Wikipedia:Reference desk/Archives/Science/2013 October 29

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

October 29

Voltage divider.

I'm trying to use a simple voltage divider to allow me to measure a variable resistance that ranges from around 10k ohm down to nearly zero using an Atmel microcontroller (like in an Arduino). The circuit is the one shown at right (from our Voltage divider article - which I've read).

Vin is 5 volts and the analog-to-digital converter is connected to Vout and measures 0 to 5 volts with a precision of one part in 1024. I'm trying to figure out whether I should make R₁ or R₂ be the variable resistor - and what value to use for the other (fixed) resistor. From the math:

V_out/V_in = R₂ / (R₁ + R₂)

Which suggests that if R₁ is the variable resistor, then R₂ should be very small in order to use the entire range of the AtoD converter. (For example, if R₂ is 10k ohms - then all of my results will be in the range 2.5v to 5.0v and I'll be wasting one whole bit of the result...but if it's just 10 ohms, then I'd be using almost the entire range of the AtoD. I'm betting that something will go horribly wrong if I make R₂ be as small as 10 ohms. What's the gotcha and what's the recommended value?

(I'm not really interested in investigating more complicated circuits - I'm engineering down to a price here!).

SteveBaker (talk) 02:57, 29 October 2013 (UTC)

Wouldn't it be simpler to use a three terminal variable resistor, with one end of the resistor connected to +5 V, the other end to ground, and the wiper connected to the A/D converter? This would assure a low (0.5 mA) current through the resistor and a negligible current through the wiper. Jc3s5h (talk) 03:14, 29 October 2013 (UTC)

If I was using a potentiometer - yes. That would be nice because R₁+R₂ is a constant and V_out is directly proportional to R₂ - but I'm not. There is just one external variable resistance that I'm measuring and R₂ is fixed. (Although if there was a good reason to do it, I could make R₂ be the variable one and R₁ be the fixed resistance. SteveBaker (talk) 03:39, 29 October 2013 (UTC)

On the "betting that something will go horribly wrong", you need to watch the potential power dissipation in each of the resistors, e.g. 0.5 A and 2.5 W with your value of 10 ohm. That's a pretty hefty resistor, and may cost a few cents more. If your variable resistor can dissipate less power than the fixed resistor, it may overheat when it is at a similar value. Another gotcha is that over most of the range of the variable resistor (> 10 ohm), your resolution is going to be far worse than if you use a larger fixed resistor. You should plot the resolution over the whole range, and how it is spread over the range compared to where you want it. For example, you may want substantially more resolution at lowish values of resistance at the expense of at the high resistance end, in which case you could use a lowish value (e.g. 1 kiloohm, 90% of your ADC range, quite nonlinear; 10 kiloohm: substantially more linear, 50% of the ADC range). In your diagram, which resistor is which makes no real difference: the results are simply reversed, non-linearities and all (subtract from 1023). — Quondum 03:48, 29 October 2013 (UTC)

Another, slightly more complicated way of doing it (I know you said you're not interested, but it's still cheap) is to convert the resistance into a delay and measure that. 555 timer IC#Example applications gives a good example of how the dirt-cheap and ubiquitous 555 timer (or similar methods) were used to measure the position of joysticks, which produced a variable resistance and had to be read with a digital device. Basically, you charge a capacitor and the resistance determines how long it takes to charge it. You can time that and know what the resistance is. The accuracy isn't great, but it is simple to set up in a way that doesn't damage components with too much current, which can be a serious problem for reading the whole range with a voltage divider. Katie R (talk) 11:47, 29 October 2013 (UTC)

The key thing the 555 brings to the table here is the ability to send a trigger pulse then wait for a response using all digital I/O. Since you have an analog input, you could just watch the voltage of the capacitor directly as it charges from a digital output if you want to avoid the chip. Katie R (talk) 11:49, 29 October 2013 (UTC)

An important piece missing is tolerance and accuracy. If your fixed resistor is 5% but 10k, you can forget being accurate at the low end. You would already be bit limited 10 ohms/bit, but also +/- 500 ohms (50 bits) for resistor tolerance. Without knowing your tolerance for error and where you need accuracy, it's not very tractable. Other alternatives are to use a equal voltage divider to supply the reference voltage to the A/D (zero it out) and then have the same circuit with the DUT in parallel to R2. The more you can ratio out error, the better your accuracy. With just the voltage divider, resistor tolerance, temperature and device selection will all impact your measurement and at the low end, it will be useless. You always want to be close to full scale.--DHeyward (talk) 12:52, 29 October 2013 (UTC)

You know more about this than I do, but my thought is that whatever you're using this for, wasting power is always bad, so you want as high a value as possible for R1 so that no matter what goes wrong with the rest of the circuit, you know your power isn't peeing away through it. You then want your other variable resistor to also be as high as possible so that you can get nearly the full range by turning it up higher than R1. Which gets us to the brass tacks: what is the resistance of your voltmeter really? A hypothetical voltmeter passes zero current, so you could use a chunk of wood for R1 and R2 and move the probe needle back and forth on it. Unfortunately this is not a purely hypothetical voltmeter, so you need R1 << maximum R2 << Rvoltmeter. Wnt (talk) 16:35, 29 October 2013 (UTC)

If you want to use a simple voltage divider, I would choose two fixed resistors, with R₁+R_var in place of R₁, R_var being the variable resistor now. V_out/V_in = R₂ / (R₁ + R_var + R₂). With R₁= 100 R₂ = 1000, R_var between 0 and 10k, Vout will be between 0.45 and 4.55V (assuming 5V input). Absolute precision goes down when the variable resistance rises, but the relative precision will be highest when R_var equals R₁+R₂. Example: Voltage difference between R_var = 100 and = 110 is 0.0344V, between 2000 and 2200 it's 0.098V, between 8000 and 8800 it's 0.044V. Ssscienccce (talk) 08:36, 30 October 2013 (UTC)

Replace R2 with alow value capacitor, make Vin come from an output of the u controller and use software (free) to. work out the restance:)--31.55.113.25 (talk) 22:28, 30 October 2013 (UTC)

Questions about Dietry Fibers

1) Why fibers are not considered Essential nutrients?

2) What is the main source of the most coveted source for insoluble fiber by humans?

3) Is there such a concept "correct ratio between soluble and insoluble fiber in a human's diet?"

Thanks very much for an answer. 109.65.50.176 (talk) 05:48, 29 October 2013 (UTC)

I numbered your Q's for easier response:

1) A nutrient is something you digest. Since fiber is not digested, it's not a nutrient. At least that's true of non-nutritive fiber. StuRat (talk) 21:59, 29 October 2013 (UTC)

What you've said is pretty much correct, but in my opinion nutrients don't *always* need digestion. Depends on your definition of digestion. If you define digestion to be any combination of "ingestion, breakdown, absorption, and egestion," then fiber could be included in digestion. If it is enzymatic breakdown only (which some sources indicate as "digestion" per se), then vitamins, minerals, and water don't qualify as being digested because they don't need that enzymatic breakdown before absorption. If you take it to mean ANY sort of breakdown, then drinking a glass of water and absorbing it would not include any sort of digestion. (However, you could make the argument that either water is NOT a nutrient or that's a unique case.) Doctorcherokee (talk) 18:13, 1 November 2013 (UTC)

Wouldn't retrograde amnesia erase a person's mental illnesses too?

If an adult with several mental illnesses or emotional problems and specific fears had an accident and woke up in the hospital with no memories at all like that was their first second being alive, wouldn't their mental illnesses, emotional problems and specific fears be gone too? — Preceding unsigned comment added by 174.65.23.49 (talk) 07:03, 29 October 2013 (UTC)

No. There are multiple problems with that idea, one of which is that mental illnesses aren't purely based on what's in a person's episodic and declarative memory. See Causes of mental disorders. Red Act (talk) 07:57, 29 October 2013 (UTC)

Life is not like a video game. We can't reset back to an earlier time. Mental disease will no more disappear just because of forgetting than grey hair will turn black. Dmcq (talk) 11:08, 29 October 2013 (UTC)

What would or could be forgotten is whatever was experienced previously. Also, is the hypothetical person able to speak and do other things normally? ←Baseball Bugs ^{What's up, Doc?} carrots→ 11:17, 29 October 2013 (UTC)

It depends on the extent of the damage that caused the amnesia. Different areas of the brain cover different functions, so damage to one (for example, to the ability to recall memories) would not necessarily cause problems for other areas of the brain (for example, language processing or physical coordination). --Jayron32 12:15, 29 October 2013 (UTC)

• The answer is a bit complicated, and involves several factors.

1. It is common in movies and TV shows for somebody to wake up with no memory whatsoever, but that almost never happens in real life. Genuine retrograde amnesia eliminates memory for the recent past -- sometimes extending back for years -- but almost always leaves memory from childhood intact. There are a few exceptions, but most of them have dubious features -- in some cases it is not clear whether people are faking TV-style amnesia in order to avoid answering awkward questions.
2. The idea that all, or most, mental illness results from traumatic experience is a relic of Freudian theory that is now rejected by the great majority of psychiatrists.

So the bottom line is that (a) that doesn't happen, and (b) it wouldn't eliminate most types of mental illness even if it did. Looie496 (talk) 16:59, 29 October 2013 (UTC)

Meck, Su has this kind of retrograde amnesia and though it is rare she isn't the only person to have a complete loss of all memories. Baseball_Bugs, in this case the person would be like an infant and would have to be taught everything again. — Preceding unsigned comment added by 174.65.23.49 (talk) 02:30, 30 October 2013 (UTC)

I remember that happened to Lt. Uhura once. They managed to re-teach her everything by the very next episode! --Trovatore (talk) 02:35, 30 October 2013 (UTC)

I'm sure Kirk volunteered to show her where she sleeps at night. StuRat (talk) 02:40, 30 October 2013 (UTC)

I looked up the Su Meck case, and while it's mostly popular-press stuff, it is very much an impressive story. --Trovatore (talk) 02:53, 30 October 2013 (UTC)

Even in cases where mental illness might be caused (or triggered) by trauma, erasing the memories of the traumatic experiences would not erase the mental effects of the traumatic experiences. My cat doesn't have to specifically remember being sprayed with water to react to the sight of a spray bottle. thx1138 (talk) 18:07, 31 October 2013 (UTC)

Side-stick controls in cockpits

This is just a quick question: when Airbus designed the Airbus A320, the first fly-by-wire airliner, what were the reasons why they decided to use side-stick controls (which, at the time, was probably unheard of for commercial aircraft) instead of the conventional yokes, and what were the reasons why Boeing decided not to use side-stick controls in the Boeing 777 and Boeing 787 Dreamliner? Also, what are the advantages and disadvantages of using side-stick controls in commercial aircraft cockpits? I have read our Wikipedia articles for yokes and side-sticks, but they do not elaborate on the information I want answered here. Narutolovehinata5 ^tccsdnew 10:39, 29 October 2013 (UTC)

As a pilot that flies small planes, the Cirrus SR22 vs Cessna 210 (and both are mechanical, not fly-by-wire), the sidestick is out of the way of the display and doesn't feel all that different from the left seat (captains seat) than a traditional center yoke. There is also the feeling of more room. The downside is that the left/right seat switches are reversed (some would be anyway for throttle control while PTT). An obvious advantage of the center yoke is either hand can fly and either pilot can see and reach the other yoke. The requirements for the quick donning mask, for instance, might be easier to accomplish with this feature. Even the fly-by-wire yokes have mechanical feedbacks to mimic what would be felt in a traditional mechanical yoke so the transition is easier like the buffet before a stall is mimicked (I'm not sure about side-stick though). Personally, I like the Cirrus side stick just for the added space. Side stick fighter planes normally have the stick on the right with throttle control either on the stick or left side so opposite the captain's seat. In the end though, it comes down to transition time. If everything looks and works the same and the only difference is fly-by-wire vs hydraulic, a transition will be much easier. If there a number of differences, like how/when the autothrottle engages/disengages or where switches are located (stick or console), transition time can be longer, more expensive and possibly less safe. I suspect the yoke and controls on the yoke, the travel distance for rotation, bank angle for standard rate turn, etc, all match previous Boeing airplanes and makes transitioning easier and quicker (cheaper). The crash in San Francisco recently was attributed initially to a small difference in autothrottle behavior between types. Small things that change can have large consequences even if the change appears good on paper. --DHeyward (talk) 12:04, 29 October 2013 (UTC)

In jet airliners like Airbus and Boeing, the pilots actually spend a very small percentage of the flight with hands on the control - most of the time the aircraft is being flown by the autopilot. Pilots spend a much greater percentage of their time monitoring instruments, monitoring information displays and doing things with documents - hence the attraction of the Airbus strategy of providing a stick at the side of the pilot and a tray table directly in front. All flight controls are fully powered by servo motors so a large stick with a large moment arm is not necessary. And why has Boeing stayed with the conventional control yoke? Probably because Boeing is the market leader so it doesn't need to try new strategies as Airbus does in an attempt to establish its product and then increase market share. Also, Boeing places great importance on what its established customers say and want in their future aircraft. It seems likely to me that when Boeing asks its established customers and their pilots, all say they are entirely happy with what they have at present. Dolphin (t) 12:23, 29 October 2013 (UTC)

It is worth noting that the most important differences between the Boeing and Airbus fly-by-wire control really aren't in the shape and position of the yoke or side-stick, but rather in the way that the aircraft's computer interprets, responds to, and feeds back pressure on the yoke or stick. There is much information and discussion (correct and otherwise) available in online fora; see for instance [1]. (You will find many online flamewars about flight envelope protection philosophies of Boeing versus Airbus.) The Airbus system has a number of unambiguous benefits from a design and economics standpoint—it is lighter and mechanically simpler; it seems to be ergonomically preferable (less tiring for pilots, room for a tray in front of the pilot, no problems with controls obstructing the view of instruments); removing the yoke allows the cockpit to be slightly shallower from front to back.

The safety question is difficult to settle. What you will find fairly readily are a small number of edge-case incidents and crashes which each side's proponents like to cite over and over, saying "in this tiny one-off case, our preferred technology could have/did save(d) the day". TenOfAllTrades(talk) 16:55, 30 October 2013 (UTC)

Animal learning by experiences of other animals without direct exposure themselves...

It may sound like a confusing or misleading title, but what I am getting at is the instance of having an animal experience some sort of distress by a specific type of predator that looks a particular way. Instead of reacting by instinct, the animal that experiences the stress passes on the learning or conditioning by communication or language, no matter how rudimentary it is. That way, successive generations of the same species can learn who to avoid rather than relying on instinct all the time. In a similar story about lyre birds, if I remember correctly, young male lyre birds can learn the songs of older lyre birds and imitate them. I am wondering if there are any more specific cases in the animal kingdom, but are more sophisticated/complex than the ones I've described and are occurring in nonhuman species. I already know that humans can do it. So, please don't list homo sapiens or any member of the homo genus unless it is an exception. 140.254.227.60 (talk) 13:32, 29 October 2013 (UTC)

It has been shown shown that primates do not have an inborn fear of predators such as snakes but they learn this behaviour from other primates. However, although they can be conditioned in the lab to fear predators by showing them videos of other members of their species showing a fear response, the same process won't necessarily condition them to fear other objects. As it says here "When videos were spliced so that identical displays of fear were modeled in response to toy snakes and flowers, or to toy crocodiles and rabbits( (M. Cook & Mineka, 1991), the lab-reared monkeys showed substantial conditioning to toy snakes and crocodiles, but not to flowers and toy rabbits". So they suggest that there is also an evolutionary component to the selective learning. Richerman (talk) 14:23, 29 October 2013 (UTC)

Concidentally or otherwise, the information you're stating was covered in an NPR discussion just yesterday. :) Which means there could be a link to it on their website, if someone wants to hear more about it. ←Baseball Bugs ^{What's up, Doc?} carrots→ 23:19, 29 October 2013 (UTC)

Are you going to link the article? 140.254.226.247 (talk) 16:18, 30 October 2013 (UTC)

We have a moderately lengthy article on observational learning, with some material about animals. A Google Scholar search for "observational learning animals" will give you a lot more material -- for example, rats learning which foods are good to eat by observing other rats. Looie496 (talk) 16:44, 29 October 2013 (UTC)

There was a study of crows which seemed to show that they recognize specific humans and pass a fear to those individual humans on to others. The people wore masks which made identifying them a bit easier, though. StuRat (talk) 22:05, 29 October 2013 (UTC)

Might be the masks made them scarier. As this BBC thing I just read suggests, humans are naturally freaked by not-quite-human faces, due to the expectation of a regular one. Crows have lived among us as long as we have, so it seems (to me) they could have the same issue. Might be why scarecrows exist. InedibleHulk (talk) 06:04, 30 October 2013 (UTC)

One of my psychology students had trained a rat to press a lever repeatedly to get a reward, and wanted to see if a "student rat" could learn to press the lever to get the reward just by watching the "teacher rat" do the act. The result was that Student rat just hung around by the water dipper to enjoy the reward which was earned by the Teacher rat. The Student did not learn that pressing the lever was what brought the reward. Edison (talk) 04:24, 30 October 2013 (UTC)

And did they call this Student Rat a StuRat, for short ? :-) StuRat (talk) 17:07, 31 October 2013 (UTC)

Does anybody know how many citations an article has been cited?

Does anybody know how many citations an article has been cited? I know there is a function on some website that allows people to track how many people have cited the article in their own articles as part of their references/citations/bibliographies. By the way, have there been documented cases where scientists harshly criticized their rivals' work, where both scientific teams are engaged in the same field and both want the glory of making some important distinguished landmark? I seem to recall a story in the field of chemistry about some guy who discovered Helium and then lost to the guy who discovered Hydrogen... or was it the reverse? I forget. 140.254.227.60 (talk) 14:19, 29 October 2013 (UTC)

To answer the first part of your question, the number of times a scientific publication is cited is the main criterion used to decide its quality and it is something academics are obsessed with. There are a number of ways find out see:[2] for example or google "how many times has my article been cited" As for the second part - scientific rivalry often spills over into harsh criticism of one anothers' work, and the awarding of the Nobel prizes are often a great source of controversy as to whether the recipient really should have got the credit - see Nobel Prize controversies. Richerman (talk) 14:37, 29 October 2013 (UTC)

I beg to differ that it is not academics who are obsessed with citation count, but the people who evaluate academics (for allocating university rankings, research funding, etc). I'm sure most academics would happily not care about such dubious measures but are forced to if they are to survive. 31.52.246.217 (talk) 23:17, 30 October 2013 (UTC)

• If you look up an article on Google Scholar, it will show you the number of citations and allow you to list the citing articles. The number is often not completely accurate -- if it is important, you should check through the list to make sure that each entry is legitimate. Looie496 (talk) 16:48, 29 October 2013 (UTC)

Google Scholar is a useful tool but it has the drawback that it only includes journals that have been indexed online. In some fields this can omit the majority of publications. SpinningSpark 17:27, 29 October 2013 (UTC)

The old standard was ISI Web of Science, but in some way that I don't understand they made themselves so much of a pain in the ass for so many researchers to use (both in terms of interface and who was allowed access from where) that for a while they've seemed to have become more obscure. Wnt (talk) 21:11, 29 October 2013 (UTC)

Fighting scientist? doi:10.1029/2011JE003880 against doi:10.1029/2010JE003599. Or an old example: The discovery of thallium doi:10.1098/rsnr.1984.0005. The favorite one is Cassiopeium and Aldebaranium are no elements because the French chemist was heading the naming committee and not the Austrian chemist. The bloody fight about polywater and the devastating resistance about plate tectonics are other examples.--Stone (talk) 12:30, 30 October 2013 (UTC)

If your German is good enough just read what Hermann Kolbe wrote about his colleagues in his journal (one example of many is doi:10.1002/prac.18820260121).--Stone (talk) 12:33, 30 October 2013 (UTC)

My German is extremely poor. A translation of the document or a summary written in English would be helpful. Thanks. 140.254.226.247 (talk) 15:55, 30 October 2013 (UTC)

"...Adolf von Baeyer is an excellent experimentor, but he is only an empiricist, lacking sense and capability, and his interpretations of his experiments show particular deficiency in his familiarity with the principles of true science..." To make it clear 1905 Baeyer was awarded the Nobel Prize in Chemistry.--Stone (talk) 19:01, 30 October 2013 (UTC)