I know there have been a couple of threads here on the New Class A Technics, those with synchro bias chips sets. I can't comment on those as I've never heard one. But I would like to talk for moment about the Class AA Technics that use the Matsushita Bridge, also referred to as a Wheatstone coupling. First of all, I don't really understand exactly how these work other than it uses a class A voltage stage with a current compensating A/B stage. If I understand correctly these are fully discrete components as well. Having two of these in my stable now I can say these are some of the best amplifiers I've owned (not that I've owned a shit load of gear). They sound like a bottomless well of current, and dare I say tubey. Some may remember this thread where I lamented the need for more power for some four ohm Snell Ci's, http://www.audiokarma.org/forums/index.php?threads/amp-selection-for-some-snell-cis.654021/, I did eventually purchase an Adcom GFA555 (325w into 4 ohms) and while it sounded good, it wasn't quite the improvement I'd hoped for. Enter the Class AA Technics with the Matsushita bridge, I threw this in on a whim just to test it out and make sure functioned correctly, and knew after the first song the Adcom was a goner. The bottom end in particular filled out like I'd never heard before. I've since been running my 2.7 Magnepans on this integrated and it sounds wonderful, I was a bit worried as I've heard nothing but that Magnepans need tons of current, but this amp handles them with ease, rarely does the volume go past nine o'clock. Output is 100w into 8 ohms and 150 into 4 ohms. THD for these amps is extremely low as well at .003% So as a heads up, if anyone happens to run across one at thrift, Craig's, or if you see these on the bay cheap, don't hesitate as these are really nice sounding units that can handle mildly difficult loads easily. These are not BPC as the front is extruded aluminum. They also support MM/MC phono as well. Hopefully SVI12004a or Macphantom will chime in as I'm sure they may have a bit more info on these nice sounding amps. Here's a couple of pics for anyone who may have not seen the inside of one.
SVI has probably talked about it at some point. http://audiokarma.org/forums/index....=Class+AA&t=post&o=relevance&c[user][0]=66945
Thanks Glen, reading some of SVI's older posts confirms my opinion that these are a whole different animal than the computer drive new class A Technics.
I have an Su-v85a, too! I'm sure someone poured a drink into the back chassis as I can see the arching traces from a massive short. (I guess she got pissed off or he was beaney) Saving it hopefully to get to it as a near full rebuild on the main board. The amp AA as referred to. usually extremely low THD% as you noted and many of these models were avail in px mil. stores in the EU and pacific but not many show up in the usa. They have the conversion switch to 220 to 120 so mine came back from over seas. Big well built chassis and covers.. another reason I'm holding onto it. The front end selection still functions but you can imagine all the drivers and op amps will need replacing. Oh, when I first got it two of the 3 tone controls were popped! literally the shafts were out of the pot body and cracked one of the knobs. took a hell of a beating. Hell of a lot of work ahead of me.
I've already had the main board out once as well as the tone board that's easy but tone board needs a rebuild, too. The OP unit looks nice inside.. mine is a heart breaker. I have chassis rust from the arching, top cover needs paint, not even sure if I fixed the pots re assembling them the dmm showed good though. The bom isn't going to be cheap.. talking a lot of parts and work for a resale value? not sure it's worth it.
That's too bad Binkman, my second one arrived with the bass pot apart, almost as if the crimp on the pot assembly was never tight enough. But it was an easy repair, as tone board just pops out by pushing on a a couple tabs. These are very easy to work on with lots of room under the hood, but it does sound like yours needs quite a bit of attention. I'd love to find a SU-v90d, which is pretty much the same as the v85a but with optical inputs.
Hey Tony! no crap on the bass pot apart? Optical inputs? got a denon av and a Yammie with optical inputs and must say at the 15 foot cable I have optical is nice and clean. (I got the longer cable for testing Av's but nice option for a bench to a second bench. The cosmetics on 86a isn't a big deal. Rail v's are common. Been considering every time I look at this unit redesign it a bit. there is room in this box for a tube amp or a pair or turn into a pre. add a dac etc.. it's a big frigging box and enough room for a lot of stuff. even have a sony monster se transformer.
Fair enough, your wish is my command! Warning: below follows a long post! I actually had to split it up into two because of the character limit. Class AA is indeed a (Matsushita invented) name for an amplifier topology consisting of both a class A stage as a voltage control amplifier as well as a class B stage for the current drive—plus a so-called "Class AA Bridge" which links together the two amp stages. Matsushita had been in a long dispute with a Mr. Sandman who claimed to have invented that principle before, but eventually they somehow still managed to keep that name. I'm not interested in that lawsuit stuff, so I don't know the details of what happened there. Here's something about the class A and B topologies in general; if you're familiar with that (and don't want to correct me), you can skip that part: The main differences between the generally known amplifier topologies is the working range of the transistors used: Class A stages try to make use of the linear working range of the transistor only. The input signal (AC) is biased (alternating DC) so that there always is a voltage at the transistor base (resp. gate in the FET case). Consequently, the transistors are always turned on, even if the signal is completely flat (i.e., the audio source is silent). This means that the transistor on average always has to pass a certain amount of current, but not all. Therefore, it always acts as a resistor and thus dissipates a whole lot of heat, which is why its efficiency is so terrible. Class B stages in turn don't apply any bias at all to the transistors. Instead, the positive and negative half-waves of the signal is split up and amplified by two separate half-stages. Therefore, if no signal is coming through, all transistors are completely switched off and don't dissipate any heat. Likewise, if the instantaneous signal only contains positive half-waves the "negative" part of the amp stage does not have anything to do and is switched off as well. The downside to this principle is the moment of switching (that is, when the signal changes from positive to negative). Transistors are no perfect devices and they need a certain minimum voltage to turn on (around 0.75V for silicon and 0.3V for germanium). In addition, they may tend to overshoot for a small moment when turned on. These effects are the root cause for the infamous crossover distortions. Class AB is a compromise between the two: here, the transistors are biased just a little bit so that they always stay on, but convey only a very tiny amount of current if the instantaneous signal is silent. This way, switching distortions are mostly avoided. Class AB is probably the most often used amplifier topology: it's simple (requires a class B design with a few extra resistors and diodes) and effective. At the end of the 70s companies suddenly realised that they kind of need an additional selling point after having sold millions of dull class AB amps, and they all kind of simultaneously attempted to reduce distortions in a more intelligent way. Matsushita's answer to this was called "New Class A", and the idea behind it was the same like in just about all other fancy topologies as well, including JVC's Super A, Dual Class A, Denon's Optical Class A, etc. By adapting the idling current according to the audio source, the transistors received a variable bias which shifted them into the linear working range just if needed. The way this was achieved differed from company to company (Denon e.g. used an opto coupler in addition to many other things), but it worked pretty well. Probably the most famous New Class A device was the SE-A3: Most of these older ones are absolutely magnificent amps! New Class A was later on used in parallel with Class AA, but only for lower end and midi amplifiers. Later units featured a so-called "Computer Drive" chip, which is nothing more than a logical circuit adjusting the bias depending on the temperature and the input signal amplitude. They also do some kind of pre-heating (applying a high current to the output transistors to more quickly bring the into a thermally stable condition). Other than that these devices have the exact same principle as regular New Class A amps. Next came Class AA – the main subject of this thread. This topology does not have anything to do with New Class A, except that it also tries to minimise (resp. eliminate) crossover distortions! Class AA came up in the mid 80s (the first amp using this topology being the Technics SE-A100). The way class AA works is both quite simple and very effective: two amplifier stages share the workload. The main powerhouse, responsible for high output currents, is a class B stage ("Current Drive"). Since class B stages tend to produce increased distortions when the signal switches polarity, or in general if the signal is very faint (during quiet passages), the class A stage carries over and amplifies what the class B stage is unable to without producing distortions. One could say, the class A stage keeps up the voltage below the silicon's threshold of 0.75V and is hence called "Voltage Control" amplifier. If that's just a mnemonic or not I don't know. One more thing: the class AA bridge. This is basically nothing more than a resistor network linking together the voltage control and current drive stage. In higher ended units like the SE-A100, the class AA bridge may further contain some capacitors and inductors. To put it all together, this is what a class AA stage eventually looks like (picture taken from the SE-A100 service manual): Note the class AA bridge (rhombus-shaped resistor network in the right part of the schematic). Here's an inside shot of the SE-A100: (source: http://vintagetechnics.co.uk/controlpower/sea100.htm) See the two big transformers, the outer heatsinks, adjacent caps and two of the four riser cards? These are the guts for the current drive stage. See the smaller heatsinks (and smaller caps behind it) in the centre of the unit, the smaller transformer in the middle as well as the two central riser cards? You guessed it: that is the voltage control stage. Class AA was used up until the break of Technics (around 2004), the last unit featuring that topology being the SE-A7000 power amplifier: (picture by Werner; source: http://www.hifi-forum.de/viewthread-84-19003.html) However, one additional thing that sets Class AA apart from New Class A is that it was also used in CD-players, tape decks, pre-amplifiers, DATs, and tuners! Ever came across a Technics device with that "Class AA" badge on the front? That's no lie, they really used that principle even in the audio sources! Example: here's a subset of the Technics SL-P770 CD-player's service manual: Notice something? You definitely should. Why they went along and put that into a low-power, low-voltage unit like a CD player is beyond my imagination. Probably marketing. [end of part one]
[part two] Okay, while we're at it, there was one more amplifier topology used (and actually invented) by Matsushita: Class A+. The holy grail. Why that? Well, because Class A+ has been used in one device only. One. And that is the famous SE-A1. Time-wise Class A+ is some kind of a predecessor of New Class A, but it works entirely differently. In Class A+ the major amplification job is actually done by a class A output stage. However, here comes the ingenious trick: the power source for the class A stage is a class B stage. This stage is a floating power source, which does the following: It keeps the class A stage's output transistor's collector-emitter voltage constant However, during silent passages it minimises the available current to the class A stage's power transistors This way, the transistors can constantly be biased (and hence don't produce crossover distortions) but due to the current limitation there is hardly any heat being produced in the transistors. This allowed Technics engineers to make the SE-A1 produce an enormous amount of output power (350W at both four and eight ohms), while having the output stage always stay in class A mode. SE-A1 inside: (picture by Werner) Have you ever wondered why the SE-A1 has (per channel) two 22'000µF capacitors rated for 105V *and* two additional 100'000µF caps rated at 10V? Well here's your answer: the two high voltage caps are for the class B stage and the two small, big capacity ones (combined in one aluminum can) are for the class A stage. Class A+ was a very intelligent solution, but also very expensive to make (I reckon), which is probably why New Class A eventually took over. Now, summing up you're probably wondering about one or two things: which one sounds best? The answer: you cannot tell! In fact, you could make a class B amp that sounds just as good (resp. accurate) as a class AA device. If that's the case, why the different topologies? Well, for one part this surely was a marketing strategy. For the other, it's nice to test out different ways of doing stuff, and each of the mentioned topologies above has their own advantages and disadvantages. Which would I recommend? Once again: this question cannot be answered that easily. There are both New Class A and Class AA devices which are outstandingly great, and there are units in both cases which are of lower build quality. Very often, cost-cutting was achieved in both cases by implementing that topology in one or two integrated circuits (SVI/STK/RSN ICs). They don't necessarily sound worse than discrete amps, the only problem they have is that due to their tight package they are prone to overheating. Most famous example: the Technics SE-A900S and SE-A1000 power amps (Class AA). Using two ICs for voltage control resp. current drive. Heatsink's too small, so they need a (input signal-controlled!) fan. Better stay away from them. This is a somehow random and incomplete list of good and recommendable units: New Class A: Integrated: SU-V10 (first New Class A device ever), SU-V6, SU-V8, SU-V7, SU-V707, SU-V9, SU-V909, SU-V85, SU-V90D, SU-V100D (Japan only). Power amps: SE-A3 (MK2), SE-5 (MK2), SE-A7, SE-A70 (USA only) Class AA: Integrated: SU-V900, SU-MA10. Power amps: SE-M100, SE-A50, SE-A100, SE-A2000, SE-A3000, SE-A5000, SE-A7000. Pre-amps: SU-A40, SU-A60, SU-A200, SU-A1010, SU-A2000, SU-A3000, SU-A5000, SU-A7000 There are also some good New Class A receivers (e.g. SA-616, SA-818, SA-1010), but surprisingly there are no Class AA receivers iirc. And no, the SA-1000 isn't anything of the above, but just a big plain class AB receiver. Let me know if you want some more elaboration on certain topics, or if something is missing or wrong!
Holy shit Macphantom, thanks, that's way more than I thought you might contribute to this thread. Again, thanks for explaining what are to me, excellent sounding under the radar amplifiers.
This thread is old, but I did not find many on this topic that had any useful content - thank you MacPhantom, great post ! To elaborate on that: the voltage amplifier gets feedback through the line departing from the speaker. As I understand it, this should keep the tension on the output constant whatever change in load resistance is occurring. If for instance at some frequency the speaker exhibits a resonance increasing its resistance, the tension on the output of a normal amplifier would go up, worsening the resonance - here, the current will be lowered (because of the bridge) to keep the tension constant. (I never understood why speakers require controlled tension although it is current that generates the electromagnetic force - but that is the way they are constructed). So the class A voltage amplifier works in a feedback mode all across the (not so good class B) current amplifier stages, comparing input and output tension singnals to make sure that both are the same by feeding any difference into the input so that it is eliminated - I suppose this meant by voltage clamp (VC). On the way this eliminates most problems of distortion, I guess. Feedback is not a new concept but as I remember from some experiments as a teen, it is (or better was at the time, differential amplifiers were unheard of) difficult to realise in detail so that it works really well. (It seems that Quad had already had this idea, but at the time I believe the electronics was not so well developed.) In then makes sense to put class AA into CD players (I also wondered before what that is good for before this post) - This way, whatever shitty follow up amplifier you have (with changing input impedance), at least the output voltage of your player will not budge.
I got an 8055 (class AB), a V5 (new class A), a VX2 (new class A) and a V45a (class AA) from eBay for comparison. The 8055 sounded like a transistor, the V5 had impressive dynamics but was (much) too heavy on the high frequencies, the VX2 had a very neutral sound in any aspect. The V45a had a more transparent, cleaner sound than the VX2 although the difference was less obvious than among the others. However, the V45a exhibited a slight lack of dynamics compared to all others. I found similar descriptions on some blogs since, in particular regarding the V5 and the V45a - so it is not only me ! I suspect that the 8055 and V5 sounded much better when they were new, and they need to be recalibrated or even need to have a change of some capacitances (the power meters of the 8055 show different values for both channels for unknown reasons, the V5 is only slightly better). I further suspect that Technics played with the dynamic response of the amplifiers - in particular if you have a VC, it does not seem difficult to make the cheaper systems sound a bit weaker that the expensive ones. I also suspect that class AA compensatesin an excellent way even for ageing components.
