Capacitor Replacement: ESR, Capacitance & Thresholds (47uF/16V Example) - When to Change?

Good morning,
when replacing capacitors, check their capacitance and ESR. I wonder when a capacitor with increased ESR is suitable for replacement.

I suggest this table:


It is a pity that there is no specified ESR parameter range that you can stick to. Here a fixed number is given and the more it deviates from it, the lower the quality of such a capacitor. The question is, how wide should such a border be?

For example, I now have a CD on the table, I desoldered some of the capacitors for verification. These are 47uF/16V capacitors and according to table ESR should be 1 Ohm, with mine it is on average 2.1-2.2 Ohm. Not much more, but more than 100% of what the table shows.
Capacitance relatively preserved, 45-47uF.

Is such a capacitor eligible for replacement?
Are there any limits to the ESR parameter that should be strictly adhered to?

In my opinion, it depends on the function of the capacitor in the system.
If it is a capacitor that filters the power supply, then with such an increase there will be ripples and the capacitor will heat up, if it is a coupling capacitor, it will still work.

Capacitors probably filter the power supply, I don't have a schematic and can't find it anywhere.
I ordered a few pieces of new ELNs, I will measure them and compare them to the 26-year-old Nippons.

For comparison, from the same equipment, ELWA has a slightly increased capacity and ESR lower than that in the table.

Added after 14 [hours] 9 [minutes]:

I have one more question, can the new capacitor, as a result of work in the system, burn-in reduce its ESR? Or just the other way around?

Added after 3 [hours] 8 [minutes]:

I measured the new Nichicon FG electrolytes, they have an ESR of 0.91, so the choice is clear.
I would be grateful if someone could tell what possible range of ESR is acceptable for capacitors, any table would be welcome.

Hello

From what I noticed capacitors that worked a bit have lower esr than new ones. These were probably panasonic FC and FM, measured with the atlas esr meter, but these were not big changes.
The best source of information, which should be the esr, is the documentation, but this is a problem with older capacitors.

Regards, Romek

And if the ESR is 0, is it already to be replaced?

jakubtopolski05 wrote:

And if the ESR is 0, it is already to be replaced?

.
Is this a mockery or a genuine question?

>>21447144 When you were a beginner, did they write you like that too? Master yourself

Until switched-mode power supplies came into daily use displacing continuous power supplies, hardly anyone considered the ESR parameter. No one was bothered by capacitors 10, 20 or older, except of course in cases that necessitated replacement.

It is easy to buy any meter today, so anyone can try their hand at it. One sometimes reads posts that the replacement has significantly improved the sound of the amplifier.
Maybe it did, after all there are new capacitors with low ESR. .
It doesn't matter where, if the electrolyte is the one to replace.

And this is how home-grown professionals often replace old, often good quality capacitors with modern crap, but new and importantly with low ESR.

And the reality is that this parameter is only important in impulse circuits, where a large current flows through the capacitor. This is the case in switching power supplies when the capacitor is charged by a pulse with an almost rectangular slope.
This is what we are dealing with in impulse power supplies on the secondary side. This is where improved capacitors are needed to minimise the heating of the electrolyte leading to gradual drying out and loss of capacity.

jakubtopolski05 wrote:

>>21447144 When you were a beginner, did they write you like that too? Get a grip

.
Some difference, because I started more than half a century ago and imagine there was no internet and everything was looked up in magazines or books.

>>21447144 Sometimes it is better to refrain from hastily criticising others.

Use an instrument, not a tester

HD-VIDEO wrote:

Use an instrument not a tester

Then perhaps this is how you should have answered the colleague who asked the question?
I proved that the question may have been legitimate.