It is currently Thu Sep 21, 2017 2:56 pm

What's new about ESR?

Valves, Transistors, IC's and Components. Bench test techniques, Test Gear - Valve Testers, Scopes, CRT Testers, Bridges, Sig Gens, Pattern Gens, Workshop tours etc
 
Posts: 3781
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

What's new about ESR?

Post by Terrykc » Mon Oct 15, 2012 2:12 pm

Yesterday, Mike (Crackle) produced the batch of capacitors he changed to cure his recent microwave oven problem and demonstrated the bad ESR.

It became apparent that, although everybody knows that ESR is BAD for you, it is not obvious what is is or even what the initials stand for - although the fact that is measured in ohms is a good pointer! (Actually, it is Equivalent Series Resistance ...)

It doesn't help that, if you Google ESR, all you seem to turn up are very complicated articles riddled with complex formulae!

Another thing is that, to anybody who has dabbled in electronics for a long time - and particularly anybody who has come back to it after a long break - it is a relatively new thing that was never heard of years ago.

That doesn't mean that ESR didn't exist but that it was something that nobody needed to know anything about - unless, of course, you were a capacitor manufacturer or a designer of particularly demanding equipment.

There is no such thing as the perfect capacitor. A practical capacitor will have all sorts of unwanted 'additions' - pure resistance and inductance, for example. The effect of all of these is lumped together to give the ESR, in other words, the equivalent - I prefer effective - total series resistance between the capacitor proper and the outside world. ESR is not restricted to, but is of particular importance to us in connection with, electrolytics. Of course, the current flowing into and out of the capacitor creates a voltage drop across the ESR, which creates heat ...

Although the term wasn't in common use, the effects of ESR are prominent in the marking of one important component in every vintage piece of mains operated equipment. Just take a look at the smoothing can - assuming it is the usual multiple type - and note the red paint dot on the reservoir section. There will be one, even if both sections are the same value, as they commonly were. It was imperative that this section was connected to the HT rectifier because of its superior ripple current rating - often marked on the can.

The reservoir cap was the section wound on the outside of the others, so that the can could be used to help dissipate heat to the atmosphere but, of course, it was the ESR that determined how much heat was produced!


As the capacitor aged, its capacitance could drop or its ESR could increase - or both. The effect on the radio or TV would be much the same in any event, which would result in replacement. Everybody assumed the capacitor value had dropped, but it didn't matter anyway - nobody had an ESR meter and, in truth, nobody needed one!

So what changed?

The answer to that is the introduction of the Switch Mode Power Supply. ESR increases with frequency - note the inductive component I mentioned earlier on. Our old reservoir capacitor was quite happy with its low 50 or 100Hz charge/discharge cycle but increase that a thousandfold and everything changes!

Add an ever increasing demand for miniaturisation - thus more heat inside and around ever smaller capacitors and ESR becomes a very critical factor.

In Mike's case, the ESR increased so much and destroyed the capacitor's efficiency so much that, although his oven failed, the failure mechanism was self limiting. Indeed, it is what probably happens with most caps or they'd be exploding by the million every day!

So, what's new about ESR? Nothing! What is new is its increasing importance do anybody who want to dable with the more modern forms of electrickery ...

I've tried to keep this simple and, in all honesty, made it up as I went along but I hope it explains what this curious ESR thing is. If anybody wants to challenge the fine detail, might I respectfully point them in the direction of Google and all those complex equations ...?
Last edited by CTV on Tue Nov 05, 2013 7:56 am, edited 5 times in total.
Reason: Not really a lounge subject more of a workshop toolbox talk type thread so I moved it. It was suggested by another forum member that the information regarding the reservoir capacitor was important for beginnes so I highlighted it.

 
Posts: 7156
Joined: Sun Aug 14, 2011 11:02 am
Location: Co. Limerick

Re: What's new about ESR?

Post by Michael Watterson » Mon Oct 15, 2012 2:49 pm

Also Terry,
1)Power has maybe gone up, maybe 20W to 100W instead of 8W
2)Voltage has gone down, 3.3, 5, 12 20V rather than 200V to 500V
3)Pentode output stage or Electrodynamic speakers reduce effect of Hum and earlier stages and O/P screen grid likely have more decoupling.


This means the ESR needs to be 20 to 1000 times lower on "modern" equipment!

Also a valve set had well ventilated set and died quickly in an enclosed space. A set-box may be in an enclosed space and while more efficient, the capacitors may actually be at a higher ambient temperature.

In 1950s and 1960s the TV was only on maybe a short while in afternoon and evening. My mother (80) puts on both TVs when she gets up and turns them off when she goes to bed.
Last edited by Michael Watterson on Mon Oct 15, 2012 2:52 pm, edited 1 time in total.

 
Posts: 423
Joined: Thu Aug 04, 2011 8:50 pm
Location: Evesham

Re: What's new about ESR?

Post by Mark Hennessy » Mon Oct 15, 2012 2:49 pm

What's new is the very high currents being asked of small capacitors.

By "new", we're talking about the last 30 years, but that's new in the context of this forum.

By "large", we're talking about supplies that might be 10s of amps.

And by "small capacitors", we're talking about 1000uF or so...

Because these supplies run at 10s or 100s of kHz, the energy supplied per cycle is much less, so you need radically less capacitance for a given current output compared to a conventional 50/100Hz PSU. But as ESR is proportional to capacitance (very, very roughly speaking), the sorts of capacitors that might have seemed OK based on just volts and micro-Farads considerations turned out to be poor in switched mode power supplies, and the reason for that is ESR.

Big capacitors - the 10,000uF bottles you'd find in an old Quad 405, for example - did have very low ESR, but ESR wasn't the most important parameter in a linear power supply...

Over the years, capacitors have emerged that have much lower values of ESR, and these are principally used in power supplies. They have electrolytes with much better conductivity, and are normally rated to run at higher temperatures too...

So, if you have an ESR of 0,1 ohms, and you're putting a mean current of 10 amps through one of these, it will dissipate 0.1 watts of heat. Not a huge amount, you might think, but it will cause some self-heating. And this self-heating, added to the surrounding heat of the immediate environment, will cause the electrolyte to warm up and slowly evaporate. As the electrolyte largely determines the ESR of a typical capacitor, you can see that a capacitor that is literally drying out will have a higher ESR than it did when it was new...

But as soon as the ESR starts to rise, what happens? More self-heating (I-squared R), so more heat to evaporate the electrolyte. Which means the drying out process is accelerated - it's like positive feedback. And, sooner or later, the surrounding circuit will become unhappy with this, as it requires a certain value of ESR to perform to design...

People are very quick to blame capacitors for being very unreliable - and the "badcaps" scandal didn't exactly help with that - but the point is we are asking very much more of our capacitors than we ever used to...

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: What's new about ESR?

Post by sideband » Mon Oct 15, 2012 3:22 pm

Terrykc wrote:I've tried to keep this simple and, in all honesty, made it up as I went along but I hope it explains what this curious ESR thing is. If anybody wants to challenge the fine detail, might I respectfully point them in the direction of Google and all those complex equations ...?


Simple but very effective and straight to the most important points! Excellent...might be worth making this a sticky as this question crops up a lot. You're right, it isn't new but more people are hearing about it now and believe that it might be new.

ESR is less important on the old electronics that we deal with. It is very important, as you quite rightly say, in switched mode supplies. The old favourite being the start-up capacitor in certain VCR PSU's. Everything OK until the machine is disconnected from the mains and upon reconnection, it flatly refuses to do anything. What I always try to convey to people is that testing an electrolytic on their digital meter won't tell them a lot unless it's open circuit or short. It could well indicate near normal value but still refuse to work in a switched mode circuit. Also be aware that the caps in a SMPS are often low ESR 105 degree types. Don't fit 85 degree types...chances are the circuit won't work (been there done that in the early days).

If you are still repairing SMPS, then an ESR meter is a useful addition to the test kit. It isn't really necessary for vintage gear though.


Rich.

 
Posts: 423
Joined: Thu Aug 04, 2011 8:50 pm
Location: Evesham

Re: What's new about ESR?

Post by Mark Hennessy » Mon Oct 15, 2012 3:27 pm

sideband wrote:The old favourite being the start-up capacitor in certain VCR PSU's. Everything OK until the machine is disconnected from the mains and upon reconnection, it flatly refuses to do anything.


Start-up problems are often caused by high value resistors going high or O/C. Normally in the ~300k region. And nasty because often there isn't anything else in the supply to discharge the main smoothing capacitor, that sits there with ~325V on it for days on end...

 
Posts: 3781
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: What's new about ESR?

Post by Terrykc » Mon Oct 15, 2012 4:18 pm

sideband wrote:... What I always try to convey to people is that testing an electrolytic on their digital meter won't tell them a lot unless it's open circuit or short. It could well indicate near normal value but still refuse to work in a switched mode circuit ...


Yes, Rich, high ESR and low capacity don't have to go hand in hand. It is perfectly possible to find a capacitor with 100% of its rated capacity* yet a very high ESR

sideband wrote:... Also be aware that the caps in a SMPS are often low ESR 105 degree types. Don't fit 85 degree types...chances are the circuit won't work (been there done that in the early days) ...

Yes, I alluded to high temperatures but didn't attach a figure to it.

* Of course, electrolytic tolerances are quite wide, so finding one that is close to 100% of nominal doesn't mean it hasn't lost capacity!

Years ago, it wasn't unusual to find the main smoothing rated at +100% -50% tolerance! Today, ±20% would be common but a look at the selection parameters on Farnell's website for Aluminium Electrolytic Capacitors gives this selection, which shows that wide tolerances are still not unusual. (numbers in brackets relate to numbers of products listed.)

+100, -10% (1)
+150, -10% (2)
+30, -10% (4)
+50%, -10% (92)
+75, -10% (162)
-10%, +30% (70)
-10%, +50% (59)
-10%, +75% (1)
± 10% (55)
± 20% (8873)

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: Symbols and their usage

Post by sideband » Mon Oct 15, 2012 8:44 pm

I'll let this thread run a little longer and then I'll make it a sticky so it can be used as a reference. After it becomes 'sticky' if anyone else has information to add, please contact one of the mods or Chris with the info and we can add it.


Rich

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: What's new about ESR?

Post by sideband » Mon Oct 15, 2012 10:02 pm

Further information added by jimmc.

Congratulations on a well written article Terry.

But, doesn't ESR decrease slightly as the frequency increases (at least over the normal range of operating frequencies)?

I can't find a simple example but this note by nichicon contains some relevant data...
http://www.nichicon.co.jp/english/produ ... uminum.pdf

see PDF pages
p06, 1-6-2, fig 1.8
p07, 1-6-4, fig 1.11
p22, 2.9.5, para 2

Jim

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: What's new about ESR?

Post by sideband » Sun Nov 04, 2012 8:55 pm

Further information added by Red to Black.

A very informative thread Terry,
I would just like to add the following comments:

Strangely enough when the ESR rises I have found the capacitance also rises, significantly in some cases, I am not sure by what mechansim this occurs.

I have witnessed this phenomena first hand with over several years usage of an ESR meter, I was using the Peak Atlas ESR meter, as this measures the capacitance as well as ESR.

I have also compared the 'old component' against new components, and other old components of the same make, type, value and age on the same PCB (but in different positions), on the same item in for repair.

I wonder if this could be just an anomaly with the measurement technique of this particular meter ?

Any thoughts or explanations of this phenomena would be welcome.

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: What's new about ESR?

Post by sideband » Sun Nov 04, 2012 11:38 pm

Further information added by Michael Watterson


Strangely enough when the ESR rises I have found the capacitance also rises, significantly in some cases, I am not sure by what mechansim this occurs.


Most modern meters read high with a leaky capacitor. Yes, you can have high series resistance AND low Parallel Resistance.

100 Ohms (Series) is very high ESR.
100K ohms is very high leakage, a low parallel resistance. Even 1k (a short really) is still high ESR.

You can verify this using actual resistors on a known good capacitor.

If a capacitor dries out it may look like a resistor?

 
Posts: 2920
Joined: Wed Jul 27, 2011 7:13 pm
Location: Croydon

Re: What's new about ESR?

Post by sideband » Mon Nov 05, 2012 9:13 pm

Further information added by refugee


If you do not own an ESR meter don't panic there is another way.
Put a scope on the suspect capacitor and measure the ripple.
If it is greater than 20% of the expected DC voltage that should be across the capacitor it has gone high ESR. This would have to be estimated to 5% of the expected voltage if the rail is monitored by an opto coupler as this will be the critical rail in the circuit.
If the PSU is ticking like a clock the trace will be diagonal on the scope screen and the ripple will have to be estimated from a display more like the IF waveform in an AM radio. Also the sweep could be increased on the scope and the brightness increased if the tube is good enough and the rise times be observed. A smoothing capacitor on a modern power supply should not have fast rise times across the smoothing capacitors. If they do the ESR is high and the trace will have vertical bits.
On 240V the pulses these things have to handle are twice the current and half the duration of the ones found in 110V and units that are rated 100V to 250V without a voltage adjustment are the worst for this. The capacitors get a hard life in these type of power supplies when running on 240V.


Return to In The Workshop



Who is online

Users browsing this forum: No registered users and 1 guest