It proves that it is possible to experience interference even from a CE approved device to a CE approved device, which can be very annoying. It does not prove that either the transmitting or the receiving device does not comply.
Let's get back to topic though, which is Class ABBA.
Do I summarize correctly that:
- The merits of Class ABBA should be compared to an EF with double output transistors.
- There are doubts on the bias current stability of Class ABBA.
arjen6t8 - Those are my concerns.
Distortion in a class AB emitter-follower is a function of amplitude, bias current, and load impedance. Due to cancellation, it is possible to get very low distortion under specific conditions. However, one can't design an amplifier that does that consistently.
The one thing I have learned is that the DoubleCross circuit does reduce distortion at low amplitudes by introducing another cancellation point. The tradeoff is less margin against thermal runaway.
Ed
Distortion in a class AB emitter-follower is a function of amplitude, bias current, and load impedance. Due to cancellation, it is possible to get very low distortion under specific conditions. However, one can't design an amplifier that does that consistently.
The one thing I have learned is that the DoubleCross circuit does reduce distortion at low amplitudes by introducing another cancellation point. The tradeoff is less margin against thermal runaway.
Ed
correctly (simplified and condensed) summarized - we are on the way to Nick's individual (or global) solution!
In the end, we will probably all realize that the solution to the problem has recently been revealed to us, as have the causes.
Nick's circuit diagram (with component values and designations) and some fundamental explanations will confirm my feeling, I am convinced of that.
Without a schematic it remains a gut feeling. We all cook with water - but ABBA will be much easier and more reliable to rebuild than doubleCross by Ian H.
What is ABBA supposed to mean or say?
CLASS AB (equal to Self's diction) || BA SSALC - or EF2+EF (n times EF parallel) in Class B ?
kindly,
HBt.
First we have to find a complementary pair of transistors, i.e. an NPN-BJT that is exactly the opposite of the PNP-BJT and vice versa - this is still our cardinal problem.
How do we manage to make the WingSpread-diagram, which is ready to help, produce a real horizontal line, perhaps with three PP emitter followers connected in parallel as EF2+n (?) of course.
DoubleCross
is (in my view) too complex and complicated (but also really interesting).
ABBA
is only a name 😉.
How do we manage to make the WingSpread-diagram, which is ready to help, produce a real horizontal line, perhaps with three PP emitter followers connected in parallel as EF2+n (?) of course.
DoubleCross
is (in my view) too complex and complicated (but also really interesting).
ABBA
is only a name 😉.
The AM US range is within 530-1600KHz while the Hypex switching frequency is around 460KHz, that s why your receiver collect nothing.
https://audioxpress.com/article/fre...-amplifier-module-doubling-down-on-a-flagship
They define the allowed emissions within used frequencies, FI the Hypex switch at 460KHz while the Purifi is at 520KHz,
this way their EM pollution occur outside of the official AM range, so it s clear that the switching frequencies were not
chosen randomly as if these were within the AM range then the amps would be deemed as non conform,
that s how they manage to be within the emisions requirements.
Anyway classes A, AB and for that matter ABBA, produce no perturbation, only disadvantage is the poor efficency,
but at reasonable powers, say 50W, it doesnt matter much.
As for performance they are still above class D in many metrics.
Last edited:
Even when switching just outside the AM range, the harmonics which are almost as strong, are smack in the middle of the AM range.
They do not limit EMI by trying to stay outside AM range. They do that by good design.
The other thing is that designs like the Purifi ('Eigentakt') change their switching frequency if the load changes, so the exact switching frequency is undefined.
Jan
They do not limit EMI by trying to stay outside AM range. They do that by good design.
The other thing is that designs like the Purifi ('Eigentakt') change their switching frequency if the load changes, so the exact switching frequency is undefined.
Jan
Last edited:
That s not true, what you call good design is precisely to set the switching frequency outside of the radio range.
The Hipex has 0.4V residual at 460Khz, you think seriously that it would be within the norms should the switching
frequency be set at say 550kHz.?
The first harmonic at 910kHz is at 100uV, so that s relatively low and is eventualy within the accepted EM pollution level.
The Hipex has 0.4V residual at 460Khz, you think seriously that it would be within the norms should the switching
frequency be set at say 550kHz.?
The first harmonic at 910kHz is at 100uV, so that s relatively low and is eventualy within the accepted EM pollution level.
Just mesaure the frequency under load. I did.
You won't like it.
Everything else is speculation.
But anyway, you went from 'the authorities don't know about this' to discussing specific frequencies.
That's progress. The rest will come with time.
Jan
You won't like it.
Everything else is speculation.
But anyway, you went from 'the authorities don't know about this' to discussing specific frequencies.
That's progress. The rest will come with time.
Jan
Last edited:
Since you did measurements, what is the frequency and amplitude of the switching residuals.?
As said for the Hipex it s 0.4V, also when loaded and with volume at 0.
Also you said that the Purifi has variable frequency, but seems that you didnt even check the curve
i linked because it s clearly visible that the frequency is shifting but it doesnt seems to exceed 530kHz,
and we can se that the amplitude is at -35dB of the main signal, FTR the Hipex s 0.4V
is at -40dB below a typical 40V Pk output level, so that s comparable to the Purifi.
In a class AB amp there s a lot of low amplitude harmonics that extend well in the AM range
but they are of infinitesimal amplitude, in class D the distorsion is 1% according to the numbers above
but since it s at very high frequency it s practicaly at 0, neverless class D amount just to shift the distorsion
frequencies in the 100s KHz range albeit at a very high ratio.
Also those class D amps are of lower linearity at 1kHz even if it s ppm class THD, i posted a graph of a class AB design
that has the main residual (of 2Khz) at 100nV amplitude for 4V pk output, of course you ignored the number because
it s a prove that class D bring nothing at low powers, we re talking of residuals that do not exceed -152dB,
can you point me a class D amp that goes that low..?.
As said for the Hipex it s 0.4V, also when loaded and with volume at 0.
Also you said that the Purifi has variable frequency, but seems that you didnt even check the curve
i linked because it s clearly visible that the frequency is shifting but it doesnt seems to exceed 530kHz,
and we can se that the amplitude is at -35dB of the main signal, FTR the Hipex s 0.4V
is at -40dB below a typical 40V Pk output level, so that s comparable to the Purifi.
In a class AB amp there s a lot of low amplitude harmonics that extend well in the AM range
but they are of infinitesimal amplitude, in class D the distorsion is 1% according to the numbers above
but since it s at very high frequency it s practicaly at 0, neverless class D amount just to shift the distorsion
frequencies in the 100s KHz range albeit at a very high ratio.
Also those class D amps are of lower linearity at 1kHz even if it s ppm class THD, i posted a graph of a class AB design
that has the main residual (of 2Khz) at 100nV amplitude for 4V pk output, of course you ignored the number because
it s a prove that class D bring nothing at low powers, we re talking of residuals that do not exceed -152dB,
can you point me a class D amp that goes that low..?.
Actually, radiated magnetic emissions are allowed to be 10dB higher below 490kHz, so Hypex would have more room to operate withing compliance. But AFAIK, for both the switching frequency is not fixed, rather it is modulated by the signal, and of course there will be harmonics anyway that are within the more restricted range.
Bottom line, I would assume both do not only comply to the limits, they also are below the AM range limit at all frequencies. So it is quite possibly not correct to imply they -- Hypex, that is -- only passed EMC test because the switching frequency fundamental might be in the less restricted range below 490kHz.
These allowed emissions amount to 200uV/m and 70uV/m for the two first plateau asuming that the recipient is a tuned antenna,
notice that the distance is 30m, much more than the distance between two apartments in a same flat.
Obviously that s quite loosy standards, and that will be the case only if the speakers cables are short enough,
according to Maxim semiconductors it should be no more than 2m if the device want to pass the EM validation,
not sure that all Class D users are aware of this fact.
Also the varying switching frequency of the Purifi is on purpose as this allow to spread the residuals energy
within a larger bandwith and hence reduce the fundamentals and harmonics RMS amplitude at a given frequency,
this amount to spread the trash over a larger surface to make the ground looking cleaner.
To get back on topic of this thread methink that those laborious tries to reduce the THD by doubling the OS
and use different biaisings is useless, very low distorsion can be achieved with some design effort without
requiring such non stable biaising ratios, it s jist that people are stuck in old concepts and are quite conservative
when it comes to amplifiers toplogies and designs, as i already pointed it -152dB/1W/1kHz level residuals is possible
without that much transistors.
notice that the distance is 30m, much more than the distance between two apartments in a same flat.
Obviously that s quite loosy standards, and that will be the case only if the speakers cables are short enough,
according to Maxim semiconductors it should be no more than 2m if the device want to pass the EM validation,
not sure that all Class D users are aware of this fact.
Also the varying switching frequency of the Purifi is on purpose as this allow to spread the residuals energy
within a larger bandwith and hence reduce the fundamentals and harmonics RMS amplitude at a given frequency,
this amount to spread the trash over a larger surface to make the ground looking cleaner.
To get back on topic of this thread methink that those laborious tries to reduce the THD by doubling the OS
and use different biaisings is useless, very low distorsion can be achieved with some design effort without
requiring such non stable biaising ratios, it s jist that people are stuck in old concepts and are quite conservative
when it comes to amplifiers toplogies and designs, as i already pointed it -152dB/1W/1kHz level residuals is possible
without that much transistors.
Maybe you are a radio amateur, and from that point of view the standards may look lose to you.
So you may critisize these standards, but not the products that comply
Anyway the AM disturbance, if existant, is of no relevance to 99.99% of the population
So acording to your 99.99% there s no more than 8000 people in Germany who eventualy fire an AM radio to do
some DX listening or use whatever such sensitive devices, and much less radio amateurs, and no more than 800k
people worldwide.
FTR there s 75000 radio amateurs in Germany, that s almost 0.1% of the population, so your estimation is off
by just an order of magnitude, and that s only radio amateurs, not people doing DX listening.
Notice that Canada wich is much less populated has 81000 hams while it s over 800 000 in the USA.
some DX listening or use whatever such sensitive devices, and much less radio amateurs, and no more than 800k
people worldwide.
FTR there s 75000 radio amateurs in Germany, that s almost 0.1% of the population, so your estimation is off
by just an order of magnitude, and that s only radio amateurs, not people doing DX listening.
Notice that Canada wich is much less populated has 81000 hams while it s over 800 000 in the USA.
Last edited:
Ok, one 9 to much, its about 99.9% who do not care.
I stand corrected.
This gives a tremendous different picture of the importance of this issue worldwide.
I stand corrected.
This gives a tremendous different picture of the importance of this issue worldwide.
So you are saying that everyone should use inefficient technology (class AB, not to mention class A) just because a hobby of a very small number of people will potentially be affected.
Were you so adamant against public restrictions on smoking or banning diesel powered vehicles from city centers? Those decisions impacted a much larger number of people.
+1
What would be the modern, correct (in the sense of reason) design concept?
Gentlemen,
could we please concentrate again on "the problem of transferring NPN to PNP crossover and vice versa ".
Electrosmog is a topic in its own right, and an unspeakable one at that. It should be clear to every humanoid being that it is not well-disposed towards us.
#
Class AB is overbias Class B,
what is ABBA supposed to mean,
Agnetha Björn Benny Anni-Frid ?
Indeed there s even les people who are wheel chaired, so we should stop building acess ramps, because after all...
I ve never been a proponent of class A, beside do you think that those 2 x 200W class D amps are more efficient overall
than my 2 x 15W class AB amp that use a toroid wich pull 2W when idle.?
Most of those amps use an SMPS btw, is a 500W SMPS idle power lower than 2W.?.
That being said with my band we used class D amps for PA purposes in the late 90, that s one case where
they make sense but for domestic use they are just some bragging rights tool, as who need 2 x 200W at home, seriously.?
@hbtaudio, i wont publish any schematic, only hint i ll give is that i used as basis the amp schematic that i posted
in another thread but that i radicaly redesigned with some old yet quite luminous ideas that are not used in any
design posted in this forum.
Certainly that people like Jan Didden could design such an amp but apparently they went to class D believing that it has much greener pastures, we ll see in a few years once they ll have explored the thing, eventualy they ll find themselves like those people who divorced a woman that they thought was too demanding to just fall in another one that end being overly tyrannic once
the honeymoon was over.
Maybe we should stop feeding the troll.
When we explain he is wrong on a subject, he moves to another subject. Where he is equally wrong.
Classic trolling.
Jan
Lol, all i do is answering to people who talked to me, so to be polite and not leave their posts
unanswered is to troll according to you.?.
It s you who are changing subject and going as far as insulting someone because you dont agree with him,
guess that when i demonstrated you 10 years ago that an IPS is dominant in matter of distorsion and that
you said that you were unaware of this fact it was also trolling, isnt it...
That s quite low from your part Ian, you just lost my respect with your arrogance and insults, indeed that s the only things
left for someone who is not up to a debate.
unanswered is to troll according to you.?.
It s you who are changing subject and going as far as insulting someone because you dont agree with him,
guess that when i demonstrated you 10 years ago that an IPS is dominant in matter of distorsion and that
you said that you were unaware of this fact it was also trolling, isnt it...
That s quite low from your part Ian, you just lost my respect with your arrogance and insults, indeed that s the only things
left for someone who is not up to a debate.
Last edited:
I appreciate your advice very much; it would be all the more regrettable if you did not publish your dimensioning and modifications of the well-known basic topology, such an important blueprint should not remain hidden.
Actually, I also assume that Jan D. can put together a really good analog Class B.
that can be, for Germany, the following highlights are still missing:
"Learning and teaching sustainability" "Driving digitalization forward", i.e. "sustainable and digital, as well as CO2-neutral".
That is a very apt comparison!
I hope so,
Nick S. can take the lead in this thread again soon and give us all the details about his Class AB-BA concept, so that no questions remain unanswered for any interested self-builder.
I look forward to a controversial discussion, friendly, open and profitable.
Kind regards,
HBt.