Capacitor Leakage, and ESR Explained

[Skywave]

'Dymo' labelling machine, the early one: I remember it well. After having made several long labels, didn't your fingers and wrist ache!

And yes, using 'Letraset': again, pure nostalgia: those little pink and black packs. I might even have one knocking around somewhere, but I'll be surprised if the lettering (and its retention to a clean surface) is any good by now.

Al. / Aug. 3, 2013. //

[Wanderingfree]

Hi Yorkie.
I am a new member and would like to pick your brains please.  You mentioned Bush TR82C  (Mk1) Caps being problematic.
Are there any you would change as a matter of course.  I did note your comment about not changing for the sake of it.

Mike

[Crackle]

I have only just noticed this thread as a result of it being posted to recently.
I thought ESR is the same thing in wax paper as it is in electrolytics, the effective difference being that in electrolytics it is more noticeable because the capacitive values are higher, this coupled with them needing to be reformed will give a very high ESR value.
The cause of a high ESR will, be as already mentioned, allow DC current to flow through the capacitor even after it is charged.
With small value wax paper caps in very high impedance circuits such as the grid coupling capacitor this can be critical even though the actual current may be miniscule. With large electrolytic capacitors in the HT smoothing circuits the ESR of a capacitor that is failing can cause very significant DC currents to flow and have been known to damage rectifiers, speaker energising coils and mains transformer secondaries.

One additional comment I would make is I have often found modern electrolytics to have very poor ESR values and sometimes the 50 year old ones show a lower ESR.

Regarding the Bush TR82C see my comment on your introduction thread.
Mike

[Crackle]

The cause of a high ESR will, be as already mentioned, allow DC current to flow through the capacitor even after it is charged.

I wonder if this sentence in the above post could be changed to;
The result of leakage will be a high ESR and as already mentioned, will result in DC current to flow through the capacitor even after it is charged.

Thanks
Mike

[ppppenguin]

....The cause of a high ESR will, be as already mentioned, allow DC current to flow through the capacitor even after it is charged.

Mike, while much of your explanation makes sense, this bit does not. ESR and leakage are different things. Leakage can be often be modelled as a parallel resistance, but cannot, by its nature be a series one.

This article isn't bad: https://www.avnet.com/wps/portal/abacus/...apacitors/ but I haven't yet found anything online that gives a concise explanation of ESR and why it matters to us when restoring vintage equipment.

The article I cited has a nice diagram showing the imperfections of a real capacitor. Series and parallel resistance, plus series inductance. How perfect does a cpacitor need to be? Depends on the application, there isn't a single short answer. Behind this lies questions like "Why do we mak such horrible capacitors?" The answer is nearly always because we want a lot of capacitance in a small space at low cost. And we're prepared to trade off other things against that.

An example...
A reservoir capacitor will often be 100s or 1000s of microfarads. Totally impractical other than as an electrolytic. But electrolytic must be conencted the right way round, are leaky by their nature and will have series resistance which tends to increase with age. A small leakage won't matter -  what of the odd milliamp when you're dealing with amps. But that cap has a lot of current flowing in and out of it. The series resistance may be called "equivalent" but it will still make heat when current flows through it. Heat isn't good for capacitors, expecially electrolytics. So you need low ESR in this role.

So we all need to understand the whys and wherefores of capacitors when we're working on vintage kit. Or modern kit for that matter. I've got a lot of experience and I still run into capcacitor problems. See this thread: https://golbornevintageradio.co.uk/forum...p?tid=8219

So don't be ashamed of not understanding, please ask for help and we'll do our best. Even if we're not always right.

[Crackle]

Mike, while much of your explanation makes sense, this bit does not. ESR and leakage are different things. Leakage can be often be modelled as a parallel resistance, but cannot, by its nature be a series one.

So am I right in assuming that Electrolytics can have both ESR and leakage properties then, otherwise if it was just ESR it would not display leakage properties, i.e. passing DC current when reformed.

Mike

PS. the links in post #6 & #10 dont appear to work any longer, I wonder if that can be fixed.
Edit by mods: links now fixed

[ppppenguin]

Mike, all electrolytics leak to some extent, it's part of how they work. Too much leakage is obviously harmful. I've seen tables and/or rules of thumb for how much leakage is tolerable for a particular size and rating of capacitor but can't find them when I want them.

Likewise all capacitors have ESR. It's a question of how much is tolerable for your application.

For most vintage applications, neither leakage nor ESR is usually too critical. For a switchmode PSU, ESR matters a lot.

PS: This data sheet has some leakage numbers: https://www.mouser.com/datasheet/2/88/WBR-32673.pdf

[Kalee20]

Mike, while much of your explanation makes sense, this bit does not. ESR and leakage are different things. Leakage can be often be modelled as a parallel resistance, but cannot, by its nature be a series one.

So am I right in assuming that Electrolytics can have both ESR and leakage properties then, otherwise if it was just ESR it would not display leakage properties, i.e. passing DC current...

Yes. Both exist (all capacitors, actually, but electrolytics are more noticeable).

Leakage results in the capacitor slowly discharging over time, instead of just sitting on the shelf and staying at the voltage they were left at - or, putting it another way, passing a bit of DC in a circuit. It can be modelled, crudely, as Equivalent Parallel Resistance.

ESR - Equivalent Series Resistance - stops the capacitor's impedance from falling towards zero as frequency rises. It's the resistance if the leadout wires; the resistance of the metallising in film capacitors, and in electrolytics, the significant resistance of the electrolyte (which we'd like to be a perfect conductor, but unfortunately isn't).

ESR is what makes a capacitor get warm when it is passing lots of ripple current in a power supply, and it's why manufacturers put a limit to this.