Has anybody tried heavy local feedback around a V->I-Stage (with a high-speed op-amp, in a style similar to Fig.11 in this LT AppNote)? I think gm-linearizing/driving a MOSFET in a common-source circuit with a suitable modern op-amp (like THS6012 or TPA6120) might be of benefit in an amplifier concept. I once tested that (only experimentally) with an Aleph-style output stage, I managed to get it stable but didn't go any deeper into that.
Another idea I worked on quite a while ago is a special form of cascoding two matched (MOS-)FETs in a way that one gets an output characteristic that is akin to ultra-linear mode of pentodes, that's why I labeled it "UL-FETRON" (no active feedback involved, this time). In the sims that characteristic looked extremly flat and linear in all dimensions (Id, Vgd, Vgs), a really big linear area for even the worst elliptical loadlines, low dynamic r-ds ad all. I also tested that (with a quite complex op-amp driver stage) with some small signal dual MOSFETS to see if the basic principle worked, and it did. I didn't make serious efforts to build/measure a serious amp stage.
Regards, Klaus
Another idea I worked on quite a while ago is a special form of cascoding two matched (MOS-)FETs in a way that one gets an output characteristic that is akin to ultra-linear mode of pentodes, that's why I labeled it "UL-FETRON" (no active feedback involved, this time). In the sims that characteristic looked extremly flat and linear in all dimensions (Id, Vgd, Vgs), a really big linear area for even the worst elliptical loadlines, low dynamic r-ds ad all. I also tested that (with a quite complex op-amp driver stage) with some small signal dual MOSFETS to see if the basic principle worked, and it did. I didn't make serious efforts to build/measure a serious amp stage.
Regards, Klaus
I think an amp using this feature (and well, why not all amps) should incorporate an Input/Output-Comparator (like older Crown amps had) which could alert the user/tester of any I/O-difference more than 0.1% (or 0.01%). Could even be expanded to a real-time distortion meter...Bob Cordell said:
Regards, Klaus
john curl said:Clipping? The last time I tried to drive my amps into audible clipping, the police came and closed me down!
Also, everyone had left the listening room due to the racket. They would have probably cited me for disturbing the peace, if I was not holding an SPL meter.
Most clipping in early solid state power amps was due to V-I protection circuitry, and it came in early with REAL loudspeaker loads. Tubes can't drive very much, so they have to clip gently.
Now let's use an example: My JC-1 and the WATT 1 loudspeaker (my personal speaker)
First the 1W sensitivity is 91dB, without a subwoofer. So at 10W, I should have 101 dB spl, and at 100W maybe 111 spl, and at 400W 117 spl.
Darn, a lease breaker if there ever was one! And that is for one channel only, without a subwoofer.
The lesson is: 'Real man' amps don't easily clip. That includes Nelson and Charles in this design group. It is only the wimpy solid state power amps of 50W or less that are going to have any significant problems, and I recommmend making a bigger amp, instead of putting a 'band-aid' on a small one.
Hi John,
You have obviously never listened to Rickie Lee Jones' "Ghetto of my Mind" in a situation where you can capture and read out in real time the transient levels going into the speaker. That is why you should have come to our audio workshops. Go to my site at www.cordellaudio.com and read the workshop summary for that work shop.
The people who attended dropped their jaws when they listened to it and saw the peak power levels and realized that the sound level was not driving them out of the room (a hotel room, no less). The speakers we were using were about 85 dB SPL and we were clipping a 250 wpc big amp (not wimpy) on the snare drum thwaks. This is easily demonstrated using the Peak-Average meter I developed for this purpose. Crest factor on well-recorded music can really get you. I'm surprised that you don't seem to understand this. On the other hand, we certainly don't see as much of that kind of music around as we would like, so maybe you just haven't come across any lately.
I am certainly in the bigger amp, higher current camp, but even those will sometimes clip.
Cheers,
Bob
KSTR said:
Hi Klaus,
Yes, these and similar kinds of clipping indicators are fairly popular on pro amps, but almost never found on home audio amps. Also, some clipping detectors are much better than others. Some will only illuminate on fairly gross clipping, while some will catch a clip on a single transient that clips for only 10-50 microseconds. One of the versions of the soft clip circuit that I designed is set up to give a clipping indication whenever it is clipping the signal.
Cheers,
Bob
How important Clipping?
Started by Pete, post #1055,
As often in such a case. I would like to look at the big picture (known in grandspeak as a holistic approach, I believe). This more easily said than done.
Putting clipping behaviour high up on my list of design priorities always seems like making a car twice as heavy with anti-skidding equipment, for use on present-day tarred roads. The trivial solution: You chose the 25-50-100-200W amplifier; keep the level from clipping or choose more intelligently!
As admirably outlined in previous posts amplifiers will clip when overloaded. Semiconductor ones particularly heavily. For some reason I do not feel particularly inclined to design a 100W amp to behave like a 200W when clipping ("soft", that is). Firstly, the volume level difference is only 3 dB, which most will agree is not a dramatic difference. Ergo, loudness difference between a 100W amp and a 200W one, both just prior to clipping is to me unimportant. If it really worries anybody, fork out and go to >400W as John suggested - the decision in terms of cash is a personal one. (Or turn the darn thing down! - similarly the difference is undramatic.)
I have installed an overload indicator in all my amplifiers to assist the listener; the extra cost is negligible and the value apparently reassuring. But all this naturally in the context of proper design basics. A modern amplifier totally "hanging up" is unacceptable; the design techniques are known and they do not present the proverbial "arm-and-leg" complexity/cost wise. Only I stop short of considering circuit gymnastics that will start increasing distortion well before overload level. I am there to listen to and enjoy music within an amplifier's (system's) capabilities, not too much of the time outside it.
Started by Pete, post #1055,
As often in such a case. I would like to look at the big picture (known in grandspeak as a holistic approach, I believe). This more easily said than done.
Putting clipping behaviour high up on my list of design priorities always seems like making a car twice as heavy with anti-skidding equipment, for use on present-day tarred roads. The trivial solution: You chose the 25-50-100-200W amplifier; keep the level from clipping or choose more intelligently!
As admirably outlined in previous posts amplifiers will clip when overloaded. Semiconductor ones particularly heavily. For some reason I do not feel particularly inclined to design a 100W amp to behave like a 200W when clipping ("soft", that is). Firstly, the volume level difference is only 3 dB, which most will agree is not a dramatic difference. Ergo, loudness difference between a 100W amp and a 200W one, both just prior to clipping is to me unimportant. If it really worries anybody, fork out and go to >400W as John suggested - the decision in terms of cash is a personal one. (Or turn the darn thing down! - similarly the difference is undramatic.)
I have installed an overload indicator in all my amplifiers to assist the listener; the extra cost is negligible and the value apparently reassuring. But all this naturally in the context of proper design basics. A modern amplifier totally "hanging up" is unacceptable; the design techniques are known and they do not present the proverbial "arm-and-leg" complexity/cost wise. Only I stop short of considering circuit gymnastics that will start increasing distortion well before overload level. I am there to listen to and enjoy music within an amplifier's (system's) capabilities, not too much of the time outside it.
IMHO, given the fact that the raised distortion appears only during transients (and many listeners might not be able to hear soft-clipping at all at the resulting SPL), I'd say that soft-clipping is worth the tradeoffs from a consumer's standpoint, especially in low power amps. I'd even go further and would like to see some rudementary form of limiting when permanent/repeated overload is detected. This is done in many pro amps (with adjustable thresholds to adapt to speaker power handling) and active studio monitors. I think it's good to have your tweeters protected (at least more likely than with a hard-clipping amp) if your young-one happens to turn up the volume while the amp is unattended/off.
Regards, Klaus
Regards, Klaus
Apology for late reaction
Reacting to the previous post (Glen) I found to my embarrassment that I missed this line of argument of some time ago - thus the heading!
I have great respect for JL-H (have corresponded with him) but - I have done similar measurements, disagreeing with the JL-H graphs given by Jan (his post#1022). This puts me in a position. I am also going to contact Andy_c and hope to be favoured with a copy of the said article. (I doubt that I will be able to find the original locally, where libraries are not well-stocked to say the least).
I sincerely trust that the method used by JH-L will clear up the mystery - and with that, I should shut up until I can present my own findings in graphic form (at present they exist somewhere in tabular form).
I would agree with Bob's explanation to another member, of NFB generating e.g. a 9th product from a 3rd - this is apparent from the mathematics involved. Mercifully this is not of practical implication except for unreal amounts of NFB plus heroically poor circuit design. E.g. (round figures) 0.1% of 3rd will give rise in this way to only 0.001% of 9th. etc.
With this I sincerely do not want to start a shoot-out or anything of similar ilk; I am sure there is a reason for JL-H's findings, as there always was for what he said.
Reacting to the previous post (Glen) I found to my embarrassment that I missed this line of argument of some time ago - thus the heading!
I have great respect for JL-H (have corresponded with him) but - I have done similar measurements, disagreeing with the JL-H graphs given by Jan (his post#1022). This puts me in a position. I am also going to contact Andy_c and hope to be favoured with a copy of the said article. (I doubt that I will be able to find the original locally, where libraries are not well-stocked to say the least).
I sincerely trust that the method used by JH-L will clear up the mystery - and with that, I should shut up until I can present my own findings in graphic form (at present they exist somewhere in tabular form).
I would agree with Bob's explanation to another member, of NFB generating e.g. a 9th product from a 3rd - this is apparent from the mathematics involved. Mercifully this is not of practical implication except for unreal amounts of NFB plus heroically poor circuit design. E.g. (round figures) 0.1% of 3rd will give rise in this way to only 0.001% of 9th. etc.
With this I sincerely do not want to start a shoot-out or anything of similar ilk; I am sure there is a reason for JL-H's findings, as there always was for what he said.
.
Re: How important Clipping?
These days, with modern digital audio sources, it’s easier than ever to set up an amplifier so that it simply cannot be driven into clipping.
I’ve got an old 10W per channel Rotel in my study connected to my computers CD player. The line level input is set so the max p-p output of the CD player produces the amplifiers rated output swing – i.e. maximum volume = 0 clipping.
Connected is a pair of reasonably efficient speakers, and the combo makes plenty of noise. Even for “HiFi” listening in a reasonably large living room, the setup would be perfectly adequate.
I idea that we must all have >400W amplifiers for sufficient sound pressure levels and freedom from clipping is just plain silly.
Cheers,
Glen
Johan Potgieter said:
These days, with modern digital audio sources, it’s easier than ever to set up an amplifier so that it simply cannot be driven into clipping.
I’ve got an old 10W per channel Rotel in my study connected to my computers CD player. The line level input is set so the max p-p output of the CD player produces the amplifiers rated output swing – i.e. maximum volume = 0 clipping.
Connected is a pair of reasonably efficient speakers, and the combo makes plenty of noise. Even for “HiFi” listening in a reasonably large living room, the setup would be perfectly adequate.
I idea that we must all have >400W amplifiers for sufficient sound pressure levels and freedom from clipping is just plain silly.
Cheers,
Glen
Re: Re: How important Clipping?
You do not know what you are speaking about. Bigger room, larger audience and background noise (whisper, chat) and certain kind of music and you get into clipping easily with 200W amp / 90dB speaker. We have enough evidence. (Sessions and signal recording).
G.Kleinschmidt said:
You do not know what you are speaking about. Bigger room, larger audience and background noise (whisper, chat) and certain kind of music and you get into clipping easily with 200W amp / 90dB speaker. We have enough evidence. (Sessions and signal recording).