5 watts class A is supposedly Heaven. It was found almost regardless of system or room size most people listen typically at 5 watts maximum if sine wave like signals. The extra being for transients that seem to resemble triangle waves ( 30 % of program I think ) . It seems that people buy speakers to suit that need. One could speculated forever about why. Simple to say big room allows big speakers which should be efficient in most cases. One engineer said that the one thing that can not be upgraded is magnets beyond a certain point. High efficiency is a highly desirable feature as the distortion is likely to be lower for an equal wattage. Having used Klipsch Forte in the small room I think it is totally OK to do that. Strangely the Forte still liked big transistor amps. Most oddly Naim NAP 250 ( and two box CD player ). I find that for transients an amplifier can not be too big. That is as long as it satisfies this class A bit or looks class A at 0 to 5 watts.
"Switching" in class AB MOSFETs is very soft, they can and like to be biased hotter than BJTs, and they are fast. All of these factors contribute to good sound.
The lower required "on" voltage is characteristic of lateral MOSFETs, while the verticals typically have "on" voltages in the range of 2-4 volts. The verticals retain the advantages of the laterals, but have higher transconductance and need some modest bias temperature compensation. They also have higher input capacitance, so definitely need buffering after the VAS. They are very fast, so require more care in design and layout to avoid parasitic oscillation.
Cheers,
Bob
I'm one of those strange people,
, who likes the sound "hot", i.e. the amp is running close to flat out - approaching clipping on peaks - a lot of the time. Of course, what that does is nothing more than produce decent sound levels, close to the "real thing"; if the system is working well then what you get is intense, satisfying sound rather than just "loud" hifi. Solo piano recordings tend to be a good guide - the volume needs to be set only a couple of clicks below maximum for the right, in room volume ...
Last edited:
Bob:
this is a small (very - sorry) schematic of the HMA-7500 mk1 from maybe 1980.
the bigger models add more embellishments, but this one uses almost the databook circuit with regulation of the front.
I'm probably showing my age when I say I saw this back in IEE Transactions In Consumer Electronics back in 1979 and contacted hitachi for some samples. I remember building the databook circuit back in 1979 or so; it was too temptingly simple. By the time I got something I was happy with, it was no longer simple.
mlloyd1
this is a small (very - sorry) schematic of the HMA-7500 mk1 from maybe 1980.
the bigger models add more embellishments, but this one uses almost the databook circuit with regulation of the front.
I'm probably showing my age when I say I saw this back in IEE Transactions In Consumer Electronics back in 1979 and contacted hitachi for some samples. I remember building the databook circuit back in 1979 or so; it was too temptingly simple. By the time I got something I was happy with, it was no longer simple.
mlloyd1
Attachments
... because by then I had seen the hafler DH-200 schematic and got that working better
now, i know a lot more now but have less parts and free time ...
oh well ...
mlloyd1
now, i know a lot more now but have less parts and free time ...
oh well ...
mlloyd1
I'm one of those strange people,
Me too. The 5 watt bit is to retain ears.
I have a current dumper/ class G version of the Hitachi in mind for my boss ( he hinted he would like one ). He has a Halcro or whatever and I suspect his Clearlight Audio speakers give it a good workout. It will use 1000 VA 55-0-55 transformer ( I have 2 very high grade ). Exicon 10N/P25 and 250 V cheap bipolar dumpers. I intend to hold the dumpers off until 36 watts 8 R ( 18 4 R ). No emitter resistors except 5A fuses in class C. I suspect 95% of the time the dumpers won't do a thing. As the FET's have 4 amps of spare current the dumpers should have a healthy driver. Complexity is zero. Possible VAS and CCS 2SA1209 and 2SC2911 on +/- 75 V regulated. The dumpers raw DC. Input 2SA1085 whilst still possible. As it will be regulated I might keep the tail resistor. When I did listening tests it was a winner. The capacitors will be 100 V types. Posh ones local to devices ( 4700 uF 100 V 105 C ). On another thread we talked about the series inductance of the first caps to be something we could exploit to make a Pi filter. Thus nasty high ripple cheapies first. Personally I think 2200uf per device enough for the posh ones. Regulator will be simple amplified zeners. If it's good enough for Goldmund it's good enough for me. The VAS will clip before the outputs as that's the way I like it. The FET's I suspect would be fine on raw DC. This will be a Pro Audio style amp with a little bit of Audiophile thinking. C-Audio comes to mind.
And the answer is no. 2: "amplifier has low output power".
The Perreaux was nominally the most powerful amp I've regularly used, but its shortcomings meant that its theoretical capabilities were not actually available. In raw form the power supply would sag and excessively modulate once past a certain volume, audible "compression", as some people call it, could be heard - this is something that I worked on quite a bit, and learnt a few things from over the years. In those days most amps were fairly poor in this respect; there was no point in "approaching clipping", the amp was misbehaving well before such a point.
The Perreaux was nominally the most powerful amp I've regularly used, but its shortcomings meant that its theoretical capabilities were not actually available. In raw form the power supply would sag and excessively modulate once past a certain volume, audible "compression", as some people call it, could be heard - this is something that I worked on quite a bit, and learnt a few things from over the years. In those days most amps were fairly poor in this respect; there was no point in "approaching clipping", the amp was misbehaving well before such a point.
Thanks for that schematic. Don't worry about your age - it is just a number!
I'm showing my age when I say I built a Citation 12 clone in 1971.
Cheers,
Bob
I agree with PMA here.
If the answer is it's a low power amp . . . Then it's going to be clipping most of the time.
The reason for a 200 watter in a home environment is it brings 6-10 dB of extra dynamic range to a system that a small amp cannot.
However, for lots of music, a 15W or 20 W also does fine.
Harkling back to the Otala/Lohstroh 1973 paper, which you have ibviously not read, they quoted authors who claimed that the transition from class A to class be shoud occur 17 dB below nominal power, to be on the safe side.
So bugger any X watts or Y watts, work it out. It's not as if one size fits all.
Amen!
I think I quoted my own experience. A friend and I adjusted my speakers, delivering 92 dB/2.83V/1m, to deliver 96 dB SPL at 2m in my room, using an SPL meter and generated sine wave at 1 kHz. The average power was manageable, but peaks required over 70W into a nominal 8 Ohm load, which is just below the clipping LED indicator on my power amp, a Marantz 170DC, with a very beefed up PSU.
On the other hand, that was the SPL which I for one couldn't bear for longer than 20-30 minutes, that's simply too loud for me.
Nigel,
In a source follower stage the gate capacitance is to some extent cancelled out. This is explained by Cordell on page 7 here:
http://cordellaudio.com/papers/MOSFET_Power_Amp.pdf
Incorrect. The low power amp is not clipping, if you do the maths with a normal sensitivity speaker then the peak amplitudes are fine; and transient clipping, if it occurs, should be inaudible. I always think here of a local, live recording of Beethoven's Ninth I have - goodness me!, in Audacity there are quite a number of points of straight out clipping, surely this will be obvious!! Well, I have not yet heard the slightest thing untoward in the replay - the flat top goes past so quickly my ears never pick it ...
Now, the reality of conventional amplifiers is that, yes, they do sound weak-kneed when you put the accelerator down; hence the general recommendation to have as much power as you can get.
But, my experience is otherwise: a 15, 20W amplifier working properly has no trouble generating intense, "room pressurising" sound, where it completely dominates every other noise source in the area. And this is not "loud hifi" sound, it is convincing music reproduction - very different beasts ...
Or music that is NOT compressed but recorded to allow the 16 (or 24) bits to be used (if digital).
Eg. Arvo Pärt "Sarah Was Ninety Years Old" Trk 3 Miserere (ECM 1430) needs nearly 25db more volume knob than Madeleine Peyroux's "This Is Heaven To Me" to sound "similarly loud". But the loud bits of Miserere still get to close to 0db.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.