Most MOS-FET Power amps exhibit "soft clipping" when they become saturated which is less annoying (to the ear) than BJT clipping. I addition MOS-FET's do not suffer from "thermal runaway" and become less conductive if they overheat..protecting themselves. Some of the best "sounding" amps EVER made by man are the basic MOSFET "chip amps" where the FETS are laser-etched and perfectly matched.
Have a couple of amps running tracer matched quads of properly biased & heatsinked Renesas J162/K1058 & Exicon 10N20S/10P20S lateral fets in a minimalistic circuit and... well, not in the market for any amps anytime soon!
Well I helped make a wonderful amp at rockford fosgate only 50 made. Some badged hafler..jf 2000. The fosgates were rf 2000 ..model.. they used v fets...and follow ohms law..
All active devices depend on the circuit design to sound good. Don't focus on what is inside the box! Instead, LISTEN To It and decide on that basis whether it is worthwhile or not. How the results are obtained should not matter to the end user.
I had a Fujitsu 10 Eclipse radio in a car. It used MOSFETs and was an excellent sounding unit, one of the best I’ve had. Even non-audiophiles commented on how good it was. Sadly Fujitsu 10 exited the consumer market as they got so much 4:43 OEM business for auto manufacturers that they had to convert factories for that market, and thus abandoned the consumer aftermarket. Atoll Electronique from France makes MOSFET-based amps and are more like a good tube amp, although they aren’t quite SET in their presentation.
All modern car radios use a version of that same ic.. pioneer called it the pal 002..but it was actually an st 7380 series part.. kenwood used the ta 7384..and alpine used the ta 7560 part..but theyre all very close the same part..made by st parts.. yes. Mosfet with internal crowbar protection.. ill go so far as saying ford used the same output in all of their oem radios..fujitsu and matsushita both agreed and used the part series. IT was huge advancement from the hitachi ha13151 part that used to be the bench mark.. yup..i worked for matsushita and hitachi both.. haha..PAUL.. DO I GET AN INTERVIEW YET?
Hey Paul, I agree with everything you say, but I might add that component matching, ie. amplifiers with different speakers, preamp, dacs, and more. will change the characteristics of the sound. Thanks for the great work and PS Audio.
my answer: if you have a tube preamp and you want a detailed realistic tone just use old telefunken tubes, it's what sound engineers use on mic preamps, if it's too cold and sterile you can try siemens or amperex, enjoy!
also tried all nos clear top rca’s for my cayin ha-1a mkii & noticed something similar with the rca’s.. when compared too new gold lions, alsoo they last much longer
It all depends how each is used in a curcuit there is no aptsolutes which audio has an issue with, blanket statesments that CD sound is inferior to LP's, Tubes sound better than soild state, belt-driven TT sound better than direct drive TT. spekars sound better on spikes, and so. One BIG absolute is your room and it's sound and impact on your whole system. As far as tweaks go anything you do to support your speakers or gear will give you a different sound, be the cost $20.00 of $1,000. 4 Hard apples under your preamp will sound different than the feet on that preamp as an example. Stands? What are they made of, tone arms the same, cartridge all sound different, TT paltter designs and mats sound different, how well the TT is setup and cartridge makes a huge difference. So the only right sound is the one in your room, that you enjoy but it is far from how the master tape sounds in the studio on playback. Enjoy your audio system for what it is and don't let the newest and greatest new thing sawy you, 40 years ago I heard systems that sounded more real then systems today, no fancy power cords, AC wall outlets, power cords or cables. Just gear made to sound musical to the recording and make it enjoyable to listen to at home. That should be the goal.
The best amplifier in my opinion has no sound to it as it simply amplifies the input voltage to the output voltage. This is true for both a preamp and a power amp. Still I agree amps often have a sound to them but is that really what we want?
100% agree with you. The optimal amplifier and preamp only amplifies the sound without changing the sound. Mo sound to it. That is the ideal, but it is almost impossible to achieve.
I have been using an A-B testing box that can switch instantly between amps with a remote control. It's pretty simple to make it a blind test as well. I've been amazed at the difference in the sound of amps. And I am clearly a "tube guy" as the tube amps give a body and life to the soundstage ("bloom" I suppose) that make the experience much more musical to me. This holds true with both my NHT and Martin Logan speakers. By contrast, the SS amps sound more "etched" and 2D. Do I want accuracy or more enjoyable sound? Is "accuracy" 100% relevant when the recording setting and equipment is already going to color the sound when contrasted with a live performance anyways?
@@flargosa But some amps are more pure than others. I don’t want tubes in my signal path. Adding a tube to make the sound become more musical with some degree of added even harmonics is not pure amplification. Adding a tube to get the effect of a softer nicer sounding clipping can make some sense but many decades ago NAD showed the world this can be done also with transistors and more reliable. Adding tubes because they create a planned obsolesense could be a real business case, and for example, serious profit has been made on amps where tube replacement could cost you 1000s of Dollars over the life of the amp. Besides tubes change sound as they warm up and wear out. So why in 2024 do they still exist in new design???
@ThinkingBetter First of all, let consumers decide what they want. If it's tube amplifiers, there will be customers. Secondly, there is no pure signal amplification. Even very flat instrumentation amplifiers do have a certain distortion figure. These are designed with a specific bandwidth and amplification factor. But this is not for audio. In every HiFi component you have amp stage, mostly small signal amps. Sometimes differential, sometimes non symmetrical. Along the audio path signals undergo a lot of amp stages, feedback loops, voltages and current amplification stages. A good designer knows all that and can predict & simulate. A tube stage itself is only one part in a chain of audio (analog domain) processing. Never forget your speakers at the very end. Their distortion figures are magnitudes higher then even lover class amps. It is sometimes a sonic profile (THD/THD+N)transient response, multitone response & distortions and many more. Rather complex as our hearing is way more advanced than any measurementvmicrophone etc. You might be surprised if all is super linearvwithout distortion incl. totally flat speakers how cold and sterile it might sound. Music then stripped of its soul and magic. And of course your listening room. We all hear different.
A BJT has 3 materials or segments in a row doped differently. N materials (like silicon) have excess electrons. P materials have excess "holes" or a lack of electrons. Usually either NPN, (N materials connected to a P material connected to an N material) or PNP. A barrier voltage is created where the N and P materials meet stopping electrons from passing between them. A basic design is a voltage source, electrons are connected to one end and the load to the other end. But no electrons flow because of the blocking alternate material between the two common ones. But when a small voltage is applied to that center alternate doped material, the barriers between the segments is overcome and a larger amount of electrons flow from the PS source through all three materials and to the load. The amount of electrons allowed to flow is controlled by the amount of voltage applied to the center barrier. A varying signal to the middle is amplified. FETs are like hoses. The water/ electrons will flow from the input to the output when first connected. No outside signal needed. If the main material is N, a P material is "wrapped" around the center and like a pliers squeezing the middle of the hose, water/ electron flow is restricted in ratio to the pliers pressure until it is shut off. MOS - Metal Oxide Semi Conductor. It describes what materials are used with some additional contraction. So a BJT has to be turned on before it starts conducting. Depending on materials used the input signal needs to be higher than 0.7Vdc before it turns on. Which is why Class AB is so popular. It runs the BJTs with a slight bias to keep them on all the time when no signal is applied. While FETs are always on and need to be turned off by the input signal.
@@pcm9969 Regurgitating what I learned many decades ago but got shaken out of the cobwebs by the question! I'm known for random neuron firings being triggered.
@@glenncurry3041 Amazing Glen! That is out of my league though😅 According to Paul it looks like JFETs are sometimes chosen over BJTs at the input stage after the tubes for less 'sterile' sound and both handle voltage. MOSFETs on the other hand are at the output stage and handle current. Is this correct?
@@wyup Remember FET and BJT are design types of solid state devices. Each can be made of a variety of materials. Paul said he used tube stages at the input. I believe he said they use BJTs as driver stage. They can be set into a well controlled voltage output mode. Then FETs for the output for the current./ load matching. He then mentions in other designs replacing BJT stages when they are sounding to sterile with FET stages. Now going back to materials used. MOS (Metal Oxide Semiconductior) designed FETs tend to have superior high current capabilities. JFETs compromise current handling for things like bandwidth, size, .... Gallium Nitride (GaN) is catching on for Class D systems.
Do not use tubes they are just so WEAK and go wrong all the time. Their power output is just almost nothing, meanwhile Transistors are STRONG and can output enormous power!💪
That was a ignorant statement. If handled correctly and used for the right purpouse tubes have a sound quality the no transistor can match and they will last for at least 10 years. PS Audio's BHK amps are excellent example of using each component where it have it's best qualities and don't need to have a task they are not well suited for. Tenor Audio, Riviera Labs, Lamm Ind., Counterpoint, NYAL, David Berning, Conrad Johnson. There are quite a few companies that designed great products through combining tubes and transistors.
@@flex-cx9biI’m not saying that tubes are useless but there is mosfet and fets amplifiers that were able reproduce most if not all of the characteristics of tube amps, it’s all depends on the circuit (using cascode of fets for exemple). The price of tubes isn’t really worth their qualities aside from very special installations and the use of transformers is redundant and introduce distortion
@@omnirath Depends on what you sonic preference are. If you look for natural harmonics of the distortion pattern you can get a bit of that with single end cascoded jfets. Mosfets and BJT can not get it right regardless what you do with them. Tubes have some virtues that is unique if not used in a differental gain stage or in a push pull configuration. As a gain stage I would say that nothing can beat a single ended tube. They just breath musicality in a way no transistor can do. Listen to Riviera Labs hybrid amplifiers over UA-cam and compare to the best transistor amplifiers. To my ears Riviera Labs still wins. For me as a designer of audio equipment I would say that it is the tube that brings the magic sauce to the mix of tube, BJT and Mosfets used in their products. I have studied extensively on how we percieve the harmonic products of distortion and it's essential that you get it right if you want a neutral and natural sound from your amplifier. Tubes have the best chance of keeping the harmonics intact as they where recorded.
@@flex-cx9bi listen to a Le Monstre or look at the harmonic content of the distortion there’s no audible difference, there have been quite a lot of improvement since with those kind of amps too but the choice of the transistors and the architecture of the circuit is what’s important
Bjts have thermal runaway issues..and they get more apparent as imp gets lower. Vfets ..matter of fact ALL N CHANNEL is the best output stage..but many fan cooled pro amps use triple darlington bjts in output because of their shear ability..but the temp must be controlled and I mean every device the same temp..or they domino. Kaboom.
Most MOS-FET Power amps exhibit "soft clipping" when they become saturated which is less annoying (to the ear) than BJT clipping.
I addition MOS-FET's do not suffer from "thermal runaway" and become less conductive if they overheat..protecting themselves.
Some of the best "sounding" amps EVER made by man are the basic MOSFET "chip amps" where the FETS are laser-etched and perfectly matched.
Have a couple of amps running tracer matched quads of properly biased & heatsinked Renesas J162/K1058 & Exicon 10N20S/10P20S lateral fets in a minimalistic circuit and... well, not in the market for any amps anytime soon!
Well I helped make a wonderful amp at rockford fosgate only 50 made. Some badged hafler..jf 2000. The fosgates were rf 2000 ..model.. they used v fets...and follow ohms law..
All active devices depend on the circuit design to sound good.
Don't focus on what is inside the box!
Instead, LISTEN To It and decide on that basis whether it is worthwhile or not.
How the results are obtained should not matter to the end user.
I had a Fujitsu 10 Eclipse radio in a car. It used MOSFETs and was an excellent sounding unit, one of the best I’ve had. Even non-audiophiles commented on how good it was. Sadly Fujitsu 10 exited the consumer market as they got so much 4:43 OEM business for auto manufacturers that they had to convert factories for that market, and thus abandoned the consumer aftermarket.
Atoll Electronique from France makes MOSFET-based amps and are more like a good tube amp, although they aren’t quite SET in their presentation.
All modern car radios use a version of that same ic.. pioneer called it the pal 002..but it was actually an st 7380 series part.. kenwood used the ta 7384..and alpine used the ta 7560 part..but theyre all very close the same part..made by st parts.. yes. Mosfet with internal crowbar protection.. ill go so far as saying ford used the same output in all of their oem radios..fujitsu and matsushita both agreed and used the part series. IT was huge advancement from the hitachi ha13151 part that used to be the bench mark.. yup..i worked for matsushita and hitachi both.. haha..PAUL.. DO I GET AN INTERVIEW YET?
Hey Paul,
I agree with everything you say, but I might add that component matching, ie. amplifiers with different speakers, preamp, dacs, and more. will change the characteristics of the sound.
Thanks for the great work and PS Audio.
But how?
my answer: if you have a tube preamp and you want a detailed realistic tone just use old telefunken tubes, it's what sound engineers use on mic preamps, if it's too cold and sterile you can try siemens or amperex, enjoy!
also tried all nos clear top rca’s for my cayin ha-1a mkii & noticed something similar with the rca’s.. when compared too new gold lions, alsoo they last much longer
It all depends how each is used in a curcuit there is no aptsolutes which audio has an issue with, blanket statesments that CD sound is inferior to LP's, Tubes sound better than soild state, belt-driven TT sound better than direct drive TT. spekars sound better on spikes, and so. One BIG absolute is your room and it's sound and impact on your whole system. As far as tweaks go anything you do to support your speakers or gear will give you a different sound, be the cost $20.00 of $1,000. 4 Hard apples under your preamp will sound different than the feet on that preamp as an example. Stands? What are they made of, tone arms the same, cartridge all sound different, TT paltter designs and mats sound different, how well the TT is setup and cartridge makes a huge difference. So the only right sound is the one in your room, that you enjoy but it is far from how the master tape sounds in the studio on playback. Enjoy your audio system for what it is and don't let the newest and greatest new thing sawy you, 40 years ago I heard systems that sounded more real then systems today, no fancy power cords, AC wall outlets, power cords or cables. Just gear made to sound musical to the recording and make it enjoyable to listen to at home. That should be the goal.
The best amplifier in my opinion has no sound to it as it simply amplifies the input voltage to the output voltage. This is true for both a preamp and a power amp. Still I agree amps often have a sound to them but is that really what we want?
100% agree with you. The optimal amplifier and preamp only amplifies the sound without changing the sound. Mo sound to it. That is the ideal, but it is almost impossible to achieve.
I have been using an A-B testing box that can switch instantly between amps with a remote control. It's pretty simple to make it a blind test as well. I've been amazed at the difference in the sound of amps. And I am clearly a "tube guy" as the tube amps give a body and life to the soundstage ("bloom" I suppose) that make the experience much more musical to me. This holds true with both my NHT and Martin Logan speakers.
By contrast, the SS amps sound more "etched" and 2D. Do I want accuracy or more enjoyable sound?
Is "accuracy" 100% relevant when the recording setting and equipment is already going to color the sound when contrasted with a live performance anyways?
all amps have a sound signature.
@@flargosa But some amps are more pure than others. I don’t want tubes in my signal path. Adding a tube to make the sound become more musical with some degree of added even harmonics is not pure amplification. Adding a tube to get the effect of a softer nicer sounding clipping can make some sense but many decades ago NAD showed the world this can be done also with transistors and more reliable. Adding tubes because they create a planned obsolesense could be a real business case, and for example, serious profit has been made on amps where tube replacement could cost you 1000s of Dollars over the life of the amp. Besides tubes change sound as they warm up and wear out. So why in 2024 do they still exist in new design???
@ThinkingBetter First of all, let consumers decide what they want. If it's tube amplifiers, there will be customers. Secondly, there is no pure signal amplification. Even very flat instrumentation amplifiers do have a certain distortion figure. These are designed with a specific bandwidth and amplification factor. But this is not for audio. In every HiFi component you have amp stage, mostly small signal amps. Sometimes differential, sometimes non symmetrical. Along the audio path signals undergo a lot of amp stages, feedback loops, voltages and current amplification stages. A good designer knows all that and can predict & simulate. A tube stage itself is only one part in a chain of audio (analog domain) processing. Never forget your speakers at the very end. Their distortion figures are magnitudes higher then even lover class amps. It is sometimes a sonic profile (THD/THD+N)transient response, multitone response & distortions and many more. Rather complex as our hearing is way more advanced than any measurementvmicrophone etc. You might be surprised if all is super linearvwithout distortion incl. totally flat speakers how cold and sterile it might sound. Music then stripped of its soul and magic. And of course your listening room. We all hear different.
Tom in the polite state of Wisconsin waves that red flag component stuff in front of angry audiophile bulls. 🙃
Motorola bipolar junction transistors 😊
2n6400 series was very good indeed..yup bernies dad designed them..
I'm still confused between MOSFET, JFET and BJT!
A BJT has 3 materials or segments in a row doped differently. N materials (like silicon) have excess electrons. P materials have excess "holes" or a lack of electrons. Usually either NPN, (N materials connected to a P material connected to an N material) or PNP. A barrier voltage is created where the N and P materials meet stopping electrons from passing between them. A basic design is a voltage source, electrons are connected to one end and the load to the other end. But no electrons flow because of the blocking alternate material between the two common ones. But when a small voltage is applied to that center alternate doped material, the barriers between the segments is overcome and a larger amount of electrons flow from the PS source through all three materials and to the load. The amount of electrons allowed to flow is controlled by the amount of voltage applied to the center barrier. A varying signal to the middle is amplified.
FETs are like hoses. The water/ electrons will flow from the input to the output when first connected. No outside signal needed. If the main material is N, a P material is "wrapped" around the center and like a pliers squeezing the middle of the hose, water/ electron flow is restricted in ratio to the pliers pressure until it is shut off. MOS - Metal Oxide Semi Conductor. It describes what materials are used with some additional contraction.
So a BJT has to be turned on before it starts conducting. Depending on materials used the input signal needs to be higher than 0.7Vdc before it turns on. Which is why Class AB is so popular. It runs the BJTs with a slight bias to keep them on all the time when no signal is applied. While FETs are always on and need to be turned off by the input signal.
@@glenncurry3041 Great explanation! You refreshed what I learned (and since forgot!) what I learned decades ago in electronics class. Thank you.
@@pcm9969 Regurgitating what I learned many decades ago but got shaken out of the cobwebs by the question! I'm known for random neuron firings being triggered.
@@glenncurry3041 Amazing Glen! That is out of my league though😅 According to Paul it looks like JFETs are sometimes chosen over BJTs at the input stage after the tubes for less 'sterile' sound and both handle voltage. MOSFETs on the other hand are at the output stage and handle current. Is this correct?
@@wyup Remember FET and BJT are design types of solid state devices. Each can be made of a variety of materials. Paul said he used tube stages at the input. I believe he said they use BJTs as driver stage. They can be set into a well controlled voltage output mode. Then FETs for the output for the current./ load matching. He then mentions in other designs replacing BJT stages when they are sounding to sterile with FET stages. Now going back to materials used. MOS (Metal Oxide Semiconductior) designed FETs tend to have superior high current capabilities. JFETs compromise current handling for things like bandwidth, size, .... Gallium Nitride (GaN) is catching on for Class D systems.
Do not use tubes they are just so WEAK and go wrong all the time. Their power output is just almost nothing, meanwhile Transistors are STRONG and can output enormous power!💪
That was a ignorant statement.
If handled correctly and used for the right purpouse tubes have a sound quality the no transistor can match and they will last for at least 10 years.
PS Audio's BHK amps are excellent example of using each component where it have it's best qualities and don't need to have a task they are not well suited for.
Tenor Audio, Riviera Labs, Lamm Ind., Counterpoint, NYAL, David Berning, Conrad Johnson. There are quite a few companies that designed great products through combining tubes and transistors.
@@flex-cx9biI’m not saying that tubes are useless but there is mosfet and fets amplifiers that were able reproduce most if not all of the characteristics of tube amps, it’s all depends on the circuit (using cascode of fets for exemple). The price of tubes isn’t really worth their qualities aside from very special installations and the use of transformers is redundant and introduce distortion
@@omnirath Depends on what you sonic preference are.
If you look for natural harmonics of the distortion pattern you can get a bit of that with single end cascoded jfets. Mosfets and BJT can not get it right regardless what you do with them. Tubes have some virtues that is unique if not used in a differental gain stage or in a push pull configuration.
As a gain stage I would say that nothing can beat a single ended tube. They just breath musicality in a way no transistor can do.
Listen to Riviera Labs hybrid amplifiers over UA-cam and compare to the best transistor amplifiers. To my ears Riviera Labs still wins. For me as a designer of audio equipment I would say that it is the tube that brings the magic sauce to the mix of tube, BJT and Mosfets used in their products.
I have studied extensively on how we percieve the harmonic products of distortion and it's essential that you get it right if you want a neutral and natural sound from your amplifier.
Tubes have the best chance of keeping the harmonics intact as they where recorded.
@@flex-cx9bi listen to a Le Monstre or look at the harmonic content of the distortion there’s no audible difference, there have been quite a lot of improvement since with those kind of amps too but the choice of the transistors and the architecture of the circuit is what’s important
IGBT has left the chat.
Bias the MOSFETs into Class A.
Bjts have thermal runaway issues..and they get more apparent as imp gets lower. Vfets ..matter of fact ALL N CHANNEL is the best output stage..but many fan cooled pro amps use triple darlington bjts in output because of their shear ability..but the temp must be controlled and I mean every device the same temp..or they domino. Kaboom.
Which is all very easy to do and has been solved many years ago.
@@hugobloemers4425 explain..please..what has been resolved years ago..haha..are you saying im old? My friend patented the all n channel output design.
@@davidstevens7809 The thermal run away issue can be mitigated with good engineering practices, since lets say the 60's
@@hugobloemers4425yes emitter bypass cap..yes.. they wont follow ohms law. Im the master..
Hi paul! Your first viewer in these video
ua-cam.com/video/gCCD40eB-cU/v-deo.html
Some amplifiers have LOTS of small transistors and capacitors and resistors. Is there some kind of equalizer to adjust the frequency? Cheating
✍