• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Split Power Supply for ESR Meter

Status
Not open for further replies.

JimB

Super Moderator
Most Helpful Member
Also, how does this thing know the difference between ESR and Xc? At 100 kHz, a 0.1 uF cap has Xc = 15.9 ohms, but the ESR has to be less than 1% of that. So what is this circuit actually measuring? Or am I missing something...?
A VERY good question.
Which was skirted around in the thread:
http://www.electro-tech-online.com/...lytics-not-specified-esr.145523/#post-1230336
but never really addressed fully.

I think the practical pragmatic answer is that when ESR is measured in the hobbyist world it is in connection with finding "bad electrolytic caps" which have capacitance values which are large, and so have low reactance which has a negligible effect on the impedance as measured by an "ESR Meter".


JimB
 

Nigel Goodwin

Super Moderator
Most Helpful Member
At 100 kHz, a 0.1 uF cap has Xc = 15.9 ohms, but the ESR has to be less than 1% of that. So what is this circuit actually measuring? Or am I missing something...?
It's measuring ESR, but a 0.1uF is too small to matter, and it's really for electrolytic's only - and generally larger ones.

The normal frequency used is 100KHz, and sine is preferred - see the PDF project I posted in one of the other ESR threads.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
I think the practical pragmatic answer is that when ESR is measured in the hobbyist world it is in connection with finding "bad electrolytic caps" which have capacitance values which are large, and so have low reactance which has a negligible effect on the impedance as measured by an "ESR Meter".
It's the same in the professional world as well :D

ESR has only become important because of the use of switch-mode PSU's, and the MASSIVE failure rates of electrolytic's (going high ESR).
 

Musicmanager

Active Member
The normal frequency used is 100KHz, and sine is preferred - see the PDF project I posted in one of the other ESR threads.
NG .. .. Can you give me a map reference please :)

S

Edit: As you were .. I downloaded the article a while ago .. good stuff
 
Last edited:

large_ghostman

Well-Known Member
Most Helpful Member
It's the same in the professional world as well :D

ESR has only become important because of the use of switch-mode PSU's, and the MASSIVE failure rates of electrolytic's (going high ESR).
Lucky for the hobbyist that many cap manufactures have invented an included a little dome that can normally be relied upon to pop up when a cap is on the way out! Some of mine even have a little cross on the top that is kind enough to open up so you can look inside and determine if its gone bad :p.
Those pesky old caps that dont have the modern features often last so long you need a meter to see when they are bad :p
 

Musicmanager

Active Member
I think the practical pragmatic answer is that when ESR is measured in the hobbyist world it is in connection with finding "bad electrolytic caps" which have capacitance values which are large, and so have low reactance which has a negligible effect on the impedance as measured by an "ESR Meter".
But isn't it also true, that the only relevance of ESR, whether to the hobbyist or in the professional world, is the effect it has on the efficiency of the component capacitor and therefore on the circuit as a whole. It seems to me that there has been great store placed in the ability to seek and find inappropriate ESR 'in circuit' and surely that will only ever occur during a repair. For example, I've just successfully repaired a SMPS belonging to a vacuum cleaner charger. I found a 22uf 400v capacitor which was completely open circuit, presumably an ESR meter would have been 'off the scale' .. I only discovered the faulty cap by removing it from the board and trying it with a DMM.

S
 

KeepItSimpleStupid

Well-Known Member
Most Helpful Member
The devil is always in the details. Some caps should never be used as the filter for an SMPS secondary. ESR is generally frequency dependent.
 

AnalogKid

Well-Known Member
Most Helpful Member
The devil is always in the details. Some caps should never be used as the filter for an SMPS secondary. ESR is generally frequency dependent.
Which goes back to my question about the "right" excitation freq for this kind of tester.

ak
 
Last edited:

MikeMl

Well-Known Member
Most Helpful Member
... I only discovered the faulty cap by removing it from the board and trying it with a DMM.

S
And who is to say that you didn't kill the capacitor as you de-soldered it?

In the airplane world, we call it "maintenance induced failure"...
 

KeepItSimpleStupid

Well-Known Member
Most Helpful Member
But the SOP would probably say to replace it with a new one.

I did know a guy (Friends father) that was an Avionics tech.
 

Nigel Goodwin

Super Moderator
Most Helpful Member
If de-soldering might have killed it, why not initial soldering ? Are they really that vulnerable ?

In my world, we call that hamfisted !! :D
If someone is that hamfisted, then they shouldn't be near anything electronic - or indeed anything more 'fragile' than a granite boulder :p

I seriously doubt it would be possible to damage a capacitor by de-soldering it - with the obvious exception of ripping the pins out of it :D (but that's not 'de-soldering'. it's merely ripping the capacitor out of the board).
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Lucky for the hobbyist that many cap manufactures have invented an included a little dome that can normally be relied upon to pop up when a cap is on the way out!
Only happened sometimes - however, if the cap is domed, then it's duff (even if it doesn't test as such, it should be replaced) - but if you only replace the domed ones you're going to have a LOT of repeat failures.
 

Musicmanager

Active Member
Morning NG

On form this morning !!

I've been thinking about this in the light of AK's explanation of the circuit function and I've got a couple of conclusions .. ..

1. The Amidon ferrite core listed in the alternative circuit does not seem to be easily available in the UK ..

2. And more especially, I think it might be quite interesting to expand the circuit I've built, using fixed and variable resistors in series and 'tweak' each resistance to a specified and some cases matching value and examine the results. Although this will probably produce an impractical piece of equipment to use through physical size, might be interesting and informative nevertheless.

Any comments ?

S
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Morning NG

On form this morning !!

I've been thinking about this in the light of AK's explanation of the circuit function and I've got a couple of conclusions .. ..

1. The Amidon ferrite core listed in the alternative circuit does not seem to be easily available in the UK ..
It seems a pointless complication, although Americans seem very found of toroidal cores for some reason :D

2. And more especially, I think it might be quite interesting to expand the circuit I've built, using fixed and variable resistors in series and 'tweak' each resistance to a specified and some cases matching value and examine the results. Although this will probably produce an impractical piece of equipment to use through physical size, might be interesting and informative nevertheless.

Any comments ?
Why? - you stick it across a suspect cap, and read the result - how would complicating that process help?.

It's not rocket science, you don't want to know if it's 0.05 or 0.06 ohms - you want to know if it's 0.05 ohms or 7 ohms :D

They are really VERY easy to use, and I threw my home made one away and bought a Peak one (which is absolutely fabulous).
 

Musicmanager

Active Member
Why? - you stick it across a suspect cap, and read the result - how would complicating that process help?
Fully acknowledged, as an ESR meter it wouldn't help at all. However .. .. .

It's not rocket science, you don't want to know if it's 0.05 or 0.06 ohms - you want to know if it's 0.05 ohms or 7 ohms
For me, in a sense, that's exactly what it is ! AK provided a blow by blow explanation of the circuit which I have taken in and vaguely understood .. .. but to be able to 'tweak' the values and examine the results would help to clarify my understanding in some detail .. .. .. remember what you called me yesterday ! :)

And I still don't have an effective working ESR meter !!

S
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Fully acknowledged, as an ESR meter it wouldn't help at all. However .. .. .



For me, in a sense, that's exactly what it is ! AK provided a blow by blow explanation of the circuit which I have taken in and vaguely understood .. .. but to be able to 'tweak' the values and examine the results would help to clarify my understanding in some detail .. ..
I think the example you're building makes life far more complicated by it's use of a bridge circuit - the tried and tested PDF example I posted (which even includes a veroboard layout) is far easier to understand.

remember what you called me yesterday ! :)
Actually, I didn't call 'you' any such thing :D - I was referring to people FAR, FAR worse, sorry you got the impression it was aimed at you.
 

Musicmanager

Active Member
Hi NG

I think the example you're building makes life far more complicated by it's use of a bridge circuit - the tried and tested PDF example I posted (which even includes a veroboard layout) is far easier to understand.
Yes, agreed ... .. and that is in the plan ! I just think there is a learning opportunity for me in following up the other one and it's problems .. that way, one day I'll look at a circuit and be able to make the right choice first time !

Actually, I didn't call 'you' any such thing :D
Do you have a funny sensation in your left leg -- like someone's pulling it !! ??

:D S
 

AnalogKid

Well-Known Member
Most Helpful Member
If you want to stick with your original circuit, there are a couple of small changes that will improve its performance. It has two balanced bridge circuits in series, and the important thing to know when you are not chasing millivolts or microvolts of precision is that it is the balance, or ratio, of the two sides that matter, not the absolute values of the components.

For the sensor bridge, reduce R10 and R11 to 15 Ohms each and insert a 10 ohm pot between them, with the low end of the bridge taken off the wiper. This will let you balance out any tolerance errors in all 4 fixed resistors with one adjustment. You could just replace R10 and R11 with a 50 ohm pot, but this way gives you better adjustability.

For the diff amp (which is a balanced bridge circuit with gain), again it is the ratios that matter most. If R15/R12 = R14/R13 you will get maximum common mode cancellation. Having perfectly matched absolute values improves the differential input impedance, reduces input bias current errors, and gives you predictable and repeatable gain, but again that is the stuff of higher precision needs. Inserting a 100 Ohm pot in between either resistor pair (but not both) with the opamp pin connected to the wiper will give you a single point balance adjustment.

If cost and/or size is important, put all of the component tolerance corrections in one basket: lose the bridge pot and keep the diff amp pot. Not preferred, but better than nuttin.

To adjust the bridges, first dead-short the R8-R10 node to the R9-R11 node, disconnect the oscillator, and tweak the diff amp pot for 0 Vdc at pin 7 (measured wrt the circuit virtual ground). you should be able to swing the output above and below 0 with the pot. Reconnect the oscillator and retweak for minimum bleed-through (there shouldn't be any, but good to check). Next, remove the shorts and adjust the bridge pot for minimum signal output at pin 7.

ak
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top