Science desk
< August 29	<< Jul \| August \| Sep >>	August 31 >

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.

August 30

determining probabilities from rate laws

For something like radioactibe decay, the rate law is ${\frac {dN}{dt}}=-\lambda N$ . and for a general chemical reaction it is usually something like ${\frac {dN}{dt}}=kN^{x}$ (see rate law). Is it possible to deduce from the rate laws the probability that a reaction occurs after a particular elapsed time? 130.56.84.66 (talk) 02:20, 30 August 2013 (UTC)[reply]

The answer is essentially "yes, but...". the k term in the rate law you quoted implies an constant. In practice for chemical reactions it is a complex function of temperature. The temperature dependence is often modelled by the modified Arrhenius equation. Reactions can be considered as two sorts: a) reactions that are single step, or are time dominated by one of several steps, and b) reactions that comprise two or more steps where none dominates. Reactions of the first type show an exponentially decaying rate easily (trivially) determined from the rate equation. Reactions of the second type typically show a delay period in which not a lot happens, followed by a relatively sudden or dramatic increase in the production of product (and heat). Combustion reactions are of this sort. A "simple" example is the reaction of hydrogen and oxygen. The final stable product is water vapour, but there are several steps on the way involving over 120 simple reactions between 6 intermediate species such as OH, HO2 etc. To determine both the initial delay and the subsequent reaction rate, you need to solve the reaction rate equations for each step as a set of simultaneous non-linear equations. You probably learnt how to manually solve simultaneous linear equations in high school, using two and three equation examples. If so, you'll know that the amount of work rises very rapidly with the number of equations. To do this for real chemical reactions can take a LOT of computer time. And I do mean a lot. For working it out for say the combustion of gasoline, you need a super computer grinding away for weeks.

A practical issue is that the constants you need to use in the modified Arrhenious equation to calculate k are often not known for the reactions you are interested in. And considerable laboratory ingenuity is required to measure them.

Much the same applies to radioactive decay, however with radioactive decay unlike chemistry, there is quite often single step or single time dominant step. Even when there isn't, there are far fewer component reactions for each intermediate species than is usually the case in chemistry. So the calculation of radioactive decay is generally easy, not requiring a lot of computer time.

124.178.41.65 (talk) 03:40, 30 August 2013 (UTC)[reply]

Can you tell a person's race through genetics?

Through molecular genetics can you tell a person's race? Say if a person is black, white, native can it be found through genes? — Preceding unsigned comment added by 74.14.31.85 (talk) 03:42, 30 August 2013 (UTC)[reply]

Physical characteristics would be defined by the genes, sure. See Phenotype. ←Baseball Bugs ^{What's up, Doc?} carrots→ 03:52, 30 August 2013 (UTC)[reply]

But how can you tell the person's race, isn't there more to race like IQ averages? Wasn't that proven to be almost completely hereditary? — Preceding unsigned comment added by 74.14.31.85 (talk) 03:56, 30 August 2013 (UTC)[reply]

Why was this question removed, if I say something wrong please correct me? — Preceding unsigned comment added by 74.14.31.85 (talk) 04:05, 30 August 2013 (UTC)[reply]

Race and intelligence are unrelated, as you well know. ←Baseball Bugs ^{What's up, Doc?} carrots→ 04:30, 30 August 2013 (UTC)[reply]

^{[citation needed]} 24.23.196.85 (talk) 05:58, 30 August 2013 (UTC)[reply]

Citation needed? See the discussions below. You're obviously a race-baiting troll. I don't know why this thread is being allowed to stand. ←Baseball Bugs ^{What's up, Doc?} carrots→ 13:40, 30 August 2013 (UTC)[reply]

Actually, Wikipedia has an article on this subject: Race and intelligence. There is certainly a correlation between race and intelligence. As always, correlation does not equal causation. Furthermore, calling people "trolls" is most uncivil. Race, per se, can be a very sensitive issue, but that doesn't mean we can't talk about it in academic discourse. Therefore, if you want to engage in an discussion about race and intelligence, you have to provide references to support your point. The OP's question is concerned with the idea that a person's race can be determined by genetics. In reality, a person's race can be determined by many attributes, including place of origin. Now, if individuals live together, they may mate among each other and thus produce similar genes. Therefore, race may be attributed to place of origin or common ancestry. Perhaps, a similar scenario would be dog breeds. 164.107.102.219 (talk) 21:35, 30 August 2013 (UTC)[reply]

The OP's question implies you could discern the race in part by discerning the presumed intelligence level in their DNA (if such a thing is even measurable). That's a nakedly racist idea. And thoroughly bogus. There are lots and lots of very smart black people in this world, and lots and lots of very dumb white people. And genes aren't going to contain whatever "average" a race allegedly has, they are going to contain whatever level that individual has. Intelligence is in individuals, not in races. ←Baseball Bugs ^{What's up, Doc?} carrots→ 12:03, 31 August 2013 (UTC)[reply]

And, in many peoples' eyes, black, white and native aren't races. HiLo48 (talk) 06:10, 30 August 2013 (UTC)[reply]

Here in the Eastern Hemisphere, 'black' and 'white' are approximate descriptions of the people who are 'native'. AlexTiefling (talk) 09:02, 30 August 2013 (UTC)[reply]

Race is partially genetic, but it's also partially cultural. So, identical twins might both be determined to be the same race, genetically, but, if raised in different cultures, might self-identify themselves as different races, and might appear to be different races to everyone else. For example, whether a black person wears their hair in an afro, or straightens it, will greatly affect how they are perceived. StuRat (talk) 06:14, 30 August 2013 (UTC)[reply]

The short answer is "no". A longer answer can be found by reading our article on race and genetics which says:

In everyday life many societies classify populations into groups based on phenotypical traits and impressions of probable geographic ancestry and socio-economic status - these are the groups we tend to call "races". Because the patterns of variation of human genetic traits are clinal, with a gradual change in trait frequency between population clusters, it is possible to statistically correlate clusters of physical traits with individual geographic ancestry. The frequencies of allelles tend to form clusters where populations live closely together and interact over periods of time. This is due to endogamy within kin groups and lineages or national, cultural or linguistic boundaries. This causes genetic clusters to correlate statistically with population groups when a number of alleles are evaluated.

So if you take a large enough population there may be some statistical correlations between genotype and geographic ancestry, but at the level of individuals the answer is still "no". Certainly there is no genetic definition of "race". Gandalf61 (talk) 08:28, 30 August 2013 (UTC)[reply]

You can reasonnably reliably identify someone's ancestry through genetics. Some genes are more common amoung certain populations, so you can make a good guess of where their ancestors used to live if you find several of these population-specific genes. --Lgriot (talk) 08:58, 30 August 2013 (UTC)[reply]

Really ? So how does that work ? If there is an allele X that is very common in population Y and very rare outside of population Y then you can conclude that someone with allele X probably has at least one ancestor from population Y - but that could be a single lone ancestor. This tells you nothing about the overall span or distribution of their ancestry. Gandalf61 (talk) 09:36, 30 August 2013 (UTC)[reply]

Here is some data, as of 2008: [1]

"Using only genetic data, the researchers were able to assign, on average, 50% of European individuals to within 400 kilometers of their correct country of origin. But there was one caveat: all four grandparents of an individual had to come from the same European country for the assignment to be correct. People with mixed European ancestry tended to show up between the locations of their ancestors."

A thing to remember is that you're not just looking at one allele, you're looking at tens of thousands of alleles. So statistical correlations do come into play, even for DNA taken from a single individual. Jheald (talk) 10:08, 30 August 2013 (UTC)[reply]

With more complicated ancestries, you can estimate where each part of each chromosome is likely to have originated geographically. Here is what 23 & Me reckoned they could achieve in 2012 [2] -- scroll down about two-thirds of the page for their estimates of where each part of each of the two copies of one customer's Chromosome 2 originated. Jheald (talk) 10:22, 30 August 2013 (UTC)[reply]

23 & Me makes money by selling DNA test kits, so they are not a reliable source of information on this. This article from the Telegraph reports the views of academics on genetic ancestry tests and concludes that they are "genetic astrology". Gandalf61 (talk) 10:38, 30 August 2013 (UTC)[reply]

I'd tend to trust the data and the science and what's been published by scientists actually working in this area, rather than a few rent-a-quotes rounded up by the Telegraph.

When you can see with your own eyes the clear geographical clustering in the principal-components scatterplots, it is fatuous to deny that that clustering exists.

The algorithms 23 & Me are running for the chromosome plots are very standard data-mining changepoint detection methods, and the results are pretty much exactly what you would expect. They make sense, biologically and genetically.

The weakness is that so far they have very little comparison data for genomes from outside Europe, apart from the specific individuals in the 1000 genomes project. So their identifications of the origins of non-European parts of the chromosomes are really rather vague at the moment, and should only be considered provisional -- in time, with better reference data, it should be possible to be much more specific; and it may turn out that some of the "nearest best guess" from current data may not be quite right.

In the media article, I can understand distaste at some of the rather strident hype from people like Brian Sykes or Alistair Moffat encouraging people to identify with a "tribe" according to which Y-DNA Haplogroup they belong to, and then identify with Richard III or whoever -- when this only accounts for one part of their ancestry, and may have come into the country in many different ways. I can see why somebody might challenge that this is no more significant than somebody's star-sign. But on the other hand, if somebody in the UK has particular forms of a Haplogroup I Y-chromosome, one can say with pretty much certainty that that mutation originated in Scandinavia and was brought over by the Vikings. That only gives one line of ancestry, but go far enough back to a time when there was much less population migration for much of the population, and it probably does indicate that a grandfather or great-grandfather or great-great-grandfather came from a part of the country that had a considerable Viking ancestry. Similarly, if you look at enough Y-STR markers, there may be a particular detailed pattern that indicates that that chromosome came from a particular Ashkenazi Jewish ancestry; or a particular lineage from the Indian subcontinent. Yes, there were mistakes made when people over-claimed too early, using too few markers, before it was better known how widespread or not those combinations were in particular populations. (Some of the early claims made for the so-called Cohen Modal Haplotype based only only 6 markers, rather than up to 37 or 67 that are typical today, are such an example). But we now have that data, so we now know just how specific a particular combination of markers is.

With autosomal data, we are not at that stage yet for non-European populations. We don't yet have enough data from enough different parts of the world outside Europe to be able to definitively say in detail how geographically widespread particular values for particular genetic markers may be, and in just which populations they may or may not be most common. So this is why the ancestry assignments outside Europe for parts of chromosomes are pretty vague, and for the moment should be regarded only as provisional. But they will inevitably get better, more detailed and more certain, as more reference data comes in. The underlying science is sound. Jheald (talk) 11:52, 30 August 2013 (UTC)[reply]

This guide] from Sense About Science is a detailed objective critique of the marketing claims of 23andMe and similar commercial organisations. "Genetic ancestry tests use some techniques that have been developed by researchers for studying differences in DNA across many groups of people. The things we know about genetic ancestry, almost without exception, are about the genetic history of whole populations.Companies use techniques from this field and sell their findings to people who want to find out about their personal history. The techniques were not designed for this. The information they give is not unique to any individual." Gandalf61 (talk) 12:27, 30 August 2013 (UTC)[reply]

I think it's a fair message that you need to have your eyes open to the limitations of what a genetic test can tell you. And you should know exactly how high resolution a test you're getting. There are certainly inflated claims being made to tell you "your ancestry" on the basis of just quite a low-resolution test -- perhaps just the Y or mtDNA haplogroup; so the comments on the limitations of a test like that, and the generic un-personalised people might get back from perhaps just a dozen standard scripts (which may also perhaps be rather fanciful) are entirely on-point. Perhaps that isn't much more meaningful than a star-sign. But at the same time, that's not necessarily all one can learn, even from just the Y or mt-DNA.

On Y-DNA and mtDNA testing, they don't really say anything different from what I've said above: that each of these can only tell you about one particular line out of many. And strictly speaking it will only relate to people sharing that DNA today. On the other hand, a 37- or 67-marker Y-STR test are specific enough that they may clearly indicate that particular line is specifically related to a cluster of Ashkenazi Jewish lineages; or corresponds to a Y-STR signature strongly found in present-day Scandinavia. Does that mean you are a "Viking"? No. Does that mean that you had a direct male-line ancestor who was a Viking? If that's your Y-STR signature, then almost certainly. Is that fact, concerning just one line out of many, of interest to you? That depends on you. And of course, most likely you will get something much less revealing -- eg a very generic West European Haplogroup R1b ancestry; or alternatively, something less easily interpreted, for example a Haplogroup T ancestry like Thomas Jefferson, for which various possible routes have been suggested.

When we come to autosomal DNA, which is probably going to be the most revealing test for this Ref-desk question, there are a couple of good caveats that they make -- for example, that one can only fully securely relate what is found to the genetics of present-day populations, because there's a limit to how confident we can be about the geographical ancestral genetics; and also, purely due to randomness, the proportions of DNA correlating to particular geographical groups will only roughly correspond to the proportion of ancestry, because some ancestors won't be represented in the inherited DNA at all; and others will be more strongly represented.

What I think the pamphlet fails to connect with (disappointingly, given the high calibre of the advisors consulted) is the statistical power of moidern high-throughput genetics. It cites a paper (5) that considers what you can learn from ten to twenty loci. But the SNP chips used by 23 & Me are testing 500,000 loci. That makes it possible to do qualitatively different sorts of analyses.

The pamphlet highlights that "the Scottish Highlands do have some genetic differences from the bulk of the population, but they are not big". But with 500,000 probes, you have enough statistical power to start to pull out those differences. "There is no such thing as a ‘Scottish gene'"; but with 500,000 probes you can distinguish between really quite subtly-different probability distributions -- as the scatterplot from 23 & Me demonstrates. There is only "gradual genetic change and mixing", even across "strong cultural boundaries, such as between the Germanic and Romance language groups in Europe". But this is exactly in line with the 2008 data I quoted above, which found that "the researchers were able to assign, on average, 50% of European individuals to within 400 kilometers of their correct country of origin."

There's a relationship between how different the probability distributions are, and how many DNA loci you need from a DNA fragment to identify cluster membership and a changepoint from one cluster to another. If the populations are very distinct, you may be able to identify even quite closely spaced changepoints along the chromosome -- as the Chromosome 2 pics linked above appear to demonstrate, still being able to sharply identify the recombination points of DNA from different continents even at a depth of several generations on. If the populations are less distinct -- different European countries, say -- identifying the points along the chromosome where the DNA inheritance switches from one country to another may be much more difficult, for anything other than the very most recent ancestors.

So there certainly are some real limits to the resolution. But given that the question at the top referred to "races" -- which we might take as a not-particualarly-good shorthand for roughly continental-level population clustering, then, yes, as the Chromosome 2 pics demonstrate, it probably is possible to estimate pretty sharply which parts of which chromosomes came from which continents.

You would be relating bits of your DNA to *groups* of people, rather than getting a detailed personal history. And these are indeed groups that would have been characterised on a population-level, and are going to be pretty generic, so not going to point to any specific ancestor. (Though you might find a chunk of your DNA happens to match the corresponding chunk of somebody else's on the programme, establishing a common ancestor). However, I don't see how a breakdown of which part of your DNA comes from which group (or indeed which continent) could be said to be anything other than "unique to the individual".

Is the information going to be personally meaningful? Depends on what the individual finds meaningful, personally. Jheald (talk) 16:01, 30 August 2013 (UTC)[reply]

23andme offers "country of origin" analysis, and if it shows, based on genetic analysis that Person A is 99% European, while person B is 76% Sub-Saharan African, it is pretty clear which will have the Black physical characteristics and which will have the Caucasian characteristics. No one would be confused, looking at their pictures, as to which one was of mostly European and which one was of mostly African ancestry. "Race" is a social construct, such as Jim Crow laws or German Nuremberg Laws which made someone a Black or Jew based on "one drop of blood," or one great-great grandparent of the discriminated-against race. But there are readily detectable genetic indices of where ones ancestors lived. Edison (talk) 22:19, 30 August 2013 (UTC)[reply]

There are definitely people who claim they can reconstruct a face from DNA now [3] or in the near future [4]. But the former is an artist, and the latter is a company page about a future product (and their table of faces according to a gradation from male to female pegs my BS-ometer, because how would they get an intermediate value?). Our article on DNA profiling doesn't cover it. As of 2004 it was pretty crude [5] but one expects improvements. We'd really need to get a list of every product on the market and evaluate each one skeptically. Wnt (talk) 16:12, 31 August 2013 (UTC)[reply]

Isn't the answer here simply, "maybe, depending on what you mean by race"? If the subject has the alleles that produce the West African version of sickle-cell, or the Jewish version of Tay-Sachs, can't we draw the obvious conclusions? μηδείς (talk) 03:30, 1 September 2013 (UTC)[reply]

Yes. We can draw the conclusion that someone doesn't understand genetics. Alleles for the "West African version of sickle-cell" aren't confined to West Africans, and alleles for the "Jewish version of Tay-Sachs" aren't confined to Jews... AndyTheGrump (talk) 03:37, 1 September 2013 (UTC)[reply]

L-system and proportions

Hello. As far as I understand from the article, the commands for drawing a fractal using L system (if we draw it using turtle graphics) don't deal with the proportions of the fractal and the fact that in many fractals, deeper the recursion, smaller the segments added. Am I right? for example, the rules for Lévy C curve don't treat the fact that each segment is replaced by 2 segments that are smaller version of the initial segment (they are divided by sqrt(2)). Why is That?

Thanks! 94.159.221.4 (talk) 07:34, 30 August 2013 (UTC)[reply]

Your are correct - the iterated replacement rules in an L-system give a scale-free description of a fractal structure. How you then interpret the L-system strings to construct a family of geometric curves is up to you. You can rescale at each iteration, making each segment smaller and smaller so that overall size of the curve (as measured by, say, the width and height of a rectangle that contains it) remains the same. Or you can keep each segment the same size, in which case the overall size of the curve increases with each iteration. Both interpretations are valid - neither one is more "correct" than the other. Gandalf61 (talk) 08:56, 30 August 2013 (UTC)[reply]

Thank you very much! maybe it's worth mentioning in the article itself. 94.159.221.4 (talk) 11:19, 30 August 2013 (UTC)[reply]

Correlation between adultery and education level?

I found this interesting Psychology report on the web.

Whisman & Snyder (2007) also found support that the likelihood of infidelity decreases the more religious you are, as you age, or if you’re better educated. They also found that the risk for cheating was greater for women who were remarried (compared to those who were on their first marriage), or for either gender with the greater number of sexual partners you have.

I can easily find a possible reason for the correlation between religiosity and adultery or age and adultery. But the correlation between education level and adultery is what I find interesting. The article cites the source, and I try to find the source on the web, which brings me to the original article. Sadly, the original article doesn't really seem to support the notion of education level and adultery, besides reporting the negative correlation in the table. There is no verbal coverage on this topic. I am wondering if similar research has been done on the correlation between adultery and education level, or the possible reported reasons and explanations for why this phenomenon may be the case. 164.107.102.232 (talk) 13:54, 30 August 2013 (UTC) [reply]

DNFT--see banned user Wickwack at talk

The following discussion has been closed. Please do not modify it.

There is a vast multitude of survey based pschology "findings" out there. Most are pretty dubious, inclding those reported every month in Scientific American Mind, and in professional journals.

However, I don't find the claim that the risk of infidelity is less for the better educated at all surprising. The better educated are often (but certainly not always!) the more intelligent. That confers at least three factors: The more intelligent probably made a better and more thought out choice of partner to marry, rather than just go on lust or infatuation. They are better able to be led by their brain (seeing the negative consequences of infidelity) rather than being led by their hormones when meeting, as will inevitably happen sooner or later, an available person more sexy than their marriage partner. They are better able to reason out a commitment to do nothing that could potentially harm their children or affect their relationship with inlaws etc. A succesful infidelity-free marriage is based on three things: a) Commitment, b) Each has something the other needs, and c) a "circuit breaker" for the inevitable occaisonal silly argument. The more educated have demonstrated, whether they are intelligent or not, an ability to committ to something hard but durable - their education. The more intelligent probably have just as many "domestic tiffs" as the less bright, but are better able to get past them and get back to harmonious living. Finally, the less intelligent a person is, the more they view things as "all about me". The more intelligent tend to view things as "more about you".

124.178.41.65 (talk) 15:57, 30 August 2013 (UTC)[reply]

Interesting theory. Can you define what you mean by "intelligent" and "educated"? It appears that you use those two words differently. What type of "intelligence" are you talking about? I would have thought that there are so many confounds in regards to adultery that it would be nearly impossible to generalize accurately the relationships between adultery and its presumed causes. In addition, I think one explanation is to combine age, religion, and education into one big pot. Perhaps, as people grow older, they become wiser, more religious, and more educated - and thus less likely to cheat against each other. Many religions are deeply philosophical and may put people in what I call "deep thinking mode", a state in which people just stay calm, meditate, and think. And all that deep thinking and reasonings against unwise sexual intercourse make people less likely to commit adultery. It's a thought. 164.107.102.30 (talk) 16:56, 30 August 2013 (UTC)[reply]

Note that 124 is a banned sockpuppet so their responses are not welcome. Nil Einne (talk) 08:24, 31 August 2013 (UTC)[reply]

What is the reason for the orange bar? Why was my second post deleted? What is a sockpuppet? What is Wickwack at talk?

Of course there is an underlying assumption here that infidelity is necessarily a bad thing. Our society has basically said that it's a bad thing - but that doesn't mean that it is. From an evolutionary perspective, there is benefit to males in doing this because increasing the number of females he mates with increases the number of copies of his genes that make it into the next generation. It might be (and I have no evidence for this) that the impulse for males to be adulterous might not be negatively correlated with intelligence at all - completely the opposite in fact.

I'd also caution the results of studies that do not control for intelligent people being smart enough to get away with it more often - resulting in under-reporting of their infidelity rates.

We'd also have to be careful about our definitions of the term. Technically, it means having sex with someone else while you're still married - but what about when people are married, but separated? SteveBaker (talk) 17:47, 30 August 2013 (UTC)[reply]

Using the term "evolution" with "good" or "bad" is extremely misleading. First of all, the good/bad dichotomy may refer to morality, and by saying that "because evolutionary forces made us this way", you are automatically making a claim that somehow nature can make so-called "good" or "bad" choices. That is extremely misleading by itself. Second of all, you are assuming that the spreading of genes is desirable, when in reality it may not be in all situations. For instance, a counter-argument would be that males that invest time, effort, and commitment in the raising of the offspring ensure the viability and fertility of the next generation, while males that invest less time in such behaviors and more time in one-night-stands with various women may produce children living in destitute conditions, and such children may be left on the streets to die. Therefore, you cannot assign "good" or "bad" characteristics to nature. 164.107.102.219 (talk) 20:23, 30 August 2013 (UTC)[reply]

Ferdinand Richters (1849-1914) and Tardigrades - help!

Hello all,
Apparently Ferdinand Richters was the authority for any number of Tardigrade familia, genera and species: see Richtersius coronifer. I'm not being much use google-fu wise for the article.
Help me out here! Pete aka --Shirt58 (talk) 16:04, 30 August 2013 (UTC)[reply]

What do you need to know? I don't see a question here. SteveBaker (talk) 18:17, 30 August 2013 (UTC)[reply]

I think Shirt wants refs/sources confirming or counter-indicating Richters as the authority of the Tardigrade genera. But maybe I'm reading too much into it, or perhaps OP will clarify. SemanticMantis (talk) 18:24, 30 August 2013 (UTC)[reply]

Shirt created the Ferdinand Richters article, which currently contains one sentence. I think Shirt is looking for more information to put in the article. Looie496 (talk) 23:00, 30 August 2013 (UTC)[reply]

Here are 4 sources, Shirt: [6], [7], [8], [9]. They came from googling "Ferdinand Richters 1849 1914". -- Jack of Oz ^{[pleasantries]} 23:17, 30 August 2013 (UTC)[reply]

Thanks all. As SemanticMantis said, I'm looking for refs that verify that Richters was the binomial authority for Richtersius coronifer. And as Looie496 said, yep, I'm looking for more content for the article. De mortuis nihil nisi bonu notwithstanding, obituaries are good for verifying dates of birth and death. Looking for input from folks who have access to scholarly journals. Pete aka --Shirt58 (talk) 12:40, 31 August 2013 (UTC)[reply]

Shirt58, I found this paper [10], which led me to this [11]. Both list Richters as the naming authority for several tardigrade genera, even those which don't bear his name, e.g. Mesocrista spitzbergensis, (Richters, 1903). Other names are sourced to Richters (1908) and Richters (1911). Looks like these articles are not in English, and I don't read German [12] or French [13]. The common thing to do would be to cite Kathman and Cross (1989, my first link), as the source for the claim that Richters is a naming authority for several Tardigrade spp. and genera. SemanticMantis (talk) 15:02, 3 September 2013 (UTC)[reply]

Linear bench power supply for charging batteries

Mastech says their linear bench power supplies cannot be used to charge batteries or provide power for electroplating. They recommend specific models for those applications which have over-voltage and over-current protection. Why can't their linear power supplies also have over-voltage and over-current protection? --89.241.229.123 (talk) 18:39, 30 August 2013 (UTC)[reply]

Because both those applications tend to use very high current (or at least, there is a high risk of creating a short-circuit), and even the circuit-breakers and fuses built into bench power-supplies will have operational limitations beyond which they can not be reliably assured to provide protection. For example, see fuse breaking capacity and the corresponding section in our article on circuit-breakers. If an application is known to exceed the design capability of the protection circuitry, then the protection-circuitry is not safe and reliable. For those applications, a higher-grade protection-circuit is recommended. Nimur (talk) 20:31, 30 August 2013 (UTC)[reply]

It would not be a matter of "can't", it would be that Mastech did not put such protection in. This would be because it would have cost more, or degraded other specifications (such as size, accuracy or efficiency). For charging batteries you often need a specific charger that will not damage the batteries with overheat or overcharge, and yet not take too long. Graeme Bartlett (talk) 21:47, 30 August 2013 (UTC)[reply]

None of the above is correct, as reading the Masport information shows that overeload protection / current limitting is built in. The reason why these laboratory power supplies should notv eb used for battery charging is that they have front panel controls for voltage and current. It is all too easy through accident or ignorance to adjust the output so high the battery will be damaged and even to make it explode. The battery may discharge back into teh power supply when it is off. That flattens the battey and in some cases may damage the power supply.

Why was my earlier post deleted?

144.138.223.100 (talk) 15:31, 2 September 2013 (UTC)[reply]