how they do that?


Does some of you already seen some amp from crown.
Especially the k series. How can they produced so much output and their is NO air cooling the case is completly closed. And yet it's still pretty cool for an amp.

In my opinion this Balanced current deviced must be some sort of class c amp like the current dumping (i think it's called that way). Does nayone got info on this i would like to see that.
My guess is that they have the amp in class H. That means multiple operating voltages, which are switched on only when needed.

An amplifier with 100V operating voltage produces 96W heat when it drives 4 Watts into 4 Ohms. The smaller signal level, the worse efficiency.

Same amplifier with switchable 50V and 100V would only make 46W heat in that case, so the cooling problems are practically those of a four times smaller amplifier (that is, one that operates with real 50V all the time), with the exception of peak power.

-Kimmo S.


diyAudio Moderator Emeritus
2002-01-19 9:25 am
Hey JBL (Love your speakers),

The link below is to the Crown site. In the K-Series amps, they are using a variation of the Class-D (Switching PSU) type configuration (Super Efficient). You'll see what looks like Common-Mode-Chokes to filter out the switching noise yet allow the "DC" through to power the dynamic rails.

Sony had a class-D out in the early 80's. There were others, but I can't think of them just now.

Here's a pic of the Sony.

Rodd Yamas***a


  • sony_ta_n88_01.jpg
    38.1 KB · Views: 4,881
The K series do not use a switching PSU, they use a variation of CLASS D amplifier. The PSU is a quite conventional linear one with toroids.

Gerald Stanley's innovation in the design of the BCA Class D circuit is using two inductors connected in series; the center point being connected to the output, the ends being connected to the upper and lower transistors. Under idle conditions both transistors are turned on and off for 50% duty cycle, but they are actuallly controlled from ramp comparators running anitphase. So, at idle, both turn on and off at the same time. Current circulates thorugh the inductors, but there is no net output. With a signal, the PWM duty cycle on one set of transitors increases, while on theother set it decreases.

This configuration has several interseting advantages over a conventional half bridge or full bridge switching amplifier output stage as used by most companies.

First, it's impossible to have cross conduction through the half bridge, which is the bane of a conventional half bridge circuit, and makes the driver performance and timing quite crictical.

Second, the effective modulation ripple frequency seen going into the output smoothing filter is double that of the actual transistor switching frequency. This has the expected benefits as regards filter component requirements and HF output noise.

Third, it's possible to add additional phases interleaved, and reduce both distortion and HF noise without having to switch the transistors any faster.

There used to be a pretty comprehensive technical white paper on their site, and downloads of Gerald Stanley's conf. papers. But since Harmon has taken over Crown, things look a lot different, and the technical content is really dummed down.



There are a number a ways to make very efficient amplifiers: Class G and H are variations on rail switching, Class D is pulse width modulation (which someone else other than me will have to explain), and then there is B.A.S.H. which seems to be class D on steroids. You can insert the whole story ( pro and con) of B. Carver in here somewhere, as well.

At this point I gather none of these have a noise level that will meet audiophile levels of acceptability. The noise levels do seem to be comming down over time. Recently Doug Self published plans for a 50W class-G that is better than pervious efforts but by his own admission still falls short of really good class AB performance. There are very practical motives for all of these efforts with the advent of home theater systems where 5 or more 100+W amps need to be squeezed into a single box.

When we are dealing with only two channels, these schemes don't seem worth the trouble. But with multichannel systems, I think you will see a lot of market pressure to move in this direction. I predict (for whatever that it worth) that they will work their way into mid-fi and slowly, painfully, reluctantly into the higher-end world.

Actually, this may be a good challenge for experienced DIYers. Start with Self's scheme and see what can be done to push the noise envolpe in one direction and the power envelope in the other. Personally, I hate the looks of the ubiquitous 19" x 6" x ~16" rack-based box -- I would really like to see good equipment in something on a substantially smaller scale without having to sacrifice power.

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
The rationale for these alternative circuits is mostly
energy efficiency.

I have built them all, and can say with some authority that
none of them saves all that much energy over plain old
Class AB circuits with stupid linear supplies when used
for music.

If you are running continuous waves into industrial
equipment for long periods of time , it makes sense,
but if you are really serious about saving energy in
an audio system, use speaker drivers with greater

You can dissipate 1100 watts driving an 80 dB speaker
with a switcher, or you can dissipate 200 watts driving
a 90 dB speaker.
Even if the energy savings don't matter that much for home use, don't you still gain something by being able to reduce the size of the the transformer and the heatsinking? Some of the multichannel amps and AVRs have grown rather inconveniently large. I'm not an engineer but an industrial accountant by profession. It's not rocket science that big ticket items are the sinks, x-formers and an enclosure to hold them. Except for the for the filter caps in the PS, the passive components are pretty much chicken feed and the in quantity the active components aren't too much more. In mass market quantities, reducing the size and mass of these three or four big chunks of metal are should be pretty significant. Not to mention that there is a population that would apprieciate quality products even at a significant price but as a practical matter simply can't deal 50-100 lb 19"x19"x8" amplifiers.

I think if you just look at the cost of reduced electric bills you miss the bigger issues.

Crown claims .1% THD. They have a pretty good reputation so i expect that it really a bit better. Nonetheless, this might be considered a bit high in some audiophile circles where a couple more leading zeroes (.001%) are not un heard up. .1% is probably not audible, but distortions are cummulative and two or three steps in the chain at such a level would get you in to a range that is apearently audible to some. So I would say this is pretty respectable performance for a lot people for a lot of purposes (perhaps a sub woofer amp or a HT system) but it falls short of what some audiophiles would demand.

It's important to keep in mind though that this is much better than PWM amps could previously do. If Crown or some one else can add one or more of those leading zeros (which may well happen eventually) then it would be a whole other ball game.
THD of "modern" class D

It's important to keep in mind though that this is much better than PWM amps could previously do.

Actually, 0.1% THD is typical for today's class D amps. The Tripath TA0140 evaluation board has a THD of 0.1% into 4 ohms. TACT's digital amplifiers have a reported THD of <0.01% (20 to 20K).

No idea on the veracity of these specs, but these number are what the companies have reported.

Personally I use a class-x amplifier - this novel design uses subspace transmutational inductance theory (taking into acount the Hubbel-Einstein constant re-equalisation curve) to convey sound.

It is (when applied correctly) up to 99.9994% efficient and creates a THD of 0.0000003%.

The basic theory is to modulate electro-chemical reactions within the "listener" in accordance with the program content.

The only limitations are of the highly complex filter system used to employ the system - which creates a stop band at the ABBA frequencies to protect the listeners sanity.

Although I may have proven to be of some great insight this knowledge must remain top secret as I am still under obligation to (but no longer working for due to psychotic episodes) NASA.
Hi all

Having developed a class-d amplifier myself a few years ago I'd like to throw in my two pence.

My opinion is that the performance of switching amps will be approaching the performance of conventional amplifiers one day. Though I have a distinct feeling that they will never exceed the performance of the best conventional amps available. But for a majority of people and applications they will be fine (so they will save all the energy wasted by the class-A users ;) )

I was able to listen to two of the best switching amplifiers available today, the Sharp SM-SX100 and the TacT Millennium and I must admit that both sounded excellent. If you have a look at their interiors then you can see how much effort has been put into these (engineering and build, specially the TacT which is an excellent showcase for a well-built piece of equipment).
But I think a conventional amp for the same price sounds equally well.

One thing I like about switching amps is the fact that they are left virtually unimpressed by "difficult" loads. This is true for output signals whose frequency lies well below the output filter's cutoff frequency. For higher frequencies this is an absolutely different story though.

The Crown K-Series is not just a bridged version of a conventional class-D amplifier (JohnMarsh's description is correct in this respect) like Nelson Pass assumed.

It is originally meant to be used for professional P.A. applications. Though a friend of mine once gave them a try and he found them not beeing that useful when used in large quantities within the same rack, because they don't have any fans. Even if they dissipate only about 200 Watts/amp under full load condition this can be quite much if you can't get rid of it. But for applications where there are just a few amps in a rack they will be cool (in every respect).
I was once interested in this amplifier because I am playing the bass-guitar. For this purpose planet10's approach of high-efficiency speaker/low-power amplifier doesn't work that well because speakers that can be moved easily ANDgo really low and loud (not only in their sales brochures !!) aren't that efficient usually.

Though there is a B&O dealer close to the place I work I haven't found the time (or maybe wasn't willing to) audition their Beolab 1 speakers which are making use of their ICE power modules, so I have no idea about their performance so far. But I have studied the patent related to their feedback topology and I must admit that it is quite clever.

If anybody is interested I can try to find the circuit of my class-d amp and scan it. But this might take a few days. It's measured performance wasn't very good at all but we listened to music through it (although not intended for this purpuse !) and it sounded quite pleasing.


Thermal Advantage

Hi all,

I can see the thermal advantage of making subwoofer amps with CLASS D, but how about the trible and midrange ?

I something gained other than a lot of work building a full range class D amp ?. I mean the thermal stress of a trible amp for active systems is not a problem in consumer systems. (don't know about PA stuff)

I seams to be only the "smaller" 250 W of the B&O models that have full BW.


Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
Well, I've read what I can find, including Mr. Marsh's
comments, and I still see bridged PWM amps.

OK, they are synchronized bridged PWM amps, but
apart from that I am sceptical that:

"Crown’s BCA technology creates a totally new paradigm for amplifier design that represents the future of professional amplifiers."