John S said:
I was not aware that there was a lookup table for this. Even if I
don't need it, I'd like to see it. Where can I find it?
How did you determine a D of 0.2 is the limit? Tell me more about
D and ESR, please.
Hi John,
most of the ESR meters print a table on the front panel that shows
the expected ESR for different values of caps. There are some
examples on this page:
http://members.ozemail.com.au/~bobpar/esrmeter.htm
The Capacitor Wizard doesn't have a table. Instead, it has some
vague instructions by Doug Jones, the designer of the Capacitor
Wizard. Here is a section from one of the pdf files:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Well, finding those open caps and good caps was easy. Now lets look
at other bad caps that require a little more experience with the
Capacitor Wizard and some knowledge about capacitor TYPES and USES.
You probably found caps from 1 to 30 ohms ESR in your bad box. How
do you tell the good caps from the bad??
Whether the ESR of a particular capacitor is correct or too high can
always be determined by comparing the suspicious capacitor to a
known good one of the same value, voltage rating, and type.
Unfortunately one doesn't always have another capacitor to compare
against. Experience is the best teacher here, however there are some
general guidelines:
The higher the rated working voltage, the higher the normal ESR.
Capacitors used in Power Switching applications need to have really
LOW ESR - less than 1/2 ohm Nonpolar Caps are normally less than 1/2
ohm The next logical question about ESR is "How HIGH is TOO HIGH"?
Thats a judgement call that can only be based on experience or
comparison to a known good cap (or access to the engineering data
from the capacitor or equipment manufacturer - ha ha!). Over 10 ohms
is certainly too high for most applications. Over 3 ohms is too high
for Horiz/Vert switching applications. Over 1/2 ohm is too high for
power switching applications. By comparison you will gain experience
and know when to be suspicious. These are my opinions. Here are some
actual repair situations:
Example: 47uf @50vdc measures 25 ohms ESR in circuit - BAD CAP The
suspect capacitor is a 47uf @50vdc in a switching power supply for a
VCR. The Capacitor Wizard has measured 25 ohms ESR in circuit. That
is higher than 15 ohms and much to high for any quality cap. A new
capacitor measured 5 ohms ESR. The new capacitor fixed the VCR. In
my opinion the new capacitor was not of the highest quality (5 ohms
is too high) however it did fix the VCR. The use of these low
quality inexpensive import capacitors is probably the reason we see
so much capacitor failure in con- sumer electronic equipment! A
higher quality cap with a lower ESR of the same kind costs more
money but will measure less than 1 ohm and be more reliable.
Conclusion: This is a higher voltage capacitor and can be expected
to normally measure higher than 1/2 ohm. In my judgment any "switch
mode" capacitor that measures more than 3 ohms ESR is suspect no
matter what the voltage rating.
However you may obviously get by with the 5 ohms ESR in that
particular circuit. For comparison, the bad part was checked "out of
circuit" on a well known competitors $2000 Cap analyzer and it
determined that the cap was GOOD - even though the ESR measured 25
ohms! That manufacturer made a huge mistake by trying to calculate
good and bad ESR from entered and measured data. It can't be done
reliably. That is why we don't simply have a good/bad indication on
our meter scale. Any cap over 3 ohms is suspect. This is my
Experience.
Example: 1000uf @6vdc measures 1.5 ohms in circuit - BAD CAP This is
a little brown 1000uf 6vdc cap used in lots of VCR switching power
supplies. The Capacitor Wizard measured 1.5 ohms in circuit. Because
the capacitors operating voltage is so low (6vdc) and its used in a
switching power supply, I would expect a normal ESR reading of less
than 1/2 ohm. Comparison to a known good cap confirmed it should
measure less than 1/2 ohm. Replacing this cap cured the trouble.
This particular cap goes bad often as I have many in my box of bad
caps gathered from local repair compa- nies. If you work on VCRs, I
bet you have some too.
Summery: (mrm: sp)
Measuring ESR is a very good indicator of capacitor failure. For
switch mode circuits it is the ONLY reliable capacitor test, IN or
OUT of circuit!. Open caps and caps with really high ESR (over 10
ohms) are easy to find in circuit and need to be replaced.
Marginal caps that measure between 1 and 10 ohms ESR require some
experience with the Capacitor Wizard and/or comparison to a known
good cap of the same voltage, value, and type. Caps above an
operating voltage of 35vdc have a normally higher ESR (around 1 to 3
ohms) than caps of a lower voltage (less than 1/2 ohm ESR).
I know of no perfect formula or rule that can always tell normal ESR
from marginal ESR other than comparison to a known good part. The
obvious solution is to obtain the capacitor manufacturers data
manuals on the EXACT capacitor measured but that is not normally
practical. As a technician myself I always follow this rule: "If in
doubt, replace". You will eliminate a lot of recalls and cure many
weird and undefinable intermittent problems if you follow this rule.
Doug Jones, Designer of the Capacitor Wizard
http://midwestdevices.com/_pdfs/FirstTime.pdf
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To find a typical dissipation factor for the Tonghui, I went through
several boxes of old electrolytic caps. Regardless of the capacitor
value, there seemed to be a clear dividing line between good and bad
caps. I settled on D = 0.2 as it seemed to be a reasonable number
for most applications, such a bypass and coupling caps.
However, if the application was critical, such as a capacitance
multiplier for low level dc supply, I'd look for capacitors with the
lowest D value I could find, and put some in parallel.
The interesting thing, and this has me a bit confused, is the
dissipation factor does not seem to be affected much by the test
frequency.
I would expect the ESR to remain fairly constant with frequency, but
the capacitive reactance of course will change. So the impedance of
the capacitor will change with frequency, and I would expect the
dissipation factor to change also.
I haven't had time to sort this out yet, and I need to get some more
experience with this instrument and find out how it is making the
measurement. For example, take a known good capacitor and add some
series resistance and see what happens to the readings. But here are
some wikipedia references to start with:
Equivalent series resistance
http://en.wikipedia.org/wiki/Equivalent_series_resistance
Dissipation factor
http://en.wikipedia.org/wiki/Dissipation_factor
Mike
