Why do electronic come isodd values like 0.47 instead of 0.5?

[Beamin]

This always seemed strange to me. It seems values like 0.47 go as far back as the 40's. Where there some applications like to work off 60Hz you needed some really specific value to regulate voltage or match the internal IF of AM radio's? Kind of like all barns are red because red is the cheapest colored paint and it's the cheapest colored paint because they sell the most? Like one person ordered an extra amount of 0.47 valued parts and it lowered the cost just enough to make everyone buy it then make it cheaper to make ad Infinium? Why can't I get .50uf?

[newbrain]

A very practical reason:
E-series of preferred numbers

[IanMacdonald]

For the same reason that the Imperial tool sizes made more sense than the metric.

1, 1/2, 1/4, 1/8 - Same relative change each time.

5mm, 4mm, 3mm, 2mm, 1mm -  Increasing change each time.

[German_EE]

It's my understanding that this particular sequence of numbers are used so that you can use the smallest number of components to make any value.

Now someone tell me why 455 KHz and 10.7 MHz are used in IF strips. The case for a 9 MHz IF strip is easy when you work out 9 MHz +/- a 5-5.5MHz VFO and look at the amateur radio bandplan.

[apis]

For the same reason that the Imperial tool sizes made more sense than the metric.

1, 1/2, 1/4, 1/8 - Same relative change each time.

5mm, 4mm, 3mm, 2mm, 1mm -  Increasing change each time.

That has nothing to do with metric vs. imperial. There is absolutely no reason why you couldn't have 8, 4, 2, 1, 1/2, 1/4, 1/8 mm sized tools. Imperial is even defined in terms of the SI system today, the difference is just a conversion factor. The reason everyone should use metric is because there are synergistic effects when everyone uses the same standardised unit of measurement.

[Fgrir]

For the same reason that the Imperial tool sizes made more sense than the metric.

1, 1/2, 1/4, 1/8 - Same relative change each time.

5mm, 4mm, 3mm, 2mm, 1mm -  Increasing change each time.

1, 7/8, 3/4, 5/8, 1/2, 3/8, 1/4, 1/8 - What's your point?

[Zero999]

Let's not start another silly metric vs imperial debate. 

I often use similar values when conducting experiments. Suppose I want to measure the bandwidth of an amplifier. Rather than going up in 10Hz steps which can be tedious, if I want to cover the entire audio band, plus a bit more, say 10Hz to 50kHz, I use similar values to the E6 series, albeit rounded, to make it easier: 10Hz, 15Hz, 20Hz, 30Hz, 50Hz, 70Hz, repeatedly multiplying by ten each time, until I get to 50kHz. That way I only need to set the signal generator to 23 different frequencies, rather than hundreds.

Another set of prefered numbers is the Renard series, which is commonly used for fuses.
https://en.wikipedia.org/wiki/Renard_series

[T3sl4co1l]

It's my understanding that this particular sequence of numbers are used so that you can use the smallest number of components to make any value.

Forming an arithmetic value from a geometric series, that's more or less true; take the binary number system for example, where the step size is a factor of 2.  The efficiency is near optimal (2 is close to e; base e is the optimum for number representation).

I guess that breaks down if you need a number with the least digits, though.  In that case, infi-nary (if you will) is the least possible, i.e., you use one resistor that is exactly the value needed, in each place.

So, we need some other weight function -- some new information which says, for example, how expensive it is to stock so-and-so many resistors, relative to their unit costs.  Math alone cannot provide an answer here, other than to say it's probably on the low side (so, less than 2x between values).

As for voltage dividers (the other major application), geometric values make for poor diversity in near-geometric functions like the voltage divider formula (actually, I suppose the formula /is/ still geometric, but in a more basic sense, not in the sense of geometric series...nevermind).  Occasionally you get lucky, but E-series values tend to leave unfortunate holes in the available ratios, whereas arithmetic values give more coverage but with less diversity in the total resistance value the divider has.

Example, Say you have an arithmetic series: 1, 2, 3, ..., 8, 9, 10, 20, 30, ...  You only get a 1/7th ratio with 1k and 7k resistors, or 10k and 70k, but not 5k and 30 or 40k (35k being ideal then).  Whereas in E-series, you get, say, 1k and 6.8k, or 2.2k and 15k, or...  The total value has more diversity in the latter case, while the ratios have more diversity in the former case.

Tim

[KE5FX]

It's my understanding that this particular sequence of numbers are used so that you can use the smallest number of components to make any value.

Now someone tell me why 455 KHz and 10.7 MHz are used in IF strips. The case for a 9 MHz IF strip is easy when you work out 9 MHz +/- a 5-5.5MHz VFO and look at the amateur radio bandplan.

Less chance of beatnote interference from LO radiation.  An odd IF like 455 kHz or 10.7 MHz puts the LO at a frequency where it won't land near the center of another channel.

I'm not sure why they didn't pick (e.g.) 452.5 kHz, which would have put the image response between channels as well.   Possibly because a 2.5 kHz beat note would be more objectionable than a 5 kHz one.

[hamster_nz]

It is so you have the oplimal spacing of values.

For example, low precission resistors come in five values per power of 10. Each about 1.47x the value of the previous value. So for any desired value you can get a resistor that is about +/-25% of the 'perfect' value.

[ocw]

The affect of a numeric change is logarithmic instead of linear. 
Since there are six 20% tolerance readings they are spaced about the sixth root of 10 apart (1.47).  1.47 x 1.47 = 2.16 x 1.47 = 3.18 x 1.47 = 4.67 x 1.47 = 6.86 x 1.47 = 10.1 (because 1.47 was rounded off).
Since there are 12 10% tolerance readings they are spaced about the twelfth root of 10 apart (1.2115).
Since there are 24 5% tolerance readings they are spaced about the twenty fourth root of 10 apart (1.10).
Since there are 96 1% tolerance readings they are spaced about the ninety sixth root of 10 apart (1.024).  Keep multiplying that number by itself 96 times and you will end up at 10 covering all of the 1% resistors in that range.

[Brumby]

The effect of a numeric change is most often logarithmic instead of linear. 
Since there are six 20% tolerance readings they are spaced about the sixth root of 10 apart (1.47).  1.47 x 1.47 = 2.16 x 1.47 = 3.18 x 1.47 = 4.67 x 1.47 = 6.86 x 1.47 = 10.1 (because 1.47 was rounded off).
Since there are 12 10% tolerance readings they are spaced about the twelfth root of 10 apart (1.2115).
Since there are 24 5% tolerance readings they are spaced about the twenty fourth root of 10 apart (1.10).
Since there are 96 1% tolerance readings they are spaced about the ninety sixth root of 10 apart (1.024).  Keep multiplying that number by itself 96 times and you will end up at 10 covering all of the 1% resistors in that range.

My vote ^ ^ ^ ^

[TerraHertz]

Beamin, I can recognize a topic you created, before even looking at the user name.
Not that I'm complaining. I just wonder how you do it.

[T3sl4co1l]

Beamin, I can recognize a topic you created, before even looking at the user name.
Not that I'm complaining. I just wonder how you do it.

Posing a question that Google would provide adequate information to?

Tim

[Zero999]

It's my understanding that this particular sequence of numbers are used so that you can use the smallest number of components to make any value.

Forming an arithmetic value from a geometric series, that's more or less true; take the binary number system for example, where the step size is a factor of 2.  The efficiency is near optimal (2 is close to e; base e is the optimum for number representation).

I guess that breaks down if you need a number with the least digits, though.  In that case, infi-nary (if you will) is the least possible, i.e., you use one resistor that is exactly the value needed, in each place.

So, we need some other weight function -- some new information which says, for example, how expensive it is to stock so-and-so many resistors, relative to their unit costs.  Math alone cannot provide an answer here, other than to say it's probably on the low side (so, less than 2x between values).

As for voltage dividers (the other major application), geometric values make for poor diversity in near-geometric functions like the voltage divider formula (actually, I suppose the formula /is/ still geometric, but in a more basic sense, not in the sense of geometric series...nevermind).  Occasionally you get lucky, but E-series values tend to leave unfortunate holes in the available ratios, whereas arithmetic values give more coverage but with less diversity in the total resistance value the divider has.

Example, Say you have an arithmetic series: 1, 2, 3, ..., 8, 9, 10, 20, 30, ...  You only get a 1/7th ratio with 1k and 7k resistors, or 10k and 70k, but not 5k and 30 or 40k (35k being ideal then).  Whereas in E-series, you get, say, 1k and 6.8k, or 2.2k and 15k, or...  The total value has more diversity in the latter case, while the ratios have more diversity in the former case.

Tim

The E24 series gives plenty of nice ratios. Just use a calculator program to find them, such as on the site linked below. 1:7 would give 13k and 91k for the E24 series, which is dead on or 4k7 and 33k for the E12 series, which is only 0.3% off.
http://jansson.us/resistors.html

[Beamin]

Beamin, I can recognize a topic you created, before even looking at the user name.
Not that I'm complaining. I just wonder how you do it.

Posing a question that Google would provide adequate information to?

Tim

But I wouldn't have learned about IF frequencies or that whole numbering system you provided and lots of other things. Hope fully when people google this now they will go to this forum and see and learn all that I have instead of :
 0.47 is IEEE standard 45.98b.

[CopperCone]

i heard it had something to do with rope sizes and pullies in the old british navy

[Beamin]

Beamin, I can recognize a topic you created, before even looking at the user name.
Not that I'm complaining. I just wonder how you do it.

I ask the questions that everyone else just takes for granted. The true beauty of world lies in the little subtle details of why things work. Why is the hardest question to answer. My mind is sponge that is never saturated. I think most people have no idea why a sponge holds water other then :because it's wet. According to Veritasium most people think the sun goes around the earth for me I can't not* not know that. When I look at the sun I see a perfect natural fusion reactor full of H He N O etc... with stuff spinning around it, verses a bright light.

For me to understand something I have to know why. I was top of my class in elementary school outpacing all the other student except for spelling. It was because of things like: I before E except after C ( as a teenager it was LSD before E then K). OK why? Because it is. But to say: ice floats because it's less dense. Why? because it forms a crystal that is bigger then the liquid. Now I will remember that ice floats.

[ebastler]

Posing a question that Google would provide adequate information to?
Tim

But I wouldn't have learned about IF frequencies or that whole numbering system you provided and lots of other things. Hope fully when people google this now they will go to this forum and see and learn all that I have instead of :
 0.47 is IEEE standard 45.98b.

I don't know how you use Google, but if I search e.g. for "resistor values reason", the second result is this:
https://electronics.stackexchange.com/questions/123453/why-resistor-values-are-non-linear-what-makes-sense-in-470-ohm

Tim has a point there; a tiny bit of effort wouldn't hurt before you post these questions.

[T3sl4co1l]

Maybe I grew up at a time when Google was a good resource for questions and not burdened by its individually-tailored (and often quirky/faulty) results algorithms, but it is worth noting that, for example, Stack Exchange has a much higher PageRank than EEVBlog does.  And it's likely (though not impossible) that, if the question hasn't been asked there (or Quora, a rising resource today as well), it has never been asked, and you may perhaps consider rephrasing the question closer to one that has been asked, and which does have a reasonable answer.

In the process of searching, you still learn a little about the subject.  A given system only works in such-and-such a way, and if you ask a question that goes sideways across the system rather than following its actual structure, will generate confused looks from both practitioners of the subject, and search engines over it.

Also, knowing where to find knowledge, is itself an important skill.  I'll admit this can be much harder to figure out.  For example, I suspect (but do not know for sure -- it's been a while since I read through it) you can find a brief explanation of popular BCB radio IF frequencies in RDH4 (Radiotron Designer's Handbook, 4th ed -- it's out of copyright, freely available, and still fairly relevant today despite its dedication to vacuum tubes), which otherwise may be impossible to find via Google (I see a lot of general pages and articles, mostly about microwave systems, none about now-archaic BCB radios -- you don't get search results tailored to every aspect of your query, including possible chronological cues, rather, you get results from the current population of web pages, and those are dominated by microwave and millimeter wave articles, among other things.)

Tim

[Zero999]

Maybe I grew up at a time when Google was a good resource for questions and not burdened by its individually-tailored (and often quirky/faulty) results algorithms, but it is worth noting that, for example, Stack Exchange has a much higher PageRank than EEVBlog does.  And it's likely (though not impossible) that, if the question hasn't been asked there (or Quora, a rising resource today as well), it has never been asked, and you may perhaps consider rephrasing the question closer to one that has been asked, and which does have a reasonable answer.

In the process of searching, you still learn a little about the subject.  A given system only works in such-and-such a way, and if you ask a question that goes sideways across the system rather than following its actual structure, will generate confused looks from both practitioners of the subject, and search engines over it.

Even with the typographical errors/spelling mistakes, a Google search for "Why do electronic come isodd values like 0.47 instead of 0.5?" returns a correct answer on Stack Exchange, on the fourth hit. Of course this thread is top of the list, but if she'd just Googled it in the first place, she most likely would have got Stack Exchange.
https://www.google.com/search?q=Why+do+electronic+come+isodd+values+like+0.47+instead+of+0.5%3F&ie=utf-8&oe=utf-8&client=firefox-b

Also, knowing where to find knowledge, is itself an important skill.  I'll admit this can be much harder to figure out.  For example, I suspect (but do not know for sure -- it's been a while since I read through it) you can find a brief explanation of popular BCB radio IF frequencies in RDH4 (Radiotron Designer's Handbook, 4th ed -- it's out of copyright, freely available, and still fairly relevant today despite its dedication to vacuum tubes), which otherwise may be impossible to find via Google (I see a lot of general pages and articles, mostly about microwave systems, none about now-archaic BCB radios -- you don't get search results tailored to every aspect of your query, including possible chronological cues, rather, you get results from the current population of web pages, and those are dominated by microwave and millimeter wave articles, among other things.)

Tim

It might still be in copyright, depending on where it's published. A quick Google reveals it's by Fritzs Langford-Smith, who died in 1966 and the default copyright term in many countries, is between 50 and 70 years, after the death of the author, which would mean it expires between 2016 and 2026. Be careful when sending out PDF copies: whilst copyright might have expired in one region, it may still be copyrighted in another.
https://en.wikipedia.org/wiki/Fritz_Langford-Smith
https://en.wikipedia.org/wiki/Copyright_term

[Off Topic]Whilst I wholeheartedly support copyright, I think terms need to be drastically reduced to something more sensible such as 20 years, after the work's creation. If someone or an organisation can't produce another product after 20 years, they don't deserve any more money![/Off Topic]

[Circlotron]

When I look at the sun I see a perfect natural fusion reactor full of H He N O etc...

I think N and O only if/when a star goes supernova.

[BBBbbb]

...Quora, a rising resource today as well...

ugh, not sure I agree with you on this. They are attracting visibility in search results but offer poor quality.  IMO 90% of Q&As there are either an exact copy from other platforms (stackexchange, reddit, ...) or just marketing for some products or services masked as answers. 

[Beamin]

...Quora, a rising resource today as well...

ugh, not sure I agree with you on this. They are attracting visibility in search results but offer poor quality.  IMO 90% of Q&As there are either an exact copy from other platforms (stackexchange, reddit, ...) or just marketing for some products or services masked as answers.

[Beamin]

...Quora, a rising resource today as well...

ugh, not sure I agree with you on this. They are attracting visibility in search results but offer poor quality.  IMO 90% of Q&As there are either an exact copy from other platforms (stackexchange, reddit, ...) or just marketing for some products or services masked as answers.

    Yes Quora is horrible always wanting your email and other BS only to get the wrong answer. But I agree with the other posts I will spend more time with google and only post questions that I absolutely can't find answers to or when the answers require more thought or when sources are contrary. Tanks for all the answers I'm going to look up IF now, after I finish the wiki page on gamma rays. I can look up one thing on wiki then spend hours just chasing links trying to learn every topic in the page!