Charcoal and diamonds are both pure carbon. As Ecnerwal says, they differ only in structure - in charcoal the carbon atoms are organised in a basically two dimensional array as they are in graphite, carbon nanotubes and the fullerenes, while in diamond the atoms are organised into a regular three dimensional array.
-- Bill Sloman, Nijmegen
Modified drawing
See the modified drawing. It is an LM317
On Jan 6, 4:52 pm, James Arthur wrote: [....]
At any reasonable current nearly any bipolar will have a lower noise than the LM317.
If you'd asked me I would have said 500nV/sqrt(Hz) so he must be using a "low noise" one.
Bypassing the lower resistor of the LM317 improves stability. The best way to do it looks like this:
------- -----+----------+----! LM317 !------+----+-------+--- ! ! ! ADJ ! ! ! ! BIG === 0.1 === ------- 0.1 === \\ BIG === ! ! ! ! / ! GND GND ! ! \\ GND +------+---+----+ ! ! ! \\ 0.1 === === BIG / ! ! \\ GND GND ! GND
The 0.1s are right at the LM317's package. The BIG capacitors are physically large but to the degree it is practical, they are also close.
The LM317 likes working into a capacitive load. So long as the Vin side doesn't see too high of an impedance and the Vout to ADJ looks capacitive it is happy.
On Jan 6, 6:47 pm, Robert Baer wrote: [....]
Add one extra bend that is not obvious to the reviewer and you'd get it.
Since the noise bypass cap is connected to highish-value resistors, it's safe to use a tantalum there. And since that cap soft-starts the regulator, it becomes safe to use another tantalum as the output cap.
I think.
John
I think he may have been asking about the op-amp.
Ed
If you want a hoot, check out patent nos. 3,811,085; 3,879,622; and
4,151,431. These are perpetual motion machines! But, the patents are written in a very perverse gobbledegook way that makes it virtually impossible to determine what is going on.Some patents are designed to bilk investors and have no basis in reality and there is no requirment of that these days. Other patents, patent the obvious but the present day patent office may or may not catch it depending on the examiner and how well the inventor and his councel answer the questioning or denile of claims.
Many absurd patent go through because the present patent office is more interested in money than invention.
What you described is like a patent that I hold on a capcitive signal pickup device. It basically is a capacitor described in detail with about 20 claims. I though it absurd at the time when we applied for a patent, later grated, because it seemed so obvious and rediculous. No, I really didn't invent the capacitor but hold a patent on one none the less.
This one?
I had a subscription to NTB once, and I don't remember having had to pay for it - our tax dollars at work! ;-)
I once worked at a joint where one of the things their product could do was "molecular beam epitaxy". One day, I asked some of the PhDs and other high-level dreamers something like "Well, what happens if you use a beam of carbon ions?"
They all looked kind of baffled - sort of a blank look on their face, like they couldn't even comprehend the concept, or like an alien who's just been invited to bowl a few lines or something.
But with a suitable substrate (boron nitride?), it seems quite straightforward to me.
Charge up the BN to maybe -100,000 volts, and shoot positive carbon ions at it.
Of course, all you need for the experiment is a $750,000.00 epitaxy chamber. ;-)
Cheers! Rich
Wrong.
COAL is pure carbon (mostly). Graphite is pure carbon.
Charcoal is some damned mostly burned tree substance.
Yes, and note where that page NEVER mentions "charcoal".
- Don Klipstein ( snipped-for-privacy@misty.com)
We now stand by for COTB's explanation why he was still not incorrect, in spite of quoted facts to the contrary.
Sounds like a hell of an expensive way to get crystalline graphite, or worse, plain soot.
That's interesting, conceptually. One disadvantage is that you're relying on changes in the output voltage itself to detect, amplify, and cancel noise in that very signal. That is, you're reacting, i.e., ex post facto.
By adding his small sense resistor Mr. Wenzel gets ahead of the game, sensing the noise current *before* the output cap. can integrate it into a voltage. IOW, he gains 90 deg. of phase margin, preventing instability.
Quantatively, I read Walt Jung measured the LM317 at 120nV / sqrt(Hz) (measurement frequency unknown). Wenzel's page reports 20dB improvement with his untrimmed circuit, 40dB when trimmed.
If we assume for the moment that heavily bypassing the LM317's lower divider resistor could cut its noise gain to unity, and that Wenzel's
20dB improvement was still possible on such an LM317 circuit, that would yield about 12nV / sqrt(Hz) overall. Or, as John pointed out, you could simply do this:Cheers, James Arthur
P.S. Graham, you might enjoy this Jung article on current sources for audio use, if you haven't already seen it:
-ja
Right, that's what I was pointing out--the transistor is much quieter, so the noise-eater circuit offers high potential improvement over the LM317, bypassed or not. Wenzel puts his 2n4401's noise voltage at roughly 1nV / sqrt(Hz) at 100 Hz.
Right, looks fine except: bypassing the upper divider resistor--why? Isn't that counter productive?
Cheers, James Arthur
[...]
LDOs are stable? ;-)
Nice technique. In fact that pretty much is a bootstrapped reference. I have an old HeathKit power supply that starts with +100V and cascades zener shunt regulators, each feeding the next (for stability), to develop its +36V reference. That's how old iron did it. Kind of cool. Er, hot, actually.
Grins, James Arthur
The process pf making charcoal involves burning some of the wood in a deliberately insufficient air stream to get it hot enough pyrolyse the rest.
What is left has a high surface area per unit weight.
I was interested in establishing the point that carbon has two alternative atomic structures, rather than getting into the fine detail about the source and structure of charcoal.
"Activated charcoal" is rather more interesting than regular charcoal (and is probably what Ecnerwal had in mind).
-- Bill Sloman, Nijmegen
Yeah, I suppose anybody who can afford equipment like that to just play with probably already has all the diamonds they need. ;-)
Cheers! Rich
One more step: use a good low-noise reference into the opamp error detector and connect the opamp output directly to the 317's adjust pin. That effectively disables the 317's mediocre internal bandgap reference and uses the 317 as a nice current/thermal limited power stage. That will give an espacial improvement in very low frequency noise, where the bandgap is pretty bad.
An LM317 (or better yet LM1117) is a mediocre regulator but makes a nice power amplifier.
Intersil has some very interesting voltage references. They are a capacitor connected to a low-bias-current opamp; they charge it up to the right voltage at the factory, and ship it.
John
On Jan 8, 11:50 am, John Larkin wrote: [....]
Unfortunately there really is no such thing as a "low-noise reference" There are really noisy references and less noisy references. None I know of get down to the 1nV/sqrt(Hz) noise level I want.
[.....]It is interesting other than the huge amount of noise it makes.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.