Barrie Gilbert on the single nanotube radio
Analog giant Barrie Gilbert (pdf), inventor of the Gilbert cell, oscillator expert and contributor to EDN as well as an EDN Innovator of the year was nice enough to include me in an email thread about a nanotube radio publicized by the University of California at Berkeley. Since this was a personal email, Barrie wanted to decouple his opinions from that great analog company out in Norwood Massachusetts where Barrie is a Fellow. Barrie writes:
As for that "revolutionary" one-nanotoob radio – I’m reluctant to say a great deal. In the past, when a few crazies were talking about using a paltry 65-nm foundry CMOS to implement radios I felt obliged to point out how hard that would be (heh-heh!). My money was on the use of advanced complementary 200-GHz SiGe BJT processes; and it still is, but perhaps that’s just stubbornness.
However, I have to wring my hands in something like despair whenever I open up the latest issue of say, the JSCC and find scarcely a trace of these magnificent BJT devices. And students of Larson, Abidi, Razavi, Stayaert and many others – not knowing that the bee cannot possibly fly – are making 65-nm radios. It’s not top-flight stuff, but it’s a start.
Now, as for that little chunk of carbon that thinks it’s a radio. As a rule, I’m reluctant to say anything about which I know almost nothing. Here, it is simply what has appeared in the article in Scientific American and a few scant press releases. On the other hand, what I do know about physics (and about the utility of hyperbole to a university) suggests, in my judgment, that there’s something capricious here.
The dots just don’t join up. First, to achieve a useful SNR, an antenna, whether pure-E-mode, pure-M-mode or nicely-proportioned-EM-mode, has to intercept a certain minimum solid angle in space. I highly doubt that all the rules about noise processes are conveniently suspended in the essentially hyper-quantum domain of nanotubes thus dismissing a hundred years of radio design at a stroke. So, naturally, I wonder how a minuscule amount of intercepted RF power can do all the wonderful things claimed of the one-toober - quite apart from everything else we are informed that this gizmo performs, further down its "signal chain".
Of course, the answer to that mysterium is simply that we are not told how much transmitted power is being poured, directly, on to this long-suffering piece of lamp-black. Still, we are quite free to suspect that it represents a very considerable fraction of the lab’s electricity bill. And that being (dontcha think?) an important (and probably significant) bit of the story, I am such a committed unbeliever as to wonder whether the nonlinearities inherent in the audio amplifier used to finally drive a transducer of some sort, plus a bit of resonance in the toob, might be able to account for FM-discriminator-like behavior; and all the gain.
If you take a look at this site it tells you a lot about the discography used to do the demo, but not a thing about the test conditions, supplies or supporting instrumentation.
This leaves me with nothing to do but doubt. I doubt that this little toy, as fascinating though it appears to be, makes a very good antenna. I doubt that anything very subtle is going on in the hinterland – beyond what any good standard treatment of electron devices would explain, augmented by a little imagination and broad-mindedness. And, I very much doubt that the toob directly drives any kind of audio transducer.
This is all good PR for Berkeley, but earlier claims of cold fusion from another quarter are bound to come to mind. And of course, that very question of functional reproducibility – let alone the provision of such essentials as tuning, band-limiting and channel-selection, gain-adjust and controlled detection and/or discriminating processes – has yet to be addressed. For anyone to suggest that here is a "bold, new, fresh approach to the way future radios will be made" would be ludicrous.
My guess at this juncture is that the youngsters doing this nanotube work come from a different mind-set than radio engineers, and have simply stumbled onto a situation in which either some weak effect is being unwittingly exaggerated by, say, an excess of instrumentation, or by faulty inferences as to where all the gain (?) and demodulation are really arising.
I think Barrie may be on to something. There are other mentions of nanotube radios but they use multiple nano-tubes to make transistors and those make the radio. Lately it seems there is such pressure on academicians to publish, that they sometimes rush to judgment. Let’s hope that all of Barrie’s points are addressed so we can judge the validity of this interesting work. And oh, the cold fusion work is continuing apace, as this press release today highlights.
William Ketel commented:
More on the One-Nanotoober commented:
Joe Coppernanotube commented:
Andy T commented:















