Krell anticipator circuits of the 1990s

Very much as I thought.  No point in carrying power supply capacity that will never be used.

Both the FPB 600 and the ksa300s have the sustained plateau bias circuit, krell just lightened the power supply on the FPB series.

I'm not sure one should evaluate amplifiers by weight although when I bought my first CD player I didn't do any auditioning but just chose the heaviest in my (low) price range on the basis it probably had the best build quality.  Little did I (or anyone else then) know that the SQ in CD players (and all digital sources) depends almost only on the clock, jitter and DA converters.  Reading error even in a flimsy plastic 50c computer drive is only a couple of bits per million.

I doubt the anticipator circuits account for much weight, but compromised Class A working should allow some weight reduction as the power supply section accounts for most of the weight.

It seems like all of A ksa300s weighs 3lbs more than the krell FPB 600,  seems like it should weight more given it's double the power. The only FPB series amp I like are the FPB 750 MCX monoblocks.

Thanks, right.  It was 35,000 1993 dollars.  KRS200s were around $30,000 in the late 80s, so very much on a par.

Looking around the used ads I find KRS200s are regularly priced around double the 1990s FPBs.  I wonder if this is related to sound quality?

The krell audio standard was the first variable bias amplifier krell made, it was their flagship amplifier at the time 1993 I believe.

@invalid    I am aware of these facts.

Not so long.   I believe the first variable bias Krell was the KSA300S, launched in 1994.  The first of the FPB series started in 1997.

Krell has never been forthcoming with any of their circuit designs and schematics. They also came out with sustained plateau bias long before the FPB series came out.

@johnlnyc    Thank you for this input.

As in my OP, the anticipator circuit is claimed to monitor the output from the pre-amp and adjust the bias quickly enough so that the change has taken effect before that output is amplified.  A tall order I think.

At the introduction of the FPB series environmental issues were discussed.  My KRS200s draw more than 1kW per side.

Just because an idea is patented doesn't mean it is effective.

As a manufacturer previously claiming the benefits of pure Class A, if they wished to sell FPB prorduct it was encumbent on Krell to claim they had overcome the disadvantages of Class AB designs.

Yes the circuit could be proven on the bench but it has not been.  The time that the circuit takes to adjust the bias has never been revealed or found.  To that extent it is a secret circuit.

The paper you publish does not speak of the time it takes to change the bias.

I never said it was claimed the circuit could tell the future.  It does claim it can change the bias quickly enough in response to changes in the signal.  We don't know if that claim is true or not.

Where’d this “anticipator circuit” come from? 
I also don’t think any “green revolution” is impacting class A amplifier design and production. Anymore than it is driving development or sales of  class A/B,D etc….

As I see it, Krell has been attempting to improve their amplifiers by innovation which includes removing/addressing issues with class A amplification while maintaining the benefits. Heat, and efficiency being the problems with fully class A.

I doubt Krell is just tossing about ad claims to sell product. They are not a mass market company looking to move amps and ramp up production.

I do not, I am no engineer, think the OP bass drum analogy accurately describes what Krell is doing here. It is a sort of straw man. The process is patented. It can be proven on the bench and to the ears. It either delivers its goal or it does not.

It is not as some seem to think, a secret circuit that can tell the future. In the end does it deliver on its promise?

https://www.hificlube.net/media/11276/krell_ibias.pdf 

@invalid    You used the term 'sounds better'.   We are comparing different amplifiers all the time and expressing opinions as to which sounds better (to the listener).  None of these has the same circuit design, yet we express opinions, sometimes vehemently.

By the way, what makes you sure the circuit is fast enough??

How would you compare it to another amp, unless it has the same exact circuit design one with the anticipator circuit and one without.

@invalid    What makes you sure the circuit is fast enough?  What evidence do you have?  I know of no evidence that has been published?

Of course it can be compared in the same way any alternative components are compared.  Listen to one, then the other.

I’m sure the circuit can keep up with any musical dynamics, you may have an argument on if it sounds better than straight up class A because of the added circuitry, but there is no way to compare it unless someone comes up with an amp with a switch from this to straight up class A.

@jaytor 

Thanks for your reply and I understand your predicament. Once heard, engagement is very hard to do without.

Charles

Based on the way this circuit works, it should easily be able to keep up with a low frequency dynamic spike. By the time the waveform returns to the zero crossing, the bias will have been adjusted so the other half of the output transistors are biased into class A. I would expect this would even be the case for a high frequency spike unless the designer intentionally prevented this. 
 

@charles1dad - yes, I use both of these amps with the same speakers. The speakers are fairly efficient at 92-93db/w and are augmented by powered subs which handle up to about 70hz. So the 15w PSET amps play plenty loud. 
 

the Pass amps have a bit better control over the mid bass and are perhaps just a bit more detailed, but the tube monoblocks sounds more engaging with a deeper and broader sound stage. After going back and forth a few times, I have only been listening to the 300B amps and the Pass amps are tucked away in their boxes. I’m trying to decide whether I should sell them. 

The bias is always in class A unless you exceed a certain temperature, then the top two bias levels stay in class A/B until the temperature drops. 

@johnlnyc 

Martin Collums doesn't add anything to the debate.  He eulogises the anticipator circuit.  But he doesn't say how long it takes for the bias to be changed.  That still hides behind the emperor's new clothes.

At this point it is necessary for me to repeat the question in my OP: how long does it take for the amp to react to the increased current generated by the signal (the big bass drum hit) before the bias is restored to Class A.

No-one has told me.  But it will be longer than it takes for the initial pulse of that drum hit to pass.

Maybe Martin Collums in his Stereophile review, gets it.

“Krell's regulation feature is still more impressive when you take into account Krell's "Sustained Plateau Biasing," a patented Krell technique that provides an effective equivalent to class-A biasing—without a long-term power dissipation penalty—by anticipating the size of any and every musical event (footnote 1). It uses high-speed current-feedback circuitry to do this, then holds these required levels in a static condition for tens of seconds after the event is over. This minimizes any possible dynamic interaction of bias level with sound quality. The FPB 600's seven stages of bias represents the highest evolution yet of this technique…”

@jaytor 

Do you use your parallel  300b amplifier and Pass Labs XA 60.8 driving the same pair of speakers? Is so, this seems an interesting comparison of two very different topologies. Nice!

Charles

So the Bias slides up, big deal and down. Krell would have done time domain tests to determine if there spurs appear off the fundamental.  Something like this would not be done by a field but an engineer. We would do this in the digitally domain for the floating dynamic amps and  converters we built. It not simple but not rocket science either but timing has to perfect as did this to 60 channels in parallel. This was at Texas Instruments and several others. I doubt I changed any minds but keep an open mind that analog computing is fast.

I’m in on Pass Labs, Cherrs

@yyzsantabarbara    I am not up with the new Krell to which you refer.  However, surely if it stays in Class A all the time then it is a (pure) Class A amp.  If this is the case it has no need of iBias monitoring or adjustment as it is in Class A all the time.

So what am I missing?

@yyzsantabarbara - I was thinking about the bias mechanism used in the older FPB series which is what I used to own. There newer amps could use something completely different. 

@jaytor The type of iBias described by the WestministerLab designer seems to me different than what KRELL seems to be saying with the new iBias XD amps. I am under the impression that the new KRELL stays in Class A all the time. I was going to email Walter Schofield at KRELL and talk about the iBias but then I saw that he is no longer with KRELL.

All I know is that I put back my KRELL 175XD into my office system this morning and it sounds great. So smooth and yet powerful.

@yyzsantabarbara  - yes, I think that @clearthinker is misunderstanding the purpose of this type of circuit. It is not designed to anticipate the bias requirements to always be in class A. It is designed to adjust it's bias based on the user's load requirements so that it operates in class A "almost" all the time. 

I think most of us don't always listen to extremely dynamic music or at very high volumes all the time. For me, most of the time I am only needing a few watts of power from my amps. The plateau bias mechanism allows the amp to operate much more efficiently when high power is not needed, but has plenty of power reserves when it is. Once the bias has been ratcheted up, it will stay in high bias as long as the requirements persist (unless the heatsinks get too hot). 

If anyone is interested. I re-read the thread I posted earlier about 

WestminsterLab

They also have an iBias system similar to the KRELL XD. The designer of those amps explained how the iBias predicts the needed power. His amps go into Class AB for a second if they go beyond the existing limit of Class A. Then the Class A ramps up.

@retiredfarmer     Thanks for your understanding and kind words.  I agree with all you say.

I do take Stereophile (not AS as I have been a John Atkinson fan for the whole of his journalistic career) and am keen to keep up with all the new products.  But like with my cars I don't change equipment very often and very rarely sell anything.  Now approaching my mid-70s I am increasingly finding new products, especially cars, are not to my liking and that I get more pleasure and satisfaction with the older stuff I already have.

Neither am I much of a tweaker.  I use stock mains cable and have had my heavy Kimber speaker cables 20+ years and all Kimber silver interconnects for longer than that.  I am sure someone will tell me they oxidised years ago but the music still sounds great to me.

Perhaps one should not generalise about blue collar people, but one thing most of them do not do is love their work; maybe it's not that kind of work.  I loved my work and got as much out of it is from my down time.  Spending most of the time working it is crucial that it is found enjoyable.

Yes I love Da Vinci and Monet but also Dali, although I can't do the really modern stuff.

@clearthinker there is something  to  be said about owning  your amps for many years. First off you know them very well. Secondly  you  are off the rat race to continually  buy and sell. In my opinion  it takes a bunch of time  to get a system  at its peak. That is not in a month or two. That takes a few years. You can get them fairly close in a few months  but top performance  takes time and effort and listening.  As you eluded  to before many times the biggest effort  is put into early products.  When the person who formed the company  was in control and had the passion  for what he was making long before the bean counters  were there.  As far as time goes most everyone who plays with stereo  equipment  at this level works very hard in what ever they do. Most blue collar people think that the wealthy  are lazy that is far from the truth more like some of the hardest working  people out there. So that also brings up a couple of points spare time is limited  so what does a person  get there happiness  from in audio,  listening  to music  or discovering  each new piece that comes out?  None of those things are bad but a person  wants to get enjoyment  out of the hobby and that in a nut shell  is what the goal is. I always maintain  that what a person is doing is trying  to paint a picture with your hifi system.  There is no right artist just different  styles just because  a person's taste is a monet  doesn't  mean a DaVinci  is bad. I love hearing  other people s choices  as it is a view into that person's soul it shows in ways that people rarely show you. What makes me enjoy this thread is the passion  that many people are showing that is one thing that is wonderful  about this quest is it brings out passion.  

"it sounds like you have already decided that Krell's plateau bias technology is just marketing BS. I read through the patent in detail and, as an EE, I do think it's a clever approach and not just hype."

Then as an EE you also recognize each effort to modulate the power supply models as a non-linear s-curve - a hysteresis loop with rose and fall times that must be less than any changes in the input signal. 

 

 

 

@jaytor ​​@imhififan      Well, if you read my first post you will have seen I do not believe the claims made.  I decided that in the 90s when the FPB amps were introduced.  I examined the proposition and discussed it with a Krell service engineer.  I do not believe the current of the signal can be determined and the bias changed before that signal is amplified.  I am not an EE but I have read the patent and it says nothing of the quantum of this time delay.  If this is not known then the system's suitability for audio use cannot be approved.

Amplifiers have very many functionalities apart from audio and Classes B thru H have their valid uses where their differing topologies suit the required functionality.  But for audio amplification none can deliver the sound quality of Class A.

 it sounds like you have already decided that Krell's plateau bias technology is just marketing BS. I read through the patent in detail and, as an EE, I do think it's a clever approach and not just hype.

Totally agreed. 👍

The same can be said for class G and class H topology.

My favorite sounding Krell monoblock amplifiers to present day are the Krell MDA 300. I think the quiescent current per monoblock is 400 watts.🤒 😎

Mike

Perfectly stated @jaytor. +1

@clearthinker - it sounds like you have already decided that Krell's plateau bias technology is just marketing BS. I read through the patent in detail and, as an EE, I do think it's a clever approach and not just hype. Whether it really improves the sound quality compared to operating the same circuitry with a fixed class A/B bias, I can't say. But I was pretty impressed with the sound quality I got from this amp. When driving various Magneplanar and Revel speakers, I thought the sound quality was on par with the much less efficient Krell MDA-300 monoblocks that my FPB-300 replaced. 

When it failed after 20 years, I decided to trade it in instead of getting it fixed or I would probably still own it.

I like my Pass XA60.8 monoblocks a bit better (as well as my 300B SET amps), but my current speakers are a lot more efficient, my musical tastes have changed, and I generally don't listen at as high a volume as I used to. 

@clearthinker I agree with you new is ot a given that it is better. The idea that technology has advanced so much I find interesting what is really new technology in amplifiers? Preamplifier s? Phono stages? There have been both forward and backwards steps in digital. Some components have gotten better. In my personal opinion the levinson 33h s i have, got better with the new caps better than the original ones. I think powercords have got much more balanced in the last thirty years. The hifi receptacle s are much better than the hospital grade which were much better than standard receptacle s. Things have progressed but not everything and the old top of the line new product s are still contenders and the new ones have to be really good to be as good. The only amps that I have heard that I think are better than what I have or possibly are is a pair of de agustino mono blocks I heard in a store demo a couple years back they were there top dog at the time and were really pretty wonderful. Very few times do I listen to a demo and think I could live with that system. This was one of the few times I could say that.

@jaytor   I think you've probably been taken in by the hype, because if it could change the bias instantaniously it would indeed be 'pretty clever'.

@retiredfarmer   As in many cases, a manufacturer's first products are the purest, possibly because they are conceived without thought of saleability and profit.

As I recall it, the progression from the start in I think 1983 was: KSA50  KSA100, KMA100 - two bridged KSA100s essentially, KSA80 - replaced KSA50, [possibly a couple more], KRS100, KRS200.  As I recall that ended the pure Class A series by the late 1980s.  I still have KSA50 as well as my KRS200s, and a very good sounding amp it remains after 40 years.  All progress is not good progress.

Here is a link to Krell's patent describing their plateau biasing method. It adjusts the bias based on the amp's load current requirements, not the input signal. It seems pretty clever to me. 

Krell patent

ksa 80 from 1991 never been serviced.

Brings it everyday.

So much power it scares me.

I heard  a set of bryston  7b monoblocks  compared  to the krell fbp600. I am not s bryston  fan but I found them to sound much better in that system  that day. A friend used to have a set of krell kma monoblocks  I really liked them. I never liked the 600. 

There is opto-couplers in my ksa 300s amp. It will give you 300 watts of class A power for only a little while as the heat sinks aren't big enough to sustain it. If it gets to a certain temperature it will be class A/b on the top two bias levels until the temperature comes back down.

@dlcockrum - My friends amplifier is a Krell KSA-100S.  It has sustained plateau biasing with four lights on the front to indicate the various levels.  Not quite the same implementation, but similar concept.

+1@jasonbourne52,

Qualified Krell service engineers have not been able to explain to me how it can work.

In my opinion they could not explain how it works because they knew it was marketing baloney.

Example of marking baloney in audio see below:

This feature plateau biasing circuit allows for the Krell FPB 600 amplifier to have Class A bias output without all the wasted electricity and heat. Really?

This is all very unsatisfactory and in the result I continue to use my KRS200s.

I agree, keep the Krell KRS200s if want new ''real'' class A amps see Pass XS300 mono block amps, they will give you all the heat you need to keep your English castle warm. 😎

Mike

https://www.passlabs.com/product_tags/class-a/

Indeed.  If they don't tell us the details then we can be pretty sure it's a pig in a poke.

It may even be the iBias circuit is a spoof.  If there is anything good to tell then they will tell it.  They are protected by patents, after all.

If it sounds impossible then it probably is impossible.

It's interesting that the subject of my initial post:1990s anticipator circuits, continues into new Krell products being launched right now, nearly 30 years later.  Brought my subject right up to date.  Maybe they're still selling the same old tat.

Not at all like Krell's first product, the KSA50 which was and remains wonderful.  I still use the one I bought nearly 40 years ago, in my second system.  Not even a service required, although use has been sparse these 30 year.

I still don't believe it.

If that guy does not spill the beans, then I doubt anyone else will since I am under the belief that he was the inventor of the iBias circuit.

@yyzsantabarbara   Yes I did watch the video.  This adds nothing to the story.  The Krell personnel say the Class A watts output is varied according to monitoring of the watts of the signal output.  It does not say how it is done - 'we've found a way...', 'a very high degree...'., 'this is completely new...'.  All this is marketing puff.  It does not say how long the amp takes to adjust the bias in response to the measured output.

Since Krell must know the length of the delay, we are left to assume it must be relatively long.  If it were so short that there is an argument that its effect is inaudible, Krell would surely have told us the details.  This is all very unsatisfactory and in the result I continue to use my KRS200s.

I might be in over my head on this, and I really don’t know how Krell actually manages this, but I can imagine that if there was an opto-coupler at the amps input it could possibly alter the bias faster than the speed of sound and closer to the speed of light to adapt the output accordingly.

If Krell uses digital manipulation to accomplish this, there might be an issue of close proximity of digital noise contaminating the analog signal,

 I am curilous to know how Krell actually does accomplish this.

@clearthinker Did you watch the video I posted? I think that answers the initial question you posed on this thread.

@invalid     Sorry.   I meant to type KRS200s.

The ksa amps with the anticipator circuit also have a couple switches on a circuit board that say relaxed/normal, if switched to relaxed mode it becomes a class A/B amp.

@yyzsantabarbara 

Thanks for letting me know.  But I will stay with my KSA200s.  Don't want computerised bias shuffling.  Keep it simple, as someone here rightly said.