Hi Es,
If JC is not listening or does not want to repeat himself, I seem to recall he said 9pair.
That 10W of ClassA in an optimised ClassAB output stage implies Iq~=800mA. About 90mA per device. Re=0r27 would achieve this with Vre=24mV. If one wanted to use Vre=19.5mV then Re=0r22 would get close.
If JC is not listening or does not want to repeat himself, I seem to recall he said 9pair.
That 10W of ClassA in an optimised ClassAB output stage implies Iq~=800mA. About 90mA per device. Re=0r27 would achieve this with Vre=24mV. If one wanted to use Vre=19.5mV then Re=0r22 would get close.
SY said:
Huh? When on earth did I say that? Firstly, I specifically cited 20Hz through to 20kHz THD as a general criterion to generalise on.
I specifically mentioned the importance of other distortion mechanisms such as TIM, IMD, etc, but simply pointed out the fact that all these distortion mechanisms are related to a degree and that 20Hz-20kHz THD is a rather good figure of merit.
As for bandwidth, if an amplifier has very low THD-20, then the bandwidth is clearly sufficient.
And of course things such as overload recovery and clipping characteristics are important. That is which I deliberately said a well designed SS amp.
What do you expect a single post from me to summarise my views to be? A 50 thousand-word essay concisely covering every important detail of SS amplifier design?
Cheers,
Glen
SY said:
Well was it unreasonable for me assume that the reader would infer that the tests required would be those necessary to confirm wether or not the amplifier met or exceeded the general performance criterion I specified?
I mean, your question was a very vague one. There are so many different amplifier/speaker/source topologies and designs with mild to drastically different technical requirements for harmonious operation that I can’t answer it in anything other than general terms. Also, there are other people here with more experience and expertise than me in this regard too, who could probably answer your question better than I could. Am I specifically under interrogation here?
Cheers,
Glen
Nope, but since you're throwing around figures of merit and I'm something of a measurement geek, I want to see if we're on the same page regarding measurement regimes or where we might differ. You certainly have the option to punt and not answer. But I think it would be useful and educational if you did.
Fair enough question!
Assume that you want the amp to be a voltage source and the output to replicate a scalar times the input over some frequency range as closely as possible, i.e., not an effects box. And assume that the amp will be hooked up to some unknown set of speakers in unknown living rooms run at variable levels, i.e., real-world hi-fi usage, not sound reinforcement, musical instrument, shake tables, or dummy loads.
I define "sufficient" in the usual sense, some complete set of measurements whereby the amp can be substituted for any other amp with similar or better measurements under the same test regime and the differences will not be audible to a critical listener.
Assume that you want the amp to be a voltage source and the output to replicate a scalar times the input over some frequency range as closely as possible, i.e., not an effects box. And assume that the amp will be hooked up to some unknown set of speakers in unknown living rooms run at variable levels, i.e., real-world hi-fi usage, not sound reinforcement, musical instrument, shake tables, or dummy loads.
I define "sufficient" in the usual sense, some complete set of measurements whereby the amp can be substituted for any other amp with similar or better measurements under the same test regime and the differences will not be audible to a critical listener.
SY said:
OK, so you want to measure the amplifiers damping factor with a reactive load throughout a specific frequency range.
Firstly, I’d use my homebrew AD9831 DSS based PC driven audio sweep generator and 24 bit analogue acquisition board to measure and plot the amplifiers frequency/amplitude response over the desired frequency range into a purely resistive load of 16 ohms for a reference.
I’d then repeat the above into a 4 ohm purely resistive load. That would give me two plots from which the damping factor can be computed.
I’d then do it all again into a reactive dummy load of the desired impedance and correlate all three plots.
Then plug in amplifier-under-test #2 and repeat all of the above.
The results between both amplifiers can then be compared to see it they are sufficently similar in performance in this regard.
Usually, this isn't a significant design problem in conventional SS design, due to the large amounts of negative feedback used.
BTW, Mr Moderator, Sir. Isn’t all this deviating a bit from the BJT Vs MOSFET topic of discussion??
estuart said:
Hi Pavel,
I did have a look at that amp and indeed, the thd figures are pretty good. But I'm sorry to say, I'm not impressed. IMHO it is a rather 'insipid' design, as it is quite easy to obtain such a low thd by reverting to a power hungry class-A stage (hence it's named ExtremA?) and using an inverting input stage. Clearly, the designer of this amp wasn't challenged by the hurdles that other designers are willing to overcome.
Analogue to incandescent lamps, there should be a ban on class-A amplifiers and designers of such heaters should be punished by cutting off one ear.
To all,
I'm looking for vertical fet's like the 2SK1530/2SJ201 pair, but with a much lower Cgd. Any suggestions?
Cheers,
Yikes! Well that leaves me earless
– I’ve got two class A amplifiers on the go. One is a fully symmetrical SS, 512W rms per channel into 4R in pure-class A design with Toshiba BJT’s, error correction and tracking supply rails for the two class A output stages used per channel. Dissipates 350W per channel at idle and has so far gone through about 150 design permutations . I haven’t finished building it yet and it looks like I’m going to have to take out a second mortgage on my house in order to buy a distortion analyser good enough to measure it.
Class A may be an easy way to low distortion, but despite this and the low efficiency, it’s superiority in linearity over class AB obviously can’t be denied.
My other design is just a simple little one that I just knocked up for a bit of fun, for driving an old pair of efficient PYE speakers I’ve got in my study. I noticed that you do spice simulations by request – care to do a despicable class A? Pretty please?
My crappy simulator won't go under 0.001% THD and I couldn't be bothered learing to use LT spice at the moment.http://users.picknowl.com.au/~glenk/classa.gif
Incidentally, you can delete d3 and d4.
Cheers,
Glen
The problem with Chas. Hansen's amp measurements is that it doesn't have any global negative feedback. Can any of you do as well without any global negative feedback? I really doubt it. You use it like a crutch in order to make your amps measure OK.
In reality, I doubt that adding negative feedback really adds to sound quality. That is why many serious audio designers, after designing high feedback power amps over the decades, make an effort to make amps with limited or no global feedback. One serious exception is Halcro, and I will give them a gold star for the best measurements in commercial amps today. Most of you here are still amateurs, please keep that in mind when criticizing the efforts of others.
In reality, I doubt that adding negative feedback really adds to sound quality. That is why many serious audio designers, after designing high feedback power amps over the decades, make an effort to make amps with limited or no global feedback. One serious exception is Halcro, and I will give them a gold star for the best measurements in commercial amps today. Most of you here are still amateurs, please keep that in mind when criticizing the efforts of others.