That’s nice and all, but a uA7805 is not the same as a L7805 or an LM7805. You certainly can do things like this and especially in ha “hobby” environment, but if you want to scale this up you would certainly have to be very cautious about which particular 7805 variant you use. Also, for bigger production runs, adjusting all those pots is going to be a nuisance.
I won’t be surprised if there were over 50 variants of the “7805” regulator.
Hi, Piotr
If I understand you correctly, it seems that all of your LM317s were outside the specified voltage tolerance? I wonder if you were getting reject or counterfeit ICs in Poland. Where linear regulators are concerned, I wonder whether the 7805s have been largely replaced by low drop out linear regulators (LDOs). The power savings from low dropout is useful. When I think of other good old ICs, a couple of other National Semiconductor devices come to mind. Those are LM339 quad comparator and LM324 quad op amp.
No!
I don’t know how you could understand me that way. Really - where from you understand that all were out of range?
Most were in tolerance. I don’t remember how many of them were in ±2% and how many up to ±4%. May be enough to throw away those out of 2% but as I had to measure each of them so writing (with pencil at its big pad) the exact value (only 2 digits after the point) was not a problem.
Using all of them was compatible with that that we had 5% resistors. So we had to measure them also.
As we had both measured then better solution was to pair them all then to throw away LMs out of 2% and resistors out of 1%.
I have measured may be 500 of them. I remember one with 18 written on it and one very far of range (may be 12 on it). So there were 2. And I suppose (but not sure) one had a number greater then 30. So they were 2 or 3 out of range. So I have written that I found not one with voltage out of range.
Now I use one tool made of LM317. It is current load switched by jumper for 2, 5, 10, 20, 50, 100, 200, 500, 1000 mA. It has no radiator so when using (rarely) I grab the LM317 in a vise.
Not if you want to get 5V out of 12V.
For me advantage of some LDOs is their low quiscent current. In our RFID readers I use them to get down from 10…28V to 3V3.
Ah Hah!!!
An English Second Language grammatical Problem leading to a misunderstanding.
Adverb “even” in wrong place.
Sentence should read: I happened to find not even one LM317, with that voltage, out of specified range.
I don’t care about the grammar mistake, @BobZ won’t care, and I hope you, @Piotr don’t care either.
Your English is infinitely better than my Polish in spite of my having a Polish father. Polish was never spoken in my house so I never learned the language.
That reminded me that the word only is very sensitive to placement in some English sentences. Probably in other languages too. Compare these:
I think your version is not what I wanted to say (if I understand it correct).
Your sentence I understand as all were in range as I found not even one out of range.
In my original sentence I wanted to say that they had voltages distributed in the whole range 1.20…1.30 what is not so good as it is 4% (2 times more than 7805 2%). And it happened that more then one of them had a voltage even out of that range.
My even was about the voltage happened to being out of range.
The voltage was distributed in the whole defined range (it is ok) but it happened to be out of range (not ok) so it was ‘even out of range’.
To tell you the truth I still don’t understand why you take the word ‘even’ from ‘voltage even out of … range’ and understand it with connection with the ‘not one’ standing in the totally different place in sentence.
I was sure it is fully clear what I am writing.
Piotr: Much of that post was unclear. One reason why I took you to mean that the LM317s were out of tolerance was that I did not understand otherwise why you decided to to replace them with 7805s.
I am sympathetic to the ESL issue, but you need to appreciate that the rest of us are doing the best we can to understand. Some of us may do better than others.
I hope now it is clear why I replaced LM317 with 7805.
It was my first design that was manufactured in serie. I just didn’t counted the tolerances. 4% of LM317 tolerance + 5% of standard resistors tolerances given us 9% if not measuring and matching. When we noticed it we start to do matching, but it was time consuming.
7805 with its 2% is simply more repetitive. At least 5V was enough accurate.
Have I said anything that you can understand that I not appreciate?
I am just trying to explain where from come my problems with understand the problem source.
I was 100% sure that my even is clearly about the voltage that happened to be not only at tolerance border but even beyond it.
I have read my original sentence few times and I think I got it.
The problem is not in the word ‘even’ as I was thinking all the time.
The problem is:
I was writing: It happened to find (not one) LM317…
What you read: It happened to (find not) one LM317 …
In Polish what you read I would have to write: It happened to not find one LM317… and my original text can’t be understand as being this one.
My first contact with EMC standards (in 2002) was that after reading one standard I have written to Polish Standards Committee that there have to be an error in it as I can’t understand it. I got the answer the same day saying that it is not possible as it was translated and checked by many people knowledgeable in the field, but with answer they send me the original version of the page I questioned.
There were one multiple complex sentence making up the entire 4 or 5 line paragraph. Standards are written at A4 not in columns but from left to right so it was really long, complicated sentence.
The construction that was wrongly translated was: "You can conclude that… provided that … unless not … ". And imagine that each … were also complex statements with different conditions, conjunctions, etc. (something must have produced those 4 or 5 lines).
The translation completely reversed the meaning of the statement.
Having that I have written them where the bug is. This time I got the answer after 2 weeks. The man wrote that he was on holiday and therefore such a delay, but it wasn’t the holiday season at all.
Some time later they updated that standard.
As of your last post, now I think I do. It seems to be the combined error of the 4% LM317s and the 5% resistors. This leads me to wonder what type of resistors you were using.
During the 1970s my employer was commonly using carbon composition resistors. Whether you have 1/2 watt or 1/4 watt, the body of a carbon composition resistor usually appears to be perfectly cylindrical. During those days I began to see carbon film resistors. On carbon film resistors, the diameter of the ends of the resistor body is usually larger than the center of the body.
Electrically, carbon film resistors seem to be much more accurate than carbon composition resistors. In my experience, a 5% carbon film resistor value seems to be within 1% of the marked value most of the time. But carbon composition resistors often seem to be out of tolerance (beyond the specified tolerance band.) I seem to remember measuring some “5%” carbon composition resistors and finding the actual value to be off by as much as 12%. They may be sensitive to moisture or humidity. Carbon composition resistors are supposed to be better able to tolerate surges or high peak energy.
I pondered over this comment for some time. It was very difficult to understand. I finally concluded that, to make sense, the word “even” must be in the wrong place. I assumed the word “even” was correct as most ESL writers use the correct word but sometimes use the wrong sentence structure.
eg: “It happened to find” should read " I happened to find".
In this case, after clarification of the confusion, it seems the wrong word was used or one word was omitted.
How do these read @Piotr?
In English, any of the below sentences are now easy to understand.
I happened to find not one LM317 with that voltage, some even out of specified range.
or: I happened to find not one LM317 with that voltage, some being out of specified range.
or: I happened to find not one LM317 with that voltage, some out of specified range.
My apologies to you @Piotr. In future, if I do not understand a comment, I will ask for clarification instead of attempting to interpret myself.
I had some old ones in my junk^Wspares box and often when I measured them they were well off the value indicated by the colour bands.
I remember a possibly apocryphal joke about a computer translating the English idiom: Out of sight, out of mind (meaning what you don’t see you ignore), to Invisible idiot, in another language.
We were using carbon film resistors. I can’t be sure of it but once in past after measuring a serie of 5% resistors my conclusion was that they manufacture them and after it they select from them those 1% and the rest is sold as 5%. So among 5% there were no resistors in 1% range. All were in range -5%…-1% or 1%…5%.
I don’t see the important difference between ‘It happened’ and ‘I happened’. For me if I describe some my experience I have done long time ago both these tells for me that something just happened.
In all these sentences their first part seems for me odd - what does it mean ‘to find LM317 with voltage’. I would modify it to:
I happened to find some LM317 with that voltage even out of specified range.
Trying to correctly write my original sentence with as little modification as possible I suppose the correct would be:
It happened to find not only one LM317 with that voltage even out of specified range.
Would you understand it as most were in range but few of them were out of it?
I think the main source of me writing it not correctly is that in Polish we have a word made by two words.
not = nie.
only = tylko.
one = jeden.
not only one = nie tylko jeden = niejeden.
And I just wrongly translated ‘niejeden’ to ‘not one’ instead of ‘not only one’.
My conclusion is: it is better to not use negations.
If I have written that I even found some LM317s with voltage out of range no misunderstanding would arise, I think.
Negations are tricky to translate. In some languages a double negative cancels, in some it intensifies the negation.
Maybe the language issues are better discussed via PM?
Does this still happen?
As far as I know the production process has been under tight enough control for 30 years or so that this does not happen anymore.
This is something which I have heard or imagined may happen, but I have never encountered it. I think that the better manufacturers may have done better with statistical process control. These days in USA, 1% is (in practical terms) the default tolerance for “thick film” chip resistors. If the manufacturers had to screen out many that were outside of 1%, I suspect that they would not be able to sell them all as 5%.
I would be not surprised if all resistors were manufactured now with 1% tolerance and then sold partially as 1% and partially as 5%. Only to satisfy someone who specified 5% in the BOM.
When I moved to KiCad (2018…2020) I changed my general resistor specification from 5% to 1%.
Thanks to that I have in library only one 1k 0603 resistor and not two like previously (one 5% and one 1%). Contract manufacturer sometimes asks me if he can use 5% instead of 1% for specific resistor as he has them still in stock. I think in some time he will have only 1% resistors.
No, I think.
It was in 1989 and I’m not sure if remember well. And have in mind that a socialistically managed economy produces much worse results. You can see it in Ukraine now.
Even 0.1% are affordable from the big parts distributors these days. 40 years ago these would have been a months pay each laboratory standards
Third situation - you have to use double negation and it simply means negation (not cancels, not intensifies).
For example saying: ‘I know nothing about it.’ sounds stupid in Polish. You can’t say that you know if you really don’t know. You have to say: ‘I don’t know nothing about it.’ We don’t have word for anything so nothing and anything are both replaced by our nothing (=nic).
But how to do it if three or four people take part?