Well, you can force many good classA amp into switching, leaving classA
You describe Krills switching well
But I would like to know which real time situation resembles the theoretical ones where it is clearly switching
40V have been mentioned
Anyway, it may even be questionable if you will hear the diffference between a very good AB amp and a non switching classA
If there is a difference it might have other reasons than the switching itself
You may not even notice when a classA amp leaves its pure classA and goes into a switching design
You describe Krills switching well
But I would like to know which real time situation resembles the theoretical ones where it is clearly switching
40V have been mentioned
Anyway, it may even be questionable if you will hear the diffference between a very good AB amp and a non switching classA
If there is a difference it might have other reasons than the switching itself
You may not even notice when a classA amp leaves its pure classA and goes into a switching design
Last edited:
the first sentence is the answer to the second...
because in such amps the switching occurs at very high levels,
where music vastly mask the switching products..
with class AB , this will occur at low levels, depending of the
quiescent current...
regards,
wahab
That was supposed to be my point with too, about Krill
The arguments so far seems to me to relate on high level SPL too, pushing the Krill beyond "normal" listening, to prove it is swithching
You can operate many AB amps such that low SPL listening is classA
And theres the sliding bias
If Technics new classA is one of those, I only remember it as pretty awful
My friend love his 100watt Luxman, and believes it makes 20watt in classA, but I seriously doubt that
But he doesnt notice any difference at all
I may remember about JVC doing an "almost classA" design
It got a very fine rewiev around here
Old Sansui amps too
Then came NAD, the old original one
If you read the Steve's latest posts, you'll find out that he's now claiming to bias the Krill @ 200mA with 0.1ohm emitter resistors, which is close to the optimum, anyway (26mV across emitter resistor). The 25mA bias claim was now retired, as not necessary optimal.
Compared to an EF, crossover distortions are nowhere better in Krill. To add insult to injury, and as stated a gazillion times already, the crossover distortions have little to nothing to do with output devices "switching", anyway.
When I'll see the 200W and 400W modules schematics, I may comment about. Perhaps we'll even see those claimed as 400W 0.005% distortions (open loop).
was manufactured in my country a model in this season with a circuit Super-A(non-switching).
I have the schematic, I will try my files on the HD.
Attachments
If I made a list of the op-amps that are "switching" in the op-amp "rolling" thread it might not fit into a post.
Hi wahab,
excellent, so maybe Steve can now get on with how he has improved the original Krill circuit.
thanks
Not just yet. First, I need the answer to some simple questions.
1. What is the minimum current that a power transistor (the ones in Andy's sim for example) must be conducting to be considered on?
I posted results from the sims I did using the schematic I posted. I used Andy's models.
2. Why has no one else used that schematic?
Many people here have posted circuits.
3. How many of you have posted the sim files with every circuit you posted as is being demanded of me?
The next one may be a little more difficult to answer.
4. What effect, if any, will be made on the sim results if you use the altered models posted by Andy in his last posted file, verses the ones he posted on the first page of this thread? The ones from the first page are the ones I have been using.
I expect I will be binned or banned for asking that last one, so I may not be around to post my files on here. I am willing to send my files to most of the people here. E-mail me if you want them.
So what did I miss? Post number #2 from the original Krill thread, OS biased at 5.55mV/.22 Ohms or 25ma. These were the numbers that started the discussion a year ago. I would certainly hope there are no more bannings for a while.
Attachments
Hmmm.
I just downloaded the simulation file from post #11, then extracted the "KSA_MJLmodels.txt", renaming it "KSA_MJLmodels_1.txt". Then I did the same for the file in post #210. I renamed that file "KSA_MJLmodels_2.txt". Then I opened up a command prompt, and ran the command-line file compare utility FC. The command line is:
FC KSA_MJLmodels_1.txt KSA_MJLmodels_2.txt
The result is captured below.
Steve, per your usual tactics, you provided no way for the reader to verify this for him/herself, just the usual claim without evidence.
I just downloaded the simulation file from post #11, then extracted the "KSA_MJLmodels.txt", renaming it "KSA_MJLmodels_1.txt". Then I did the same for the file in post #210. I renamed that file "KSA_MJLmodels_2.txt". Then I opened up a command prompt, and ran the command-line file compare utility FC. The command line is:
FC KSA_MJLmodels_1.txt KSA_MJLmodels_2.txt
The result is captured below.
Steve, per your usual tactics, you provided no way for the reader to verify this for him/herself, just the usual claim without evidence.
Attachments
As a minimum. I did later change that to 25 to 100ma.
So, are you saying a first draft can never be changed or improved on? I have stated several times now that I started this here before I was ready, but pain killers and steroids can make you do things differently from what you should have.
Maybe I don't like being called a liar by someone that alterers the models.
Anyone can do what what I did. Download both and go through the models looking for a difference. If I provided that information, it would immediately be labeled as questionable. This way, everyone gets to find it for themselves. The differences are there. Anyone can find them.
Is that why it took you 28 days? When you ran the files I provided, it showed I was telling the truth about the results I had posted. You either hoped no one would ever bring it up again, or you needed time to make it not work in your last posted file.
Why didn't you post the results from the file I posted. Why did you not use the correct schematic, the one I posted results from, instead of a different one?
Sorry, I'm not being too subtle am I?
And why should I believe anything you have to say now that you altered the models? Did you just forget to tell us you had updated or improved them?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.