ESR Values for Electrolytic Caps

[dark_hawk]

Hi,

I'm looking to establish a table for acceptable ESR values for Caps, I know that for each specific cap you can look at the data sheet and get the ESR and the tan d (D) to know if your cap is with spec, but when diagnosing caps in a malfunctioning circuit with a lot of caps you will need a ballpark numbers for ESR because desoldering all those caps would be a problem.

Frequent Contributor "Wytnucls" provided a table for the ballpark dissipation factor (D) for various Manufacturers and series.


Problem with ESR values is that it differs from different sources.

Here are some examples:





Taking the 10uF 25V for example, the vaules are 5.3, 14, 1.5.
Which of these tables' values seem closer to the "correct" vaules?

Thanks.

[free_electron]

that's like making a table giving the weight of a cow .. depends on the subspecies, race ,where it grew up , what it ate , when it ate , if it farted yet ...

[robrenz]

You forgot lactating or not 

[free_electron]

if capacitors lactate they are bad. 

[PA4TIM]

I only look at D and capacitance and measure (except for measuring fun and experiments)  NEVER in circuit (at 1 kHz) I'm now repairing  a Tek 2710 spectrum analyser and have replaced all caps. Just for fun I tested all caps in situ for C, D at 1 kHz and ESR at 100kHz. Then I removed the cap, measured it again using GR and IET LCR gear.
95% of the in situ measurements were so way of that I now am even more sure that methode sucks. It is nice if you have no clue about what you are dioing and a chinp can repair some consumer stuff this way.

Most caps turned out to be much better as in situ. But one bizar example: All tantaliums tested good in situ except one. That one turned out to be the only good one !  five 1 uF tantaliums measured a lot better ESR in circuit because there are a lot of caps parallel. One showed excelent ESR but the scope showed a problem.. After desoldering it turned out to have a D of 1.5 and 4 others were leaking current (but the scope shows it all)

[mariush]

All those charts are based on old series of capacitors ... some that were called Low ESR back then are almost classified in the general purpose category nowadays, there's new series or capacitors with much lower esr values than what was considered back then low and very low esr.

They're not meant to be accurate or definitive and you can't generalize something...

For example you may have a 330uF 16v  nichicon hn capacitor with 0.021 ohm impedance and you may have a 130c rubycon rx30 330uF 16v capacitor with 0.22 ohm impedance.
If you ask, you'll definitely find someone to tell you that both are low esr, but hardly interchangeable.   

I know it's technically not 100% correct but in practice it's close enough ... I usually just look in the datasheet for that particular series of capacitors at the Impedance column, measured at 100Khz and compare the ESR i get with a 30-50$ esr meter (a meter using the same principles as bob parker's esr meter).

If the impedance for a 3300uF 16v is listed as 0.015 ohm and I read 0.3 ohm, it's kind of obvious it's a 'sick' capacitor.

[kanurys]

Also, If there are multiple of the same type of capacitor, comparing their values is a useful tool. I guess it doesn't really tell you which ones are good or bad, though.

[danadak]

Compounding this is T as well as process changes over time.

Not a trivial problem you are trying to master. You cannot rely on datasheets
because over time manufacturer may rev the part with no notice (although
components EE in company can raise hell over this and disqual a vendor).

Best you can do at design time is to qual parts and their datasheet revision
at any give point in time, and pray.


Regards, Dana.,

[tigr]

For all the ESR tables, a very good capacitor.

[innkeeper]

now wait, i think the idea has merit...
Others have pointed out the pitfalls, i wont repeat them

how could this be done...
maybe you can generalize cap types and series and then give like an average esr for them.
this will also let you cover various cap types in multiple tables.
sure its many tables, but, it would let someone look it up quick and see if there in the ballpark.
in example, a table that broadly covers all 'typical' low esr electrolytics. (i know i am gonna get flamed just for saying there a typical low ESR electrolytic)

it would be a general table, if someone needs to be sure or exact, then they can go to the data sheet for the cap.

so sure, there no substitute for looking up the exact cap and knowing what the esr should be.
but as a quick reference, for those who are more the weekend engineer types could be a handy reference.
Plus, i don't think there is a modern version of this kind of table exists in one place..

[ruairi]

For all the ESR tables, a very good capacitor.

Can we not do the endless pics thing on this thread?  The formatting alone makes me queasy.

[innkeeper]

guess he had a bad reactance to pictures 

[tigr]

For all the ESR tables, a very good capacitor.

Rubycon MBZ 3300uF/6,3v max.ESR-0,012Ohm

[tigr]

it would be a general table,

Maybe this one?

[tigr]

Another ESR table. After a small adjustment, it can be used for "high-voltage" capacitors.
Only the ESR table needs to be expanded, starting at 0.1uF.
For example.New capacitor 0,47uF/400v.ESR-30Ohm.

[tigr]

Seriously, the deterioration in the characteristics of electrolytic capacitors and the processes occurring in them, are still poorly understood.
With the in-circuit measurement, looking at the ESR, it can be mistaken for a good capacitor.
Bad capacitor.
680uF/50v.

[perieanuo]

Compounding this is T as well as process changes over time.

Not a trivial problem you are trying to master. You cannot rely on datasheets
because over time manufacturer may rev the part with no notice (although
components EE in company can raise hell over this and disqual a vendor).

Best you can do at design time is to qual parts and their datasheet revision
at any give point in time, and pray.


Regards, Dana.,

Marked as best solution
Regards,Pierre


Envoyé de mon iPad en utilisant Tapatalk

[tigr]

In-Circuit ESR meter.

[tigr]

Fill in the table of ESR.
1uF/16v.
ESR-2Ohm.

[PA4TIM]

Theoretically, ignoring dielectric absorption loss, ESR=D*X, there X=1/(1*pi*f*C).

X=1/(2*pi*f*C)
And D is given most times for 120Hz, impedance is given for 100 kHz. ESR is most times not stated. 100kHz kinda works for ESR because Xc is almost zero there. But skineffect and dielectric loss can disturb ESR measurement at 100 kHz. ESR is not flat over frequency.

Use s scope, if the ripple is to high, the cap is probably not good.  For the rest, desolder, look up the data sheet. The biggest no-no is measuring caps in situ. I have done many tests on that and it is useless. A measurement that gives a high ESR in situ will be usable. But a complete dead cap can show up bad when measured in situ. And that makes in situ measurements useless

[oldway]

I find that this kind of topic in the beginner section is not very profitable.

This encourages beginners to develop a real phobia of electrolytic capacitors .... 
The attitude towards electrolytic capacitors must be moderate and mitigated. 

Systematically replacing ALL electrolytic capacitors is, in many cases, utterly stupid. 

We often bring more problems than we solve.

We must remember a few rules:
- Nobody cared about the ESR capacitors before the advent of switching power supplies, then, come to change electrolytic capacitors in a 70s amp on the pretext that the ESR is too high, it's absurd.
- Generally, a too high ESR is manifested in one way or another: ripple too high, oscillations, instabilities, ....
To understand if the ESR has or not an importance in the circuit, imagine the circuit with an additional resistance in series with the capacitor .... the operation of the circuit will be modified or not?
For example, for an inter-stage capacitor of an audio amplifier, the low leakage is more important than the ESR.
In a switched-mode power supply, the ESR can usually be checked in circuit by measuring the ripple and comparing it with the ripple specified by the manufacturer of the device.

It is normally necessary to replace:
- bulged capacitors or with leaking electrolyte
- capacitors placed near a heat source. (power resistors, radiators, ...)
- SMD electrolytics capacitors .... if we find that some are defective, it is better to change them all.

In principle, the capacity, ESR and leakage at rated operating voltage should be checked , not just the ESR.

I repair vintage audio gears, my customers know I refuse to recap their amplifiers and receivers, I only replace what is really faulty.....I have no claim, nor return, all my customers are very satisfied. 

[oldway]

I was speaking about beginners, I doubt about a beginner could be able to design a brand new mass produced device....

It was more about repair, not design.....

[tigr]

Also, ESR varies heavily on temperature, frequency and age.

I'll give an example. Here's the old board from audio equipment. The capacitors 1uF/16v are 35 years old. The table is needed to determine which ones of them bad.

[PA4TIM]

The table is needed to determine which ones of them bad.

Why you NEED it ?  People repaired audio equipment decennia long without ESR tables. The most important thing is how the circuit works. And that can be done with a FG, a scope and a multimeter. If those measurements point at a bad cap you desolder it, measure it with a decent LCR meter and if needed replace it with a suitable new one with similar specs for capacitance, D, voltage, ripple current and size (so it fits).

The problem is that caps have a bad name because the last 10-20 years manufacturers from cheap consumer stuff use caps that are not suited for their job. I have testgear from the 60's still with the original caps. I do commercial repair work, but no consumer stuff,  and I also have to replace caps so now and then but the same goes for transistors, diodes and things like 74/40/45 IC's. Most times the caps are not even really dead. I have not yet seen popped caps (I have seen them but that was only in consumer stuff I repaired for my self or from family)

[madires]

When watching videos about repairing audio gear I sometimes cringe when the expert replaces all electrolytics with low-ESR 105°C "audio grade" caps. In most cases a 85°C standard cap is totally sufficient. Low ESR caps could be even counterproductive in those applications. The ESR tables give just a rough idea about the expected ESR value of a good cap. You don't need to replace a cap if its ESR is off by a few percent. When the capacitance is ok, but the ESR is way off, it's an indication that the cap is dying. A bad cap has most likely a low capacitance and a high ESR. So a capacitance check already finds most bad caps.