Insignia NS-27LCD

[PCBONEZ]

Re: Insignia NS-27LCD

If you understood what you were reading that also enforces my argument.
.
That says is Dielectric loss changes with frequency.
That's why it is shown as a variable Resistor [Rd] in Figure 4.
Dielectric loss is part of ESR.
.
So again, your very own reference is telling you you're wrong.
.
ESR changes with Frequency. - Get over it.
.

[Paul678]

Re: Insignia NS-27LCD

I feel like a genius next to an idiot such as yourself.

I was referring to Ras in Figure. 4.

You need to study up some more....

[PCBONEZ]

Re: Insignia NS-27LCD

LOL! I like your engineering vocabulary!!! "get stuck", "clog", "hang up"!

Hilarious! Pathetic attempt to sound intelligent.

No, that's an attempt to not make you feel so stupid.
.
Well, that and it makes for a less boring read.
.
.

Here's a real explanation:

And you misinterpretation of that explanation is why you should feel stupid.

The Rs shown in Figure 1 is NOT a constant with frequency.
Never had been, never will be.
- I just explained why.

And in fact YOUR reference even specifies that it's only good at one Frequency.

We now have an equivalent circuit that is correct only at the measurement frequency.

So, even your very own reference is telling you you're wrong.
.

[Paul678]

Re: Insignia NS-27LCD

LOL! I like your engineering vocabulary!!! "get stuck", "clog", "hang up"!

Hilarious! Pathetic attempt to sound intelligent.

Here's a real explanation:

I didn't ~assume~ anything.
That's exactly what you suggested doing when you asked how to check DC Leakage with an ESR meter.
So you tell me: Why would you try to check DC Leakage with an AC signal?

You did make a stupid assumption. If you read my post, I said you
CANNOT make a DC test with that AC meter. Obviously it's impossible.

[PCBONEZ]

Re: Insignia NS-27LCD

The TRUE ESR, or pure resistance should not change with frequency.

Bullshit.
-
Electrolytic caps have chemical and mechanical properties at the molecular level that are greatly affected by frequency.
Frequency not only affects ESR, it affects it several ways.
The conductivity of the electrolyte is a component of ESR and it is affected by frequency.
The oxide layer's properties are also affected by frequency and that's another component of ESR.
These things amount to Frequency controlled variable Resistors.
-
With Lytics the OXIDE LAYER is the -actual- dielectric.
The oxide layer in Low ESR Lytic caps actually consists of tiny oxidized tubes.
The etching method [for this type cap] actually makes deep pits, like a holes drilled into a board.
When the oxide layer is formed the inside surfaces of the pits are also oxidized.
The whole reason for this is to increase surface area of the plate.
-
The liquid in the electrolyte is just an extension of one of the plates.
It's a conductor used to bridge the distance from one plate to the oxide layer on the other plate.
It 'form fits' the oxide layer by filling the tubes.
In effect the 'form fit' increases the surface area of the other plate.
-
To get AC current from lead to lead, electrons [and the ions they 'ride' on] have to travel back and forth through these tubes with the AC's constant polarity reversal.
If 'the charges' aren't in the tubes they have much less effect [due to distance] on electrons on the other side. [And other lead.]
- At lower frequencies the electrons aren't moving fast enough for their momentum to prevent [figuratively] log-jams.
They 'get stuck' or 'hang' or 'clog up' in the tubes and that results in higher ESR because less movement is 'felt' on the other side of the dielectric. [Effectively the other lead 'sees' less AC current.].
- At high frequencies [nearing GHz] they are moving so fast their movement isn't 'felt' through the thickness of the dielectric.
That's because the 'pull' isn't present long enough to budge electrons on the other side into moving. This also makes ESR go higher.
- The happy place in the middle is right around 100 kHz. [Go figger!]
Thus the nature of the dielectric raises [it's part of] ESR at frequencies both far above and just below the resonant frequency.
-
Inside the liquid electrolyte [the conductor], electrons move primarily by being carried and exchanged by traveling ions.
Because the ions have a charge, their movement is also affected by frequency.
- At lower frequencies 'the charges' move less distance per unit of time. That means fewer ions smack into each other [exchanging or freeing electrons] and that means a less conductive solution. That in turn means higher ESR.
- At higher frequencies they travel further [per unit time] so more of them smack into each other and that means a more conductive solution, and lower ESR.
[The values for this will vary widely with the type of electrolyte used.]
-
When electrons and ions smack the actual oxide it knocks 'chunks' off.
With only AC [or insufficient DC] applied the oxide will progressively become thinner [lowering ESR for AC and increasing Leakage for DC] and the frequency absolutely affects the rate at which that happens.
Additionally the 'fresh chunks' in solution dissolve into ions increasing the ion inventory which increases the conductivity.
[The values for this will vary widely with the type of electrolyte used.]
-
After we've thrown all that [and the other things I'm not going into] in the blender and pour it out on a plot that covers from 0Hz to the GHz range we end up with a curve roughly shaped like a bathtub. The drain is near 100kHz.
And whether you like it or not, it's "True" ESR because those are all "True" components of EQUIVALENT SR.
.
.

But the measurement goes up when the testing frequency is lower than the resonant frequency of the cap, and
what you see in the capacitive reactance kick in.

That's more Hog Wash.
-
To start with Xc [capacitive reactance] is not a component of ESR. They change independently of each other.
-
The microprocessors in ESR meters have no problem compensating for Xc. - Or ESL for that matter.
If nothing else that's easy to do in the firmware, particularly as the meters use predetermined frequencies.
-
And your contention that the observed change is a measurement error is complete BUNK.
If you can find it, manufacturer data shows actual ESR as the bathtub curve I described earlier with the low point at about 100kHz.
Lower than ~50kHz it rises exponentially as you approach zero Hz [aka: DC].
Higher than ~500kHz it rises slowly until much higher frequencies where it again turns into a near exponential rise.
Few publicly released tech docs even show what ESR does below roughly 1kHz or above roughly 10MHz.
It doesn't come up much even on manufacturer's sites because that's NOT what Low ESR Lytics are FOR anyway.
You CAN get your hands on that data but you'll probably have to email manufacturers to get it.
.
.

And I do build transmitters, so I DO care about ESR of pF caps.

Good for you.
I cut my teeth on Radars. They bored me to death and I don't remember much about them anymore.
This site is all about Low ESR Electrolytic caps used in IT gear and SMPS like in the screen this thread is about.
They don't make such caps in pF so not only are they off-topic for this thread they are off-topic in general.
In depth discussion about parts that aren't used in screens anyway doesn't help fix this screen.
If you want to talk about radios and pF caps go open your own thread. This one is a TS&R thread.
I'm sure at least a few members will look. We got radio buffs here too.
And if you need further remedial training on Low ESR Lytics a new thread for that wouldn't hurt either.
[Just FYI: I was dumber than dirt about ESR when I first found this site. Most people are.]
.
.

It would be stupid to assume I meant that! Why would you use AC to test DC???

I didn't ~assume~ anything.
That's exactly what you suggested doing when you asked how to check DC Leakage with an ESR meter.
So you tell me: Why would you try to check DC Leakage with an AC signal?
.
.
First pic snipped from a Nichicon Data Sheet.
Second pic shows ESR and Z for three types of lytics [Changed series names to their type names for clarity.]

[Paul678]

Re: Insignia NS-27LCD

No, you are the confused one.

The TRUE ESR, or pure resistance should not change with
frequency. But the measurement goes up when the testing
frequency is lower than the resonant frequency of the cap, and
what you see in the capacitive reactance kick in.

And I do build transmitters, so I DO care about
ESR of pF caps.

You said,

ESR meters work perfectly fine for DC resistance within their range..
They are in fact calibrated with resistors which, unlike caps, don't change with frequency.
There is no need for a 20Meg setting on an ESR meter. 100 ohms ESR would be a VERY bad cap.
- And, as the test voltage is AC it would be STUPID to use one for a DC leakage test.

It would be stupid to assume I meant that! Why would you
use AC to test DC???


Quote:

100kHz was chosen for two reasons.
- It's close to the Ripple frequency used in most SMPS. [Which was probably in part chosen for the next reason.]
- It's near the lowest point for ESR where ESR [effectively] = Z [ya, near the self-resonant frequency were ESL and Xc cancel each other out] for caps typically used in SMPS.
This is why Z and ESR are used interchangeably in data sheets. At 100kHz they are the same.
~ Self-resonant frequency doesn't vary that much in the kind of caps and applications the spec was added to data sheets FOR and it's at or very near 100kHz.
THAT is why 100kHz became the Standard used in data sheets. And yes, it's actually an Industry Standard.
~ No, it's not perfect for every cap, but it's a heck of a lot better than having a different frequency for every cap. Ya Think?

Non low ESR caps are typically rated at 100Hz or 120Hz, also because of their design application. In that case it is 2x line frequency which is found at the OP of the first rectifier. That one is an older Standard that dates back to linear PSUs.

We agree on this point. The resonant frequency is where the
capacitive and inductive reactances cancel each other out (Xc = XL),
in theory leaving you will just the pure resistance of the cap, or
the ESR. Agreed it's better to use the same testing frequency,
even though it may not be the exact resonant frequency of the cap.

Yes, I've seen datasheets that use 100Hz. Rectifiers will have
2nd harmonic of 60Hz, sure.

[PCBONEZ]

Re: Insignia NS-27LCD

You are confused.

True ESR is the purely real portion of the impedance (pure resistance),
and should ideally not change with frequency, so that's the measurement error of the technique.

ABSURD!.
If this were even close to true then at the lower frequencies of typical load variations the caps would be nearly a dead short to ground and never charge at all.
- The curve of ESR to Frequency is more or less a bathtub curve.
For the range of caps used in IT equipment [the ones where -we- care about ESR] the low point is at about 100kHz.
Generally there is a flat spot at the low point and for the caps -we- care about it nearly always crosses to 100kHz line.
- We don't give a rats patootie about the ESR of pF caps.
We aren't building transmitters, we are filtering out Ripple [Switching Noise] from SMPS and MOSFETs etc...
That noise is typically in the range of 50kHz to 400kHz and the majority of the time it's near 100kHz.

Plain, that looks like an interesting box, but it doesn't look like
it measures DC resistance, or a leaky capacitor test. Although can't you
just use your ohmmeter in a high resistance setting to check
for leakage, and just allow the meter to charge the cap, and
make sure it's an open (over scale on 20Meg setting)?

ESR meters work perfectly fine for DC resistance within their range..
They are in fact calibrated with resistors which, unlike caps, don't change with frequency.
There is no need for a 20Meg setting on an ESR meter. 100 ohms ESR would be a VERY bad cap.
- And, as the test voltage is AC it would be STUPID to use one for a DC leakage test.

Ideally, you'd test at the self-resonant frequency of the cap, but since
this varies so widely, the industry probably just chose 100kHz for the
larger caps.

100kHz was chosen for two reasons.
- It's close to the Ripple frequency used in most SMPS. [Which was probably in part chosen for the next reason.]
- It's near the lowest point for ESR where ESR [effectively] = Z [ya, near the self-resonant frequency were ESL and Xc cancel each other out] for caps typically used in SMPS.
This is why Z and ESR are used interchangeably in data sheets. At 100kHz they are the same.
~ Self-resonant frequency doesn't vary that much in the kind of caps and applications the spec was added to data sheets FOR and it's at or very near 100kHz.
THAT is why 100kHz became the Standard used in data sheets. And yes, it's actually an Industry Standard.
~ No, it's not perfect for every cap, but it's a heck of a lot better than having a different frequency for every cap. Ya Think?
-
Non low ESR caps are typically rated at 100Hz or 120Hz, also because of their design application. In that case it is 2x line frequency which is found at the OP of the first rectifier. That one is an older Standard that dates back to linear PSUs.
.

[Paul678]

Re: Insignia NS-27LCD

Either 1000 or 2000 Ohms of DC resistance would be bad
for a cap. It ideally should be a complete open at DC, or
infinite Ohms.

How do you guys usually test for leaky capacitors? By leaky,
I mean caps that fail to provide DC blocking.

[ggil44]

Re: Insignia NS-27LCD

Hi Paul678

Yes, that is the capanalyser I have. Bought it on Ebay for $120. PlainBill is correct when he says "Any piece of test equipment is only as good as the person holding the leads." I am not that good but it has identified some bad caps that did not have any physical signs of trouble. You can research their page but it does NOT measure capacitance.

[PlainBill]

Re: Insignia NS-27LCD

Plain, that looks like an interesting box, but it doesn't look like
it measures DC resistance, or a leaky capacitor test. Although can't you
just use your ohmmeter in a high resistance setting to check
for leakage, and just allow the meter to charge the cap, and
make sure it's an open (over scale on 20Meg setting)?

True ESR is the purely real portion of the impedance (pure resistance),
and should ideally not change with frequency, so that's the measurement error of the technique. But the higher the test frequency, the less you charge
the cap, and the less the capacitive reactance. Too high a test frequency
would also be bad, because then the parasitic inductance of the cap starts to kick in.

Ideally, you'd test at the self-resonant frequency of the cap, but since
this varies so widely, the industry probably just chose 100kHz for the
larger caps.

For very small pF caps, you have to increase the test frequency for ESR. These are measured around 1GHz:

DC resistance is a meaningless parameter for power supply caps. Do you REALLY care if a 1000uF, 25 volt cap has a resistance of 1000 ohms or 2000 ohms?

Lets look at some numbers. Typically they use 25 Volt caps in the 12 Volt (inverter) supply of a small monitor. The inverter draws in the 1-2 am range. Even a 100 ohm DC resistance would draw only 125 mA - not a serious additional load. However, that power dissipation (1.5 watts) would probably cook the capacitor, so it would fail ESR and capacitance tests.

The resistance itself wouldn't effect the ESR test; even a 10 ohm resistance wouldn't significantly effect the ESR test, and the cap would glow like a bulb on a Christmas tree.

PlainBill

[Paul678]

Re: Insignia NS-27LCD

Plain, that looks like an interesting box, but it doesn't look like
it measures DC resistance, or a leaky capacitor test. Although can't you
just use your ohmmeter in a high resistance setting to check
for leakage, and just allow the meter to charge the cap, and
make sure it's an open (over scale on 20Meg setting)?

True ESR is the purely real portion of the impedance (pure resistance),
and should ideally not change with frequency, so that's the measurement error of the technique. But the higher the test frequency, the less you charge
the cap, and the less the capacitive reactance. Too high a test frequency
would also be bad, because then the parasitic inductance of the cap starts to kick in.

Ideally, you'd test at the self-resonant frequency of the cap, but since
this varies so widely, the industry probably just chose 100kHz for the
larger caps.

For very small pF caps, you have to increase the test frequency for ESR. These are measured around 1GHz:

[PCBONEZ]

Re: Insignia NS-27LCD

Yes that meter is junk for low ESR applications.
Need the range of 0.01 to 1.00 ohms to be useful for that.
.
It also says it uses a test frequency -over- 100kHz, which is stupid.
The specs for low ESR caps are given -at- 100kHz in the data sheets.
... And ESR changes with frequency.
.

[PlainBill]

Re: Insignia NS-27LCD

Ggil, What capanalyzer do you have?

Is it this one?

http://www.eds-inc.com/cap.html


And a question for everyone here: I've never owned
one of these before, but the prospect of not having
to desolder caps to measure them is attractive.

But what if you have more than one cap in parallel?
Won't you measure the ESRs in parallel, and won't
that screw up the measurement? Also, most
filtering caps are electrolytic, so you pretty much are never
going to have a resistor in parallel with the cap, or
that would screw up your DC resistance measurement, right?

And this meter only works on electrolytics, right?

Any reviews/advice on this device is appreciated.....

Thanks....

Overpriced junk, in my opinion. The last good product to come out of Boca Raton was the IBM PC. Since then it's been saturated with 'enhanced' bimbos and con artists.

For about 1/3 the price you can buy an ESR-Micro v4.0 or the Bob Parker Blue ESR meter that actually give you numbers.

Any piece of test equipment is only as good as the person holding the leads. You are correct about the effects of parallel caps, although that is less of a problem than you would expect. A typical low ESR cap will have an ESR of .05 ohms or less; often the circuit will still work properly if the ESR is 10 times greater. It's not uncommon to find Capxon caps with an ESR 100 times greater than specs. 'When in doubt, rip them out.'

PlainBill

[Paul678]

Re: Insignia NS-27LCD

Ggil, What capanalyzer do you have?

Is it this one?

http://www.eds-inc.com/cap.html


And a question for everyone here: I've never owned
one of these before, but the prospect of not having
to desolder caps to measure them is attractive.

But what if you have more than one cap in parallel?
Won't you measure the ESRs in parallel, and won't
that screw up the measurement? Also, most
filtering caps are electrolytic, so you pretty much are never
going to have a resistor in parallel with the cap, or
that would screw up your DC resistance measurement, right?

And this meter only works on electrolytics, right?

Any reviews/advice on this device is appreciated.....

Thanks....

[ggil44]

Re: Insignia NS-27LCD

I bow to your superior knowledge. You saved me from stumbling into an additional problem.

Thanks,

Gil

[PlainBill]

Re: Insignia NS-27LCD

PlainBill

I tried typing in the Cap numbers into a company that sells caps and the 10uf came out as you stated. I then tried the "47uf" code (47 16V VS) and it came out "Panasonic Aluminum Electrolytic Capacitors - SMD 470UF 16V" Is there a multiplyer that I missed? Is their assesment of the numbers incorrect? I don't have a schematic so I can't tell what it should be for sure. Don't want to stick the wrong one in there.

Regards,

I just tried entering 47 UF 16V vs on Digikey's website and cape up with exactly what I was expecting - Digikey part number PCE3181CT-ND

PlainBill

[ggil44]

Re: Insignia NS-27LCD

PlainBill

I tried typing in the Cap numbers into a company that sells caps and the 10uf came out as you stated. I then tried the "47uf" code (47 16V VS) and it came out "Panasonic Aluminum Electrolytic Capacitors - SMD 470UF 16V" Is there a multiplyer that I missed? Is their assesment of the numbers incorrect? I don't have a schematic so I can't tell what it should be for sure. Don't want to stick the wrong one in there.

Regards,

[ggil44]

Re: Insignia NS-27LCD

Ignorance is a pain, and at times embarrasing. Couldn't find anything on the web that was a key to deciphering the manufacturer code.

[PlainBill]

Re: Insignia NS-27LCD

Thank you for the help Bill. It seems if they (surface mount cap makers) have room for 3 lines of information to print on the cap that they could just print the value, voltage, and tolarance in a standardized mannor so everyone could read it.... OK, I'll get off my soapbox now.

Thanks again,

ggil44

That IS a standardized pattern for SMT electrolytics.

PlainBill

[ggil44]

Re: Insignia NS-27LCD

Oh, in response to the panel, I don't think it is a panel problem because when the tv is showing a no input message box the vertical lines move across the screen in a size of the information box with the rest of the screen clear.
On a channel with "snow", it is 2 or 3 broken vertical lines end to end and they are side by side filling the whole screen but dancing around like the "snow" would normally be on the screen. IF the screen has no signal and it is muted, no "snow" then the screen is clear, no lines. I could take pictures but I wanted to try these capacitors first to see if it possibly cleaned it up. I am hoping it is not needing a new T-CON board.

Thanks,

Gil