Advice on ESR meter, I've seen all chinese stuff on ebay..

[mariush]

You will know when that happens because the LCR meter measures capacitance at the same time. If it measures a capacitance much higher than the marked cap rating, the cap will have to be isolated before testing.

I mention it because Blue ESR only measures ESR, so if he goes that route he won't see the capacitance. The ESR70 and ESR Micro I have report both and it helps.

[The Electrician]

You're mixing up two things. If you want to know if a capacitor is defective/out of spec, ESR alone is meaningless and D is a much better indicator. D is measured at 120 Hz as per the datasheet, e.g.: http://industrial.panasonic.com/www-data/pdf/ABA0000/ABA0000CE132.pdf

Even in the limited context of detecting a bad capacitor, I disagree that ESR is meaningless.  The very definition of D is D=ESR/Xc.  If the capacitance of the capacitor doesn't change, then the only way D can increase is for ESR to increase.  How then can you say that ESR is meaningless?

And, sometimes as capacitors fail, their capacitance decreases along with an increase in ESR.  If a capacitor's capacitance decreases, Xc increases and that decreases D, offsetting some of the increase in D caused by the increase in ESR.

Furthermore, as the fellow in the video I linked above says, when electrolytics fail, typically ESR doesn't just increase a little; it increases a lot.  The low cost so-called ESR meters have their deficiencies for design work, but as a repair tool they do a great job by just measuring ESR (or a good approximation to ESR).

I don't disagree that D will undoubtedly increase greatly when an electrolytic goes bad, but the low cost ESR meters don't measure D directly.  What I disagree about is your assertion that ESR is meaningless for detecting bad caps.

BTW, nice vector impedance analyzer. What make/model is that?

It's a Hioki IM3570: http://www.hioki.com/products/lcr_resistance_signal/lcr_resistance_meters/654

One way that a capacitor could have a high dissipation factor and still be ok is if the capacitor has large capacitance and therefor a low SRF.  Then, at 120 Hz, Xc might still be not much more that ESR.  I dug out an 18,000 uF capacitor and got this result from a sweep on the analyzer:



At 120 Hz, D is a little more than .5, yet this cap is perfectly good.

[Wytnucls]

Check the datasheet. There is no way D will be 0.5 for a good cap. If it is an old stock one, depending on voltage rating, D could go up to 0.4 max and still be within spec.

[PA4TIM]

First, the TS writes there is no other way to find bad caps. That is not true. If you know what you are doing you do not need an ESR meter. It is a handy replacement for knowledge. Most times you replace more caps as needed but that is often not bad. In consumer stuff, if a few caps fail, the rest will follow. Just replace them all at once instead of a few every now and then.

I repair a lot, but most of it is test gear, not much consumer electronics, and I never measure ESR in situ.  I find most problems using a scope and multimeters. And if it does not run, I desolder the caps, reform them, test for leakage, capacitance , D and ESR. If it can be powered I use a scope and if find a problem regarding caps, I desolder it and then I start measuring it but that is because I like to study component behauvior and then want to know what en why it fails.   (I have a lot of capacitance/impedance/ESR etc mearsurment gear)

I think both partys are right. ESR is meaningless and ESR is important. Depents on your point of view.

Electrician is right, ESR can be an important parameter in design.  In a lot of cases it must be low to minimise dissipation. Sometimes it must be high to keep something from oscillating. So if you repair a circuit, it can be important to use a cap with the right ESR. And that means, the same ESR. Low ESR is not defined. If I make caps and my best one is 1000 Ohm, then I am allowed to call it a low ESR cap. While an other brand can makes a cap with 0.000001 Ohm ESR, so his 0.01 Ohm version is his high ESR cap.

But in case you measure ESR need to know the specs to make a judgement, or have enough knowledge to know what would be right. Most 1uF caps are sloughtered after false accuse with an ESR meter. My IET and my homemade ESR meter measure 1 u with out problems. My homemade down too 100 nF and less accurate even lower. The IET DE-5000 measures everything you feed it. Most of my bridges do too.

ESR as a value while searcing for faults in a circuit is a bit meaningless. OK, it can be handy as a replacement for knowledge  about fault finding. But it is meaningless because you must know the ESR. And most users only use some table on the meter. But who cares. If it is 1 to 5 years old an caps fail, replace them all or junk it.

And that is the funny part. Most ESR meters measure impedance at 100 kHz and do not measure ESR.
The datasheet often states the impedance at 100 kHz. That is not the same as ESR. ESR is a part of impedance.

The datasheet states D,  sometimes also ESR as it is related to D, but at 100, 120 or 1000 Hz, not at 100 kHz.
So real ESR meters that measure at 100 kHz give you real ESR but that is a number that most times can not be verified for 100 kHz. And the not-ESR meters that measure impedance , that is in the datasheet at 100 kHz, tell you they measure ESR instead of impedance. 

I agree with the people here, that do prefere D (dissipation factor or DF, if you know it, it's a very logical parameter. )over ESR. I do too. I have several ESR meters, have a collection of bad documenteed caps and a whole bunch of bridges and analysers for caps but I use D.

There are a few failure modes:
- low capacitance
- DC leakage
- high D (and ESR)
- mecanical failures
These things can be on their own or combined.

There is one pitfall, many people who repair consumer stuff measure capacitance with a multimeter that uses DC. If ESR gets up, and if capacitance decreases also, the meter is fooled by the ESR and gives a capacitance that can be the double of what it is. The same is true for not vectorial impedance meters.

I have seen caps that failed in one of the modes above and caps that have more of the problems.
I have seen caps with a great ESR. But totally shorted for a few volts DC. The ESR meter will not find it. The capacitance meter with an in circuit usable signal amplitude will not alway find it. capacitance can be good. Only a leak test will find it. And often a good bridge will too because they some can generate a rather high voltage.
I have seen high ESR and good capacitance and no leakage. A high voltage cap had good ers and good capacitance but it looked like a geiser in circuit (blowing a lot of smoke ) It did not leak DC far below the working voltage but around 10V below the working voltage in circuit, it started leaking severe, You would not have find it if you had not seen it blow smoke and only used ESR OR lcr meter.

That is why I think only measuring ESR in situ is a bit meaningless.

http://www.pa4tim.nl/?p=3775 my page about the ins and outs of ESR measurement.
And about 1000 other topics here about ESR

[Jay_Diddy_B]

Hi,

I have spent some time studying and designing ESR meters. I also own HP 4274A and HP 4275A LCR(Z) meters. The HP  LCR meters have no trouble at all making very accurate measurements on components that are out of circuit. But this does not mean that that they are my first choice when troubleshooting.

1) A good ESR meter has some attributes that a LCR meter does not have:

It should have protection from charged capacitors and applied voltage.

It should be able to measure in circuit.

2) It does need to be super accurate to find bad capacitor.

I agree with PA4TIM that voltage measurement with a scope can be used to find bad capacitors. The ESR meter can then be used to confirm that the capacitors are bad. The ESR meter has the some advantage that bad capacitors can be found by testing without apply power to the unit being tested.

I have posted a couple of DIY designs for ESR meters on the forum.

The 5 transistor ESR meter, use an analog meter, and has limited protection:

https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/msg171364/#msg171364

My later design, ESR meter adapter, has 'best in class' protection and is very accurate.

https://www.eevblog.com/forum/projects/esr-meter-adapter-design-and-construction/msg341177/#msg341177

Have you considered building your own ESR meter?

Regards,

Jay_Diddy_B

[The Electrician]

Check the datasheet. There is no way D will be 0.5 for a good cap.

Check the data on these catalog pages from Cornell-Dubilier: http://www.cde.com/catalogs/CGS.pdf

Look under the 25 volt heading; the cap I measured is a 25 volt cap.  I see a number of capacitors with typical (not max) D greater than .5.  For example:

Code: [Select]

Capacitance (uF)       ESR      D
3000                  .235    .53
13000                 .082    .80
18000                 .052    .71
22000                 .043    .71

[Shock]

Dssence I have been through the same processes you are going through.

I brought a couple of the Chinese $25 ESR meters because they are good enough for most jobs (and in many cases upgradable).
They measure capacitance, ESR and also check transistors, diodes, resistors etc.
There is a big thread on them here: https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/
I also have a precision bench LCR meter which can pull of a few more tricks on top of that.

I will end up buying a DER EE 5000 as it's portable the best value by far and has a working 4 wire kelvin solution if I need to take an accurate measurement.
I recommend going straight to the DER EE 5000, or pick up a Chinese one if you have budget limitations or want a backup or a side project.

If not the DER EE 5000 I would get a Mastech MS5308 (it has a few quirks though and is not as good value).
If you choose the DER EE 5000 consider power supply and cable options before you decide on a seller as this can blow out the cost.

[Wytnucls]

It is so with high capacitance caps.
The datasheet actually tells you to add a factor to the dissipation factor when the capacitance is above 1000uF.
0.02 x 18, in this particular case, which would take the ballpark dissipation factor to about 0.61 for a good cap.
So the system still works, as long as you make the correction for caps above 1000uF.

[dssence]

Dssence I have been through the same processes you are going through.

I brought a couple of the Chinese $25 ESR meters because they are good enough for most jobs (and in many cases upgradable).
They measure capacitance, ESR and also check transistors, diodes, resistors etc.
There is a big thread on them here: https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/
I also have a precision bench LCR meter which can pull of a few more tricks on top of that.

I will end up buying a DER EE 5000 as it's portable the best value by far and has a working 4 wire kelvin solution if I need to take an accurate measurement.
I recommend going straight to the DER EE 5000, or pick up a Chinese one if you have budget limitations or want a backup or a side project.

If not the DER EE 5000 I would get a Mastech MS5308 (it has a few quirks though and is not as good value).
If you choose the DER EE 5000 consider power supply and cable options before you decide on a seller as this can blow out the cost.

Hry thanks so much for your advise, yep I think the DER is going to be the best viable solution. Now thing is if I could get my hands on some detailed training manual on how to use the DER to diagnose bad caps on powe supplies like from ledtv's and so on. The DER does work for in circuit testing right?. When you said about " consider power supply and cable options before you decide on a seller " what do you mean you said this referring yourself to the DER meter? You mean there are some sellers selling only the device without the cables? that 's kind of a scam I guess.

[don.r]

No scam, some people don't need the test leads or AC power adapters. Just browse the ebay listing for de-5000 and choose the one with the options you want. The alligator clips and tweezers are true 4 wire units. The PC connection seems expensive for what its is and doubtful to be useful in repair.

[dssence]

No scam, some people don't need the test leads or AC power adapters. Just browse the ebay listing for de-5000 and choose the one with the options you want. The alligator clips and tweezers are true 4 wire units. The PC connection seems expensive for what its is and doubtful to be useful in repair.

Hi I was thinking about the DE-5000 with the alligator clips and tweezers, but at the same time are those connectors propietary ? cause I see they have a black box I don't know if there's some specific circuitry for the DE-5000 or I could buy some cheap multimeter leads and I could save some money .

[don.r]

No scam, some people don't need the test leads or AC power adapters. Just browse the ebay listing for de-5000 and choose the one with the options you want. The alligator clips and tweezers are true 4 wire units. The PC connection seems expensive for what its is and doubtful to be useful in repair.

Hi I was thinking about the DE-5000 with the alligator clips and tweezers, but at the same time are those connectors propietary ? cause I see they have a black box I don't know if there's some specific circuitry for the DE-5000 or I could buy some cheap multimeter leads and I could save some money .

Nope. Dave did a teardown on the IET units (same thing) and they are just empty boxes connecting wires. No circuitry. They are there primarily for structural rigidity I would think. If you don't need true 4w measurement to the DUT, just buy the DE-5000 for $90 and use your own test leads.

[dssence]

  Does the DE-5000 use a battery like 9v ,  or a power supply to work? I believe if I buy it without all the accesories and smd tweezer it comes with the power supply right?
 

[don.r]

  Does the DE-5000 use a battery like 9v ,  or a power supply to work? I believe if I buy it without all the accesories and smd tweezer it comes with the power supply right?

9V battery or optional external power. Here is the manual: http://akizukidenshi.com/download/ds/deree/DE-5000_manu_en2p.pdf

[BravoV]

basically I'm trying to find a good ESR meter that will help me out to diagnose failing electrolitic. Cause the bulged tops which happened commmonly was the main cause you could see physically. But when they do seem fine.. you have no other way of testing rather with an ESR meter in circuit.

My later design, ESR meter adapter, has 'best in class' protection and is very accurate.

https://www.eevblog.com/forum/projects/esr-meter-adapter-design-and-construction/msg341177/#msg341177

Have you considered building your own ESR meter?

If your target application is just to hunt down bad caps in switching power supply, pc motherboard and similar while "IN CIRCUIT", I'm re-emphasizing JDB's suggestion above.

Even though its diy, for above mentioned purposes, its "MORE" than enough and do not underestimate it's effectiveness.

As usual, the devil is in the details  , its the built-in protection that is very important, and you only need one accident that you forgot to discharge the cap, that moment within few nano seconds will instantly fry your LCR meter for good.   

[dssence]

 Thanks Don, by the way I've checked DER Website and there's lots of quality instruments. I really liked the DE-208A digital meter. But I can't happen to find an US dealer to buy from, ( I might buy..it buy I found it on aliexpress at 208 U$S dollars kind of expensive I thought would be in 100 bucks range. I own a mastech-Quail MS8221C which acts good now). I've got missing a quality analog dmm and found one from DER for 49 bucks would that be a good option? thanks
  going back to the caps testing.. I have on this power supply from my led which stopped working. Was doing a relay clikc.. then turned off.. till it didn't turn on anymore . But I checked for any bulging caps and there aren't any. I checked capcitance of all of them with my mastech meter and one of them which was 25v/1000uf gave me 1700uf way out of scale.. could that cap be bothering ?
  By the way the power plug of the psu going to the mainboard of the led tv I checked all voltages without plugging to the mainboard and all voltages are as defined

[dssence]

basically I'm trying to find a good ESR meter that will help me out to diagnose failing electrolitic. Cause the bulged tops which happened commmonly was the main cause you could see physically. But when they do seem fine.. you have no other way of testing rather with an ESR meter in circuit.

My later design, ESR meter adapter, has 'best in class' protection and is very accurate.

https://www.eevblog.com/forum/projects/esr-meter-adapter-design-and-construction/msg341177/#msg341177

Have you considered building your own ESR meter?

If you target application is just to hunt bad caps in switching power supply, pc motherboard and similar while "IN CIRCUIT", I re-emphasizing JDB's suggestion above.

Even though its diy, for above mentioned purposes, its "MORE" than enough and do not underestimate it's effectiveness.

As usual, the devil is in the details  , its the built-in protection that is very important, and you only need one accident that you forgot to discharge the cap, that moment within few nano seconds will instantly fry your LCR meter for good.   

 Well if in fact I would be doing in-circuit testing of ESR , I would first of all measure the voltage on caps and discharge them to avoid frying the ESR meter.

[mos6502]

I would definitely get both the tweezers and the croc clips. The price difference is minimal. I won't say multimeter leads are useless, but ... just by moving them around you e.g. get a change in capacitance by 10pf or more. This also makes calibrating a pain because it will often fail. If you really want to use normal leads make them as short as possible.

[dssence]

I would definitely get both the tweezers and the croc clips. The price difference is minimal. I won't say multimeter leads are useless, but ... just by moving them around you e.g. get a change in capacitance by 10pf or more. This also makes calibrating a pain because it will often fail. If you really want to use normal leads make them as short as possible.

awkay but I could well use the SMD twezzers to check electrolitics right? or they're only suitable for smd caps (non electrolytic)

[mos6502]

I would definitely get both the tweezers and the croc clips. The price difference is minimal. I won't say multimeter leads are useless, but ... just by moving them around you e.g. get a change in capacitance by 10pf or more. This also makes calibrating a pain because it will often fail. If you really want to use normal leads make them as short as possible.

awkay but I could well use the SMD twezzers to check electrolitics right? or they're only suitable for smd caps (non electrolytic)

They actually work well for testing leaded caps, too. The big advantage is that you only need one hand, so it's very quick. The opening is about 10mm. For larger caps, you can just squeeze the tweezer legs outward.

[ron]

How much input protection does the DE-5000 have?  In Dave's video he said he didn't see any.  Can that be so?

[PedroDaGr8]

How much input protection does the DE-5000 have?  In Dave's video he said he didn't see any.  Can that be so?

It's not unheard of at all. As discussed earlier, this is NOT a multimeter. it is not intended to see dangerous voltages it's for testing passive devices. In fact input protection can mess with the accuracy of fuses by adding inductance and capacitance. So most of the time expect little to no input protection.

Sent from my EVO using Tapatalk

[dssence]

How much input protection does the DE-5000 have?  In Dave's video he said he didn't see any.  Can that be so?

It's not unheard of at all. As discussed earlier, this is NOT a multimeter. it is not intended to see dangerous voltages it's for testing passive devices. In fact input protection can mess with the accuracy of fuses by adding inductance and capacitance. So most of the time expect little to no input protection.

Sent from my EVO using Tapatalk

Hi pedro so this well means that It's mandatory to discharge all caps or at least test them if they're charged before testing for any nominal value such as ESR or D or whatever with the DE-5000?.  By the way I've been checking for other LCR meters and they're way more expensive than the DE-5000 why is that? It's weird casue this instrument its made in japan and I suppose it's well quality made and tuned.

[BravoV]

Do not forget the DE-5000 test signal level is 0.5 V "RMS", definitely not good for in circuit testing.

[dssence]

Do not forget the DE-5000 test signal level is 0.5 V "RMS", definitely not good for in circuit testing.

DAMN so the DE-5000 it's not suitable for in circuit testing , I asked and ppl told me it was suitable for in circuit testing. So for in circuit testing should I go back to blue esr  ?