Tolerance of resistor

Thread Starter

Missio2468

Joined Mar 18, 2022
69
Let's take tolerance of resistor be 1% and resistor be 1000 ohms. So the resistor value can be between 990 ohms and 1010 ohms. Now the resistance measured came out to be 995 ohms. Will this value keep on changing or will remain same provided all the condition of resistors like temperature, power, etc are under the rated specifications.
 

Ya’akov

Joined Jan 27, 2019
9,069
Welcome to AAC, good to have you with us.

All of the information you want is in the datasheet for the part. You can find a datasheet for the part number in question with a Google search. Here is an example from a 1% thin film SMD resistor:

1662370122494.png
Any decent manufacturer will provide similar data. All of the league suppliers have links to datasheets for the components they sell. It is a really good idea to get into the habit of reading the datasheets for parts you want to use because you can learn unexpected and possibly important tings before they present themselves as problems, as well as things you can do to improve the utility of the part.
 

WBahn

Joined Mar 31, 2012
29,976
You need to ask a more specific question. You asked if the value "would keep on changing", which implies that it has already been changing. What leads you to think that (if you do).

Are you thinking that the 1 kΩ resistor you measured at 995 Ω was initially at exactly 1000 Ω and that its value has somehow fallen to 995 Ω? That's not what is happening.

Imagine you are in the business of cutting logs to a length of eight feet and you advertise to your customers three grades of logs in terms of length tolerance - 10%, 5%, and 1%. You then go an cut lots of logs. They are not all going to be exactly eight feet long. If they are being cut by a bunch of hungover day workers with chainsaws, some of them may be way as much as a foot off. So after they are cut you have someone measure them. If they are within, say, 3/4" you stack them in a pile labeled 1%. If they are within, say, 4" you stack them on the 5% pile, and if they are within 9" you stack them on the 10% pile. All of the others get rejected. When a customer buys a 5% log, they may get one that is shorter than eight feet or one that is longer, but it won't be off by more than 5% -- and it won't keep changing it's length just because it was initially cut too long or too short by some amount.

But there may be a tendency for all of the logs to change length over time as they age and dry out. That is something very different and would be specified very differently. The lumber yard might treat them and season them in such a way that they know that they will not change by more than, say, 1% over 20 years. Other logs that have just been cut without regard to their moisture content or anything else might change as much as 5% over that time. For some applications you might need a log that is close to the advertised length and that won't change much over time, but other applications might not be sensitive to the initial length, but you still don't want it to change much. Yet other applications might need the logs to be close to the correct size initially, but you don't care how they change with time because they are only going to be in use for some temporary purpose for a month.

The same with resistors.
 

Ya’akov

Joined Jan 27, 2019
9,069
It still is hard to grasp if anyone could spread some more light on the question I raised
If you mean, “Does 1% tolerance mean that at any time the resistance will measure within 1% of the rated resistance?‘ then the answer is no. It means that the resistor was selected because it measured within 1%.
 

MisterBill2

Joined Jan 23, 2018
18,167
In addition to starting out as being within 1% of 1000 ohms, consider that it will never be exactly 1000 ohms because those fall into a more expensive class of accuracy. And the 5% tolerance parts will never be closer that 1% because that is how the sorting system works. And theten percent tolerance parts will never be within the 5% band. They are all made with 1000 ohms as the target, and then there is the sorting into the tolerance bands. At one time there were 20% tolerance resistors available, but as production control improved they have become much less common.
 

sparky 1

Joined Nov 3, 2018
756
Knowing when resistor tolerance is critical for an application usually takes a breadboard model to test (heat wave may call for a temperature-controlled room.) When the bread board's performance is lacking you can try replacing 1/4 watt with 1/2 watt, if that fails then a new approach is needed.

The arrangement and choice of resistors to compensate for the influence of temperature on resistance follows a different subject heading. "Temperature coefficient of resistance" The circuit board's copper layout and use of heat sinks. The use of parallel resistors along with combination of negative and positive temperature coefficients to arrive at a net TC of zero.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,167
In the real world, that is, in a situation where production yield is important and time for adjustments costs money, the normal practice is to create designs that do not demand exact component values, if at all possible. If some values must be exact, then the normal path is to add an adjustment capability.
There are exceptions, but in the world of most products the 1% tolerance is not part of a design.
 

Thread Starter

Missio2468

Joined Mar 18, 2022
69
You need to ask a more specific question. You asked if the value "would keep on changing", which implies that it has already been changing. What leads you to think that (if you do).

Are you thinking that the 1 kΩ resistor you measured at 995 Ω was initially at exactly 1000 Ω and that its value has somehow fallen to 995 Ω? That's not what is happening.

Imagine you are in the business of cutting logs to a length of eight feet and you advertise to your customers three grades of logs in terms of length tolerance - 10%, 5%, and 1%. You then go an cut lots of logs. They are not all going to be exactly eight feet long. If they are being cut by a bunch of hungover day workers with chainsaws, some of them may be way as much as a foot off. So after they are cut you have someone measure them. If they are within, say, 3/4" you stack them in a pile labeled 1%. If they are within, say, 4" you stack them on the 5% pile, and if they are within 9" you stack them on the 10% pile. All of the others get rejected. When a customer buys a 5% log, they may get one that is shorter than eight feet or one that is longer, but it won't be off by more than 5% -- and it won't keep changing it's length just because it was initially cut too long or too short by some amount.

But there may be a tendency for all of the logs to change length over time as they age and dry out. That is something very different and would be specified very differently. The lumber yard might treat them and season them in such a way that they know that they will not change by more than, say, 1% over 20 years. Other logs that have just been cut without regard to their moisture content or anything else might change as much as 5% over that time. For some applications you might need a log that is close to the advertised length and that won't change much over time, but other applications might not be sensitive to the initial length, but you still don't want it to change much. Yet other applications might need the logs to be close to the correct size initially, but you don't care how they change with time because they are only going to be in use for some temporary purpose for a month.

The same with resistors.
my value does not keep on changing from 1010 to 995 ohms. I am doing a project which requires accuracy so Have to use 1 percent resistor in it.
got the point from the example given. thank you
 
Last edited by a moderator:

Thread Starter

Missio2468

Joined Mar 18, 2022
69
If you mean, “Does 1% tolerance mean that at any time the resistance will measure within 1% of the rated resistance?‘ then the answer is no. It means that the resistor was selected because it measured within 1%.
Thank you,
No I didn't meant that. I meant if it was selected to be 1%. The specification of resitor like temperature ratings, power are well under the rated condition. Then will resistor keep on changing. I am not saying that after some years resistance might change but will it change in one instance and other at next instance??
 

MrChips

Joined Oct 2, 2009
30,707
No, the resistance does not change. It means that the resistance does not differ by more than 1% of the nominal value at a given temperature and power dissipation.

The 1kΩ 1% resistor measured to be 995Ω will remain at 995Ω during its life expectancy under normal operating conditions.
 

WBahn

Joined Mar 31, 2012
29,976
Thank you,
No I didn't meant that. I meant if it was selected to be 1%. The specification of resitor like temperature ratings, power are well under the rated condition. Then will resistor keep on changing. I am not saying that after some years resistance might change but will it change in one instance and other at next instance??
As long as you haven't done anything to stress it AND as long as it is staying close to the same temperature, it's value should be quite stable from run to run (using the same resistor).

If the resistor is going to see significant temperature variations, you'll want to take this into consideration. You can buy resistors with tempcos ranging from 0.2 ppm/°C (very expensive) to 350 ppm/°C (very cheap). Since 1% = 10,000 ppm, that latter choice could see a 1% change with a change in temperature of about 30°C -- and it doesn't matter if that is due to the environment changing or due to self-heating.
 

MrChips

Joined Oct 2, 2009
30,707
In addition to starting out as being within 1% of 1000 ohms, consider that it will never be exactly 1000 ohms because those fall into a more expensive class of accuracy. And the 5% tolerance parts will never be closer that 1% because that is how the sorting system works. And theten percent tolerance parts will never be within the 5% band. They are all made with 1000 ohms as the target, and then there is the sorting into the tolerance bands. At one time there were 20% tolerance resistors available, but as production control improved they have become much less common.
This is not true. If they are sorting and removing the 1% resistors then the 5% lot would have a bifurcation in distribution.

The tolerance of the end product is more a reflection of the quality of process control.
 

WBahn

Joined Mar 31, 2012
29,976
In addition to starting out as being within 1% of 1000 ohms, consider that it will never be exactly 1000 ohms because those fall into a more expensive class of accuracy. And the 5% tolerance parts will never be closer that 1% because that is how the sorting system works. And theten percent tolerance parts will never be within the 5% band. They are all made with 1000 ohms as the target, and then there is the sorting into the tolerance bands. At one time there were 20% tolerance resistors available, but as production control improved they have become much less common.
If this is the case, then nearly everything would be marked as some kind of 1% resistor. After all, if it came in at 1025 Ω why mark that as a 1000 Ω/5% resistor and not a 1020 Ω/1% resistor.

There was a time when some manufacturers made slow, rough changes to the mixture and churned out resistors that were effectively randomly distributed about a rough target. These were then measured and marked based on production targets for each kind of resistor they made. So a resistor that measured 1050 Ω might get marked as a 1050 Ω/1% resistor today and a 1000 Ω/20% resistor next month and a 1100 Ω/5% resistor six months from now. It all depends on what their volumes are of each kind and what was coming off the line.

Today, process control is so much better that, according to a conversation I had with a Vishay engineer about a decade ago, they program the machine to make a specific number of a specific tolerance resistor and it does so. It can make looser tolerance resistors faster because it doesn't have to maintain as tight a control, but when making the tighter tolerance parts, the ones that are out of spec are sufficiently few that it isn't worth trying to salvage them as a different part number -- they are simply rejected.
 
Top