Apple Designing 5G iPhone Antenna Module In-House After Being Dissatisfied With Qualcomm's Version

[victorm1]

Well Apple is far from a charity, so yeah it’s a business decision ?

But remember, my reply was in response to OP’s over the top post:

“Just like Apple tries to screw every last penny out of their customers, it's horrible when the shoes on the other foot ?”

You don’t sell $91 billion of product in a quarter if your customers feel screwed out of every last penny. On the contrary, customers must feel they are getting their money’s worth.

Generally speaking, few successful businesses run on the principle of systematically exploiting their clients. Payday loan companies are one of the exceptions.

 

[szw-mapple fan]

Doesn’t make sense to me, maybe I’m just dense. If Apple is Burger King, what do the smaller restaurants represent?

Sure Apple makes iOS and MacOS for their exclusive use, and the price is bundled into the hardware purchase. I fail to see the monopoly aspect.

Apple is the customer, Qualcomm is Burger King, and the smaller restaurants are the minor suppliers that may be able to make an antenna that fits Apple's needs but can't make as many of the same units as Apple needs for each year's iPhone, so Apple has no choice but to choose Qualcomm.

Apple isn't a monopoly in the computer or OS market, but it does have monopoly on customers who absolutely need macOS or iOS (like Logic Pro users). It's more of an artificial monopoly based on customer needs, but it is one all the same. In much the same way, there are off the shelf antenna modules that Apple can probably use, but Apple has a particular need for the design to be a certain way.

 

[PickUrPoison]

Apple is the customer, Qualcomm is Burger King, and the smaller restaurants are the minor suppliers that may be able to make an antenna that fits Apple's needs but can't make as many of the same units as Apple needs for each year's iPhone, so Apple has no choice but to choose Qualcomm.

Apple isn't a monopoly in the computer or OS market, but it does have monopoly on customers who absolutely need macOS or iOS (like Logic Pro users). It's more of an artificial monopoly based on customer needs, but it is one all the same. In much the same way, there are off the shelf antenna modules that Apple can probably use, but Apple has a particular need for the design to be a certain way.

Well now I know why it didn’t make sense—you completely changed topics.

I was responding to your post #203 response in agreement to the OP. OP said “If you need or want macOS or iOS, you have no choice but to buy the hardware that Apple sells, since they have a monopoly on those operating systems.” I didn’t realize you’d switched to talking about antenna modules.

 

[ncc1701d]

AirPod pro design is not good for me. Keep falling out. Never had the other issues described about noise cancelling though.

 

[Analog Kid]

Leaving aside your strangely competitive communication style and the ongoing debate about whether I’m stupid or merely ignorant, do I take it from the fact that you keep attacking the impedance matching strawman that you believe the Fast Company article is well researched and presented and you believe that Apple is incapable of designing a performant antenna?

My initial comment was aimed at that, after all.

Now, back down the rabbit hole:

Again, going back to the original post, the article said the antenna was inefficient and you claimed the author didn't know what they were talking about because it was all about matching. I can tell you're a circuit person because you fall into the same trap that you can fix anything by matching, which is wrong.

Ok, this is just getting silly… Keep beating that horse, it’s only mostly dead…

I have not said all problems can be fixed by matching. Nobody has. I have explicitly said that it is not true:

some problems can be fixed by better matching. That’s different than saying it’s the only parameter to optimize in a design or that all problems are fixed by matching and I don’t think anyone has made that claim.

I'm not sure why you keep going back to that. Seriously, I’m impressed you know about antenna efficiency— can we please move past it?

I can tell you don't know much about antennas because your terminology and assertions are simply wrong. For example, you keep saying "ohmic lossess" which is imprecise. We just call it losses, because there's a wide variety of dispersive loss models that do not utilize Ohm's law.

I won’t try to profile you from your posts, but for someone who just argued that d/lambda is a physical length to criticize someone else’s knowledge, terminology and assertions is kind of humorous.

Just as a point of language, though, removing adjectives makes your meaning less precise, not more. Ohmic losses is precisely what I meant. Power lost to heat whether in the conductors or dielectrics. Dispersion is a totally different concept— important to signal quality, but not power efficiency…

Transmitter people think its important. It isn't in the scheme of things, especially since your feed is so short on a cell phone. And you detune everything when somebody touches or sticks their head by the antenna. Matching is relatively easy compared to efficiency.

You’re deflecting… Regardless, once you remember that the argument that everything can be fixed by matching is a straw man of your own creation, you’ll realize that your statement that it’s “completely wrong” to say problems can be fixed by matching is not accurate.

Wrong.

Wrong.

I don’t know where you got the idea that this is an effective form of communication, but I believe most people would find it unnecessarily combative. When what you go on to say then turns out to be wrong itself, you may find that people struggle to be empathetic.

Wrong. Simple circuit theory. If you take a second to think about it, then you can easily see that in a lossless antenna, any power must be either returned to the source (stored in the matching network), or radiated. Thus, an antenna with a low Q must have higher losses, and therefore lower efficiency, because more energy is being returned.

This is the exact same concept as power factor in a power network, which is something they teach undergraduates...

Please, review the derivation of the Chu limit before you argue this any further. The Chu limit assumes free space (no dielectrics, nothing else in the near field), ideal lossless elements (no conduction losses, no power lost in the matching network) and a purely real (resistive) input impedance (nothing being returned to the source).

I don’t know what you think the Chu limit is, but it isn’t that.

Also, you might want to be careful bringing power factor into this discussion. Maybe you haven't gotten to that chapter yet, but someone who has is likely to start talking about power factor correction which smells to me a bit like impedance matching and we both know that's all just hooey…

 

[konqerror]

Just as a point of language, though, removing adjectives makes your meaning less precise, not more. Ohmic losses is precisely what I meant. Power lost to heat whether in the conductors or dielectrics. Dispersion is a totally different concept— important to signal quality, but not power efficiency…

Wrong. In electromagnetics, dispersion means epsilon and mu change with frequency. This means frequency-dependent loss, since the imaginary part of epsilon and mu describe loss. The manner the imaginary parts change with frequency are called dispersive models, e.g. Debye.

I use the exact same software that Apple uses to design their antennas and loss models are called dispersion.

Fundamentally, Ohm's law does not vary with frequency. Frequency-dependent loss models rely on parameters, beyond "resistance", that deal with the solid-state physics behind the material. This is why we refer to them as "loss" and not "Ohmic losses".

I don’t know where you got the idea that this is an effective form of communication, but I believe most people would find it unnecessarily combative. When what you go on to say then turns out to be wrong itself, you may find that people struggle to be empathetic.

The simple fact is that nearly 100% of what you have claimed so far is incorrect or incomplete. I would not take this position if you don't keep insisting that you know what you're talking about, because you don't and you keep repeating falsehoods.

The fact is that you may know analog circuits, but you keep making mistakes that reveal a lack of knowledge of electromagnetics and antennas. The fact that you think dispersion isn't related to loss is a perfect example.

I don’t know what you think the Chu limit is, but it isn’t that.

Also, you might want to be careful bringing power factor into this discussion. Maybe you haven't gotten to that chapter yet, but someone who has is likely to start talking about power factor correction which smells to me a bit like impedance matching and we both know that's all just hooey…

A failure here of engineering reasoning. Power factor does not cause any power loss in a lossless system, obviously. It tells you how much energy you unnecessarily move.

Similarly the Chu limit tells you how much energy you have to unnecessarily move, via Q. So in a lossy system, the Chu limit poses an indirect bound on wasted energy.

Simple reasoning.

 

[Analog Kid]

Wrong.

??‍♂️

Wrong. In electromagnetics, dispersion means epsilon and mu change with frequency. This means frequency-dependent loss, since the imaginary part of epsilon and mu describe loss. The manner the imaginary parts change with frequency are called dispersive models, e.g. Debye.

You’re conflating dispersion with the loss tangent. Both are related to permittivity but describe different phenomena. Dispersion is what you get when phase velocity is a function of frequency. The imaginary portion of the permittivity contributes to the loss tangent.

Dispersion distorts the signal, which can impact its information carrying capacity, but not its power and thus not its radiation efficiency.

I use the exact same software that Apple uses to design their antennas and loss models are called dispersion.

I don’t know the tool or have a manual, so I can’t speak to whether or why the tool uses dispersion to mean loss. It’s possible they mean it as a shorthand for all permittivity related parameters, or it may be that it is reporting “dispersion and loss”, not “dispersion loss”. Or the tool could be taking a link budget oriented approach and addressing it as a loss of data capacity rather than a loss of radiated power.

Fundamentally, Ohm's law does not vary with frequency. Frequency-dependent loss models rely on parameters, beyond "resistance", that deal with the solid-state physics behind the material. This is why we refer to them as "loss" and not "Ohmic losses".

I think I was pretty clear in what I meant:

Power lost to heat whether in the conductors or dielectrics.

Radiation loss is also a loss, but we need a way to separate the desired losses from the undesired ones. If you don’t like my term then call it what you want, but just calling it “loss” is ambiguous. Shall we call it “losses to heat”?

The solid state physics you’re describing includes the loss tangent which sums the conductivity with the imaginary component of the frequency scaled permittivity. You can do this because frequency scaled permittivity has units of 1/(ohm*meter). This is why the ESR of a capacitor is measured in ohms and treated like a frequency dependent resistor and the same carries beyond just lumped element analysis into the field gradients within dielectrics.

Did Ohm test his relation against frequency variant sources in the early 1800’s? Not that I know of. Did he understand that resistance can be parameterized and can vary with temperature at the least? Yes. Did he understand resistivity beyond lumped resistance? Yes. Do we see a familiar pattern among other relations for voltage, current and complex impedance that is so similar to E=IR that someone could call it Ohms Law and just about anyone would understand the meaning? Yes.

So even if not fully inclusive I feel my terminology is not without basis and should have been sufficient for you to understand my point. An inability to bridge even small differences in terminology is not an indication of a stronger understanding of concepts— quite the opposite.

A failure here of engineering reasoning. Power factor does not cause any power loss in a lossless system, obviously. It tells you how much energy you unnecessarily move.

Similarly the Chu limit tells you how much energy you have to unnecessarily move, via Q. So in a lossy system, the Chu limit poses an indirect bound on wasted energy.

Simple reasoning.

Indirect bound.

Ok, you’re getting close to an island of truth to stand on. Let me take you the rest of the way…

First, you’ve been so gracious in this discussion so far that I hesitate to point out your errors, but it’s actually higher Q that indicates more stored energy, not lower Q. Stored energy itself isn’t wasted energy, and in the Chu derivation all of the stored energy eventually radiates out because there is nowhere else for it to go. Once we depart from the Chu formulation, the ideality of the components is removed and losses are introduced, then those losses dissipate some of that stored energy over time leading to reduced efficiency. While Chu uses Q to determine the fractional bandwidth of the antenna, he does calculate it as the ratio of stored energy to radiated power in his ideal lossless circuit model. So, stored energy is proportional to Q. In most solutions to the Chu limit, the Q is a summation of powers of 1/ka, where ka is the electrical radius of the bounding sphere of the antenna. If you keep the physical dimension of that bounding sphere (’a’) fixed and reduce the wavelength then ‘k’ increases and Q decreases, meaning stored energy decreases, meaning efficiency increases. As you keep the physical dimension the same and increase frequency, efficiency improves.

So, we’re finally on the same page here, but how did this conversation start?

More importantly to the mm-wave discussion, it's not the physical size of the antenna that drives the efficiency but the electrical size. Keep the enclosure design the same and increase the frequency and the antenna gets electrically larger and efficiency goes up

Somehow, after 8 posts, you agree with me while trying even now to make it sound like I’m the one that has no idea what they’re talking about. ?

Still, while it seems to have eventually brought you to the right answer, using something like the Chu Limit to indirectly estimate efficiency is kind of pounding screws…

For one thing, substituting lossy components changes the Q, so the parameter you’re calculating is wrong from the start. For another, the loss isn’t uniform— there are different contributions from the conductors and dielectrics. And, of course, the Chu limit still assumes free space and you just spent half this thread ranting about how hard antenna design is because of all the other stuff in the near field.

A much more sensible approach is to use one of the many other models for antenna dissipation factor out there to estimate efficiency. Perhaps your confusion comes from the fact that many derivations begin from Chu’s equivalent circuit rather than his limit on Q as their starting point, but adapt it to more practical problems and then estimate losses to heat directly.

The simple fact is that nearly 100% of what you have claimed so far is incorrect or incomplete. I would not take this position if you don't keep insisting that you know what you're talking about, because you don't and you keep repeating falsehoods.

The fact is that you may know analog circuits, but you keep making mistakes that reveal a lack of knowledge of electromagnetics and antennas. The fact that you think dispersion isn't related to loss is a perfect example.

If you’re going to insist on keeping score, let’s review:

More importantly to the mm-wave discussion, it's not the physical size of the antenna that drives the efficiency but the electrical size. Keep the enclosure design the same and increase the frequency and the antenna gets electrically larger and efficiency goes up

We now agree I was right.

Actually no, loading antennas to make them electrically longer exactly results in loss. There is a relationship with the physical size, the well-known Chu limit

Not so much…

The Chu limit relates bandwidth to the electrical size (k*a) of the antenna. The Chu limit is expressly for lossless antennas.

I’m right on both counts.

Wrong. You are mistaken in calling d/lambda the "electrical length". It is not. That is the physical length of the antenna.

The Chu limit is a nice piece of theory that matches the loss due to low radiation resistance versus tricks like matching and loading.

Wrong on both counts. d/lambda is an electrical length, and the Chu limit doesn’t account for any loss beyond radiation loss.

The Chu limit is a nice piece of theory that sets an upper bound on bandwidth for an antenna of a given electrical dimension. As I said, loss (beyond radiation loss) doesn’t enter into the Chu limit at all— he assumes idealized, lossless elements.

Again, right on all counts.

Wrong. Simple circuit theory. If you take a second to think about it, then you can easily see that in a lossless antenna, any power must be either returned to the source (stored in the matching network), or radiated. Thus, an antenna with a low Q must have higher losses, and therefore lower efficiency, because more energy is being returned.

Wrong on two counts: The energy is not only stored in the matching network, it is also stored in the near field of the radiating element (which is why the match is even necessary to begin with). Higher Q circuits mean more stored energy, not less.

Please, review the derivation of the Chu limit before you argue this any further. The Chu limit assumes free space (no dielectrics, nothing else in the near field), ideal lossless elements (no conduction losses, no power lost in the matching network) and a purely real (resistive) input impedance (nothing being returned to the source).

All correct.

Similarly the Chu limit tells you how much energy you have to unnecessarily move, via Q. So in a lossy system, the Chu limit poses an indirect bound on wasted energy.

Not the intent of the Chu limit, but indirectly and approximately correct.

And as for dispersion, I’m pretty sure I’m right on that as well, but even if there is some narrow subfield fo EM theory that has expanded the definition of dispersion to focus on frequency dependent susceptibility rather than actual dispersal of frequencies that really adds nothing to your argument.

This is the problem with credentialist and pedantic language arguments-- I’ve managed to be right on the essential facts without once having to mention what degrees, title or experience I have while you appear more interested in discrediting me as you stumble through the underlying theory.

Being a cretin doesn’t make me wrong. ?

 

[konqerror]

You’re conflating dispersion with the loss tangent. Both are related to permittivity but describe different phenomena. Dispersion is what you get when phase velocity is a function of frequency. The imaginary portion of the permittivity contributes to the loss tangent.

Again dead wrong. In electromagnetics, field theory, epsilon and mu are complex values incorporating loss. Disperson is the frequency-dependent variation of epsilon and mu. The reason why you don't bother to make a distinction is due to Kramers-Kronig, which, by the way, says your second sentence is completely false.

Radiation loss is also a loss, but we need a way to separate the desired losses from the undesired ones. If you don’t like my term then call it what you want, but just calling it “loss” is ambiguous. Shall we call it “losses to heat”?

Circuit people call it radiation resistance and radiative losses. Fields people don't call radiation a loss.

Again, I use the same software Apple uses for antennas and field analysis. I click loss and it tells me the energy being dissipated thermally, not the energy being radiated.

You keep calling the language of engineers who deal with electromagnetics fields wrong. You need to admit that you don't know what you're talking about in the domain of electromagnetics.

I know what I am talking about and nothing is going to change my opinion about the inaccuracies you continue to state.

Wrong on two counts: The energy is not only stored in the matching network, it is also stored in the near field of the radiating element (which is why the match is even necessary to begin with). Higher Q circuits mean more stored energy, not less.

Again ignoring what I said. In a lossless system, energy is either stored or radiated. Energy cannot be stored with 100% efficiency, so storing energy should be avoided. A simple concept.

This is the problem with credentialist and pedantic language arguments-- I’ve managed to be right on the essential facts without once having to mention what degrees, title or experience I have while you appear more interested in discrediting me as you stumble through the underlying theory.

You may know about circuits, but little about E&M and fields. Any knowledge you have is in the low frequency regime and not in mm-wave and above. Thats from the fact that you use tand and not dispersive models.

 

[chucker23n1]

? this is amazing

 

[PickUrPoison]

? this is amazing

Yum yum yum... ?

Waiting with bated breath for the next smack down.

5,000 quatloos on Analog Kid ?
(No offense Konq!)

 

[konqerror]

Yum yum yum... ?

Waiting with bated breath for the next smack down.

5,000 quatloos on Analog Kid ?
(No offense Konq!)

Welcome to the Internet, where talking loudly > actual knowledge.

 

[PickUrPoison]

Welcome to the Internet, where talking loudly > actual knowledge.

Not sure that’s working out so well.

But Internet or not, arrogance, condescension and overconfident bluster ≠ actual knowledge. And being hard headed and closed minded is not conducive to learning.

One thing I’ve learned at my advanced age is that I’ve rarely been more wrong than when I was absolutely, 100% sure I was right ?‍♂️

In any case, it’ll be interesting to see if Apple does in fact design their own antenna module. It would be a shame if good antenna design were let down by it.

 

[konqerror]

One thing I’ve learned at my advanced age is that I’ve rarely been more wrong than when I was absolutely, 100% sure I was right ?‍♂️

I know. Explain that to Analog Kid. It's fairly obvious to me that he has worked in a much lower frequency or limited, guided wave regime of electromagnetics, not "high frequency" antennas. He won't admit his knowledge falls apart in the "high frequency" regime when you push things in cell phones and mm-wave.

 

[Analog Kid]

You keep calling the language of engineers who deal with electromagnetics fields wrong. You need to admit that you don't know what you're talking about in the domain of electromagnetics.

You brought language orthodoxy into the discussion as part of your campaign to out me as an electromagnetic fields poser, not me.

To be honest, what I’ve been doing since the first few posts started as explaining myself then transitioned to defending myself. I made a fairly straightforward comment that the Fast Company article was thin and I felt Apple wouldn’t have trouble designing an antenna. You’re strangely intent on ignoring my points but on discrediting me. Not a big deal from a stranger, but you’re so quick to judge me while not holding yourself to the same standards of rigor it seemed worth pointing out.

Now it appears we’ve become a bit of a pencil-neck cage match which probably means it’s time to walk away, but I just find it so entertaining that you think I shouldn’t express an opinion because your simulator thinks I use words funny…

And I don’t want to disappoint the Provider.

In electromagnetics, field theory, epsilon and mu are complex values incorporating loss.

We’ve been in agreement on that from the start.

Disperson is the frequency-dependent variation of epsilon and mu.

I’ll quote the IEEE Standard Definitions for Radio Wave Propagation:

• dispersion: (of a wave). The variation of the phase velocity with frequency.
• dispersion relation: The functional relationship between the angular frequency, ω, and the wave vector, k,for waves in a source-free medium.

Later you mention dispersive media, which does mean media where the parameters vary with frequency, but the IEEE explicitly notes that it is called that because such media cause frequency dependence in phase velocity, ie. dispersion.

I’m pretty sure I said there may be some narrow subfield of EM that has expanded the definition of dispersion, but we can keep arguing this if you’d like.

The reason why you don't bother to make a distinction is due to Kramers-Kronig, which, by the way, says your second sentence is completely false.

Yes, Kramers-Kronig says that if you can fully specify either the real or imaginary part of an analytic function over all frequencies, you can determine the other part. This says the real and imaginary components of permittivity (or permeability) are related. That’s different than saying dispersion and loss are the same.

As an apple grows it gets bigger and changes color. The size and color are related and part of the same underlying process, but I’d never say the radius of the apple is red.

Circuit people call it radiation resistance and radiative losses. Fields people don't call radiation a loss.

This is the equivalent circuit representing a wave which could be generated by an infinitesimally small dipole. […] The dissipation in the resistance is equal to the radiation loss in the space problem.

From his derivation. I know what you’re thinking— you’re thinking, “Wrong. That Chu guy doesn’t know anything about high frequency antenna design because engineers who deal with electromagnetic fields don’t use the term ‘radiation loss’”. But again, I'm not the one who brought him into this discussion.

Again, I use the same software Apple uses for antennas and field analysis.

Yes, I did see that. Very impressive!

Again ignoring what I said. In a lossless system, energy is either stored or radiated. Energy cannot be stored with 100% efficiency, so storing energy should be avoided. A simple concept.

Did I really ignore what you said? Go back and look at what I quoted and then responded to. I’m pretty sure you said “stored in the matching network” and that lower Q antennas have higher losses. You’ll also notice that I didn’t start screaming that you know nothing about antennas because of these oversights, I just said that if you insist on keeping score on who’s making what mistakes, those make the tally. You’d round that up to “100% wrong”, I’m merely pointing it out.

I know. Explain that to Analog Kid. It's fairly obvious to me that he has worked in a much lower frequency or limited, guided wave regime of electromagnetics, not "high frequency" antennas. He won't admit his knowledge falls apart in the "high frequency" regime when you push things in cell phones and mm-wave.

You need to admit that you don't know what you're talking about in the domain of electromagnetics. You may know about circuits, but little about E&M and fields. Any knowledge you have is in the low frequency regime and not in mm-wave and above. Thats from the fact that you use tand and not dispersive models.

Can you tell what I had for breakfast?

I’m not sure why you’re stuck on this. I keep looking at the thread title expecting to see “Guess Analog Kid’s Profession”, but every time I find it’s something to do with Apple making antennas... I can't for the life of me understand why you're reading tea leaves to guess what I might know...

Maybe it would help if I point out that “Analog Kid” is the title of a Rush song? Would I have more credibility if I'd picked "Spirit of Radio" instead?

 

[konqerror]

Yes, Kramers-Kronig says that if you can fully specify either the real or imaginary part of an analytic function over all frequencies, you can determine the other part. This says the real and imaginary components of permittivity (or permeability) are related. That’s different than saying dispersion and loss are the same.

You don't know what Kramers-Kronig said. You just read the Wikipedia article, which doesn't apply this to electromagnetics AND you failed to extend it. This is a common theme, again where you seem to know a little but it falls apart when you get into the technical details but continue to argue.

Kramers-Kronig says that the real and imaginary parts of the dielectric function are related. That is, a change in Re(epsilon) results in a change in Im(epsilon). That is, you cannot have index vary over frequency without loss and vice versa.

More specifically, in linear media, the index is the integral of the loss over frequency (with the right units and scaling).

Again, this is why we bundle loss under dispersion. Dispersion simply is the variation of complex epsilon and mu with frequency.

You should have learned this in junior level ECE under a different name, the Bode Gain-Phase relationship. Again, it's taking what you know and extending it. If you don't know that, then you shouldn't be opining on E&M.

And the Chu limit dictates the limits of efficiency of an antenna. Small antennas necessarily suck in efficiency. And you can't match your way out of small antennas.

This repeats over and over. You take an opinion. The opinion is clearly fundamentally wrong to somebody in the field. Somebody explains why you're wrong. You repeat the exact same argument.

 

[Analog Kid]

You don't know what Kramers-Kronig said.

Kramers-Kronig says that if you can fully specify either the real or imaginary part of an analytic function over all frequencies, you can determine the other part.

in linear media, the index is the integral of the loss over frequency (with the right units and scaling).

This says the real and imaginary components of permittivity (or permeability) are related.

Kramers-Kronig says that the real and imaginary parts of the dielectric function are related.

*sigh*

This repeats over and over. [...] You repeat the exact same argument.

And the Chu limit dictates the limits of efficiency of an antenna.

*sigh*

 

[konqerror]

*sigh*

If I didn't know what I was talking about, I'd be out of a job. But again, this is the Internet and winning an argument in order to feel better is more important than learning about cutting-edge wireless or even finding out the truth.