and other writings.
In my view, as a non designer, both BJT and FET topologies can be made into very high performance amplifiers.
- Status
- Not open for further replies.