Why electrolytics back to back?
Jeff Anderson
A common technique when using an electrolytic cap in a circuit where
the polarity of the DC voltage is unknown is to connect two
back-to-back. I've used this technique for years, and so has every
other engineer that I've worked with, but I've no idea *why* it works.
Does anyone have a good explanation?
Thanks,
- Jeff
David Gibson
Correctly connected an electro blocks dc. Reverse connected it will
conduct dc. Almost like a diode with a lot of shunt capacitance.
Connect two in series-opposing and it will act a non-polar cap with
one being the cap on one half cycle, the other on the other half
cycle.
I don't know how far the diode analogy can be taken, i.e. it implies a
non-linear transfer function which might distort low level signals.
David
PAUL WU -- NEW YEAR IS COMMING.
>
> A common technique when using an electrolytic cap in a circuit where
> the polarity of the DC voltage is unknown is to connect two
> back-to-back. I've used this technique for years, and so has every
> other engineer that I've worked with, but I've no idea *why* it works.
> Does anyone have a good explanation?
Jeff,
This method is to create a non-polarised electrolytic cap. ie. one
cap will take the positive cycle, and one takes the negative one.
Cheers,
Paul
*******************************
* An undergrad in NTU,Singapore *
* A personal opinion only *
*********************************
Jarek Lis
: >A common technique when using an electrolytic cap in a circuit where
: >the polarity of the DC voltage is unknown is to connect two
: >back-to-back. I've used this technique for years, and so has every
: >other engineer that I've worked with, but I've no idea *why* it works.
: >Does anyone have a good explanation?
: Correctly connected an electro blocks dc. Reverse connected it will
: conduct dc. Almost like a diode with a lot of shunt capacitance.
: Connect two in series-opposing and it will act a non-polar cap with
: one being the cap on one half cycle, the other on the other half
: cycle.
I'd rather say, that with this setup one-way leakage current will charge
both capacitor to some 'bias' voltage. So each of them are correctly
polarised, but those 'initial charges' subtract each other, and totally
you don't see it.
My hypotese should be easy to confirm - just measure voltage on
common point with meter of great input impendance!
Jaroslaw Lis
+------------------------------------------------------------------------+
| l...@ict.pwr.wroc.pl | Institute of Engineering Cybernetics |
| tel 48-71-202636 | Technical University of Wroclaw, Poland |
| fax 48-71-203408 or 517398 | |
+------------------------------------------------------------------------+
john 015
<TJC...@SPERRY-SUN.COM> said:
>CC01...@brownvm.brown.edu (john 015) wrote:
>>In article <4b94tm$g...@ixnews2.ix.netcom.com>, je...@ix.netcom.com(Jeff
>>Anderson) said:
>>>A common technique when using an electrolytic cap in a circuit where
>>>the polarity of the DC voltage is unknown is to connect two
>>>back-to-back. I've used this technique for years, and so has every
>>>other engineer that I've worked with, but I've no idea *why* it works.
>>>Does anyone have a good explanation?
>>
>>The capacitor in reverse bias ( > 15% reverse voltage) leaks charge.
>>The leakeage is limited by the presence of the other capacitor
>>(it is increasingly forward biased by the charge that leaks)
>>to a safe value (no permanent dielectric break-down) assuming
>>they are of similar construction and comparable size.
>>
>
>I'm not sure that this is the full story since when you connect two
>electrolytics back to back you don't get as much capacitance as with one
>of them on its own, but you get more than half, which is what you would
>normally expect from two equal capacitors in series. Hence it appears as
>though they are acting as two capacitors in series to a degree, but not
>fully.
I think I could explain why a
measuring device will show less capacitance than then one
but more than 1/2 but I am headed for vacation.
If you are concerned, tie the midpoint to a DC voltage
outside the range of the combined AC/DC on the caps via
a high valued resistor and voila, both caps are properly
biased.
john
Terry Clark
>In article <4b94tm$g...@ixnews2.ix.netcom.com>, je...@ix.netcom.com(Jeff
>Anderson) said:
>
>>A common technique when using an electrolytic cap in a circuit where
>>the polarity of the DC voltage is unknown is to connect two
>>back-to-back. I've used this technique for years, and so has every
>>other engineer that I've worked with, but I've no idea *why* it works.
>>Does anyone have a good explanation?
>
>The capacitor in reverse bias ( > 15% reverse voltage) leaks charge.
>The leakeage is limited by the presence of the other capacitor
>(it is increasingly forward biased by the charge that leaks)
>to a safe value (no permanent dielectric break-down) assuming
>they are of similar construction and comparable size.
>
I'm not sure that this is the full story since when you connect two
electrolytics back to back you don't get as much capacitance as with one
of them on its own, but you get more than half, which is what you would
normally expect from two equal capacitors in series. Hence it appears as
though they are acting as two capacitors in series to a degree, but not
fully.
I have used this scheme myself without problems, although I'm never very
happy about it - it seems to me that one of the capacitors will always be
reverse biased at any one time, thereby tending to reverse the process
which formed the aluminium oxide dielectric on the plates. Perhaps
someone more knowledgable in capacitor chemistry can shed some light on
this?
Terry Clark
Erik Blake
>From: l...@okapi.ict.pwr.wroc.pl (Jarek Lis)
>Subject: Re: Why electrolytics back to back?
>Date: 21 Dec 1995 20:01:47 GMT
>David Gibson (dgi...@ocean.com.au) wrote:
>: >A common technique when using an electrolytic cap in a circuit where
>: >the polarity of the DC voltage is unknown is to connect two
>: >back-to-back. I've used this technique for years, and so has every
>: >other engineer that I've worked with, but I've no idea *why* it works.
>: >Does anyone have a good explanation?
>:
>: Correctly connected an electro blocks dc. Reverse connected it will
>: conduct dc. Almost like a diode with a lot of shunt capacitance.
Hmm, seems to me that if the "active" cap is charging/discharging via leakage
through the other cap, then you might get pretty slow response - i.e. the
reverse-biased cap acts like a resistor (or is it like a diode). Comments?
Erik
--------------------------------------------------------------------------
Icefield Instruments Inc. tel: (403) 633-4264
P.O. Box 5567 fax: (403) 633-4217
Whitehorse, Yukon CANADA Y1A 5H4 e-mail: icef...@yknet.yk.ca
T. L. Fu
> A common technique when using an electrolytic cap in a circuit where
> the polarity of the DC voltage is unknown is to connect two
> back-to-back. I've used this technique for years, and so has every
> other engineer that I've worked with, but I've no idea *why* it works.
> Does anyone have a good explanation?
> Thanks,
> - Jeff
I seldom use this technique, but I know why this works. An E-cap ex-
posed to voltage with reversed polarity will have high leakage
current, which means power dissipation and eventually the destruction
of the cap. With the back-to-back configuration, there is always a
cap with correct polarity to stop the leakage current.
Regards
-TL
David Mulligan
Charles W. Hubbard
> In article <4b94tm$g...@ixnews2.ix.netcom.com>, je...@ix.netcom.com(Jeff Anderson) writes:
> >
> > A common technique when using an electrolytic cap in a circuit where
> > the polarity of the DC voltage is unknown is to connect two
> > back-to-back. I've used this technique for years, and so has every
> > other engineer that I've worked with, but I've no idea *why* it works.
> > Does anyone have a good explanation?
> Jeff,
> This method is to create a non-polarised electrolytic cap. ie. one
> cap will take the positive cycle, and one takes the negative one.
As long as we are on the subject, could somebody explain to me what it
is about a polarized capacitor that makes it polarized? I got the
impression from another post in this thread that when reversed you get
sort of a "electroplating in reverse" effect which tends to disolve
the insulating layer. Is this true? If so, why doesn't the reversed
cap in a "back-to-back" situation as described not still degrade?
Nobody has ever explained this to me before. This seemed as good a
time to ask as any.
Thanks,
C.
==========================================================================
Charlie Hubbard | As the great philosopher Bingo once
chub...@oneworld.owt.com | said, "I have seen the future and
http://www.owt.com/users/chubbard | Java is its name-o."
==========================================================================
Wayne Scharf
What happens is that an electrolytic tends to have a high leakage when
hooked up backwards, but less so in the correct direction...so the
floating junction between the two caps would eventually reach a "clamped
negative" voltage approximating, I believe, a negative voltage somewhere
in the range of the applied peak to peak...(I think that's how it goes..)
in any event, they stabilize at a point where for all practical purposes
they can both operate correctly without spending a lot of time reverse
biased. The effective capacitance is however [c1*c2]/[c1+c2].
--
NAT...@KAIWAN.COM
Silverado
Kevin McMurtrie
W. Hubbard) wrote:
>bh72...@omega.ntu.ac.sg (PAUL WU -- NEW YEAR IS COMMING.) wrote:
>
>> In article <4b94tm$g...@ixnews2.ix.netcom.com>, je...@ix.netcom.com(Jeff
Anderson) writes:
>> >
>> > A common technique when using an electrolytic cap in a circuit where
>> > the polarity of the DC voltage is unknown is to connect two
>> > back-to-back. I've used this technique for years, and so has every
>> > other engineer that I've worked with, but I've no idea *why* it works.
>> > Does anyone have a good explanation?
>
>> Jeff,
>> This method is to create a non-polarised electrolytic cap. ie. one
>> cap will take the positive cycle, and one takes the negative one.
>
>As long as we are on the subject, could somebody explain to me what it
>is about a polarized capacitor that makes it polarized? I got the
>impression from another post in this thread that when reversed you get
>sort of a "electroplating in reverse" effect which tends to disolve
>the insulating layer. Is this true? If so, why doesn't the reversed
>cap in a "back-to-back" situation as described not still degrade?
>
>Nobody has ever explained this to me before. This seemed as good a
>time to ask as any.
>
>Thanks,
Electolytic capacitors use a thin oxide layer on a metal sheet as a
dielectric. Their large capacity comes from the extreme thinness of this
dielectric. The conductive electrolyte is a means of bringing the other
electrical pole right to the surface of the dielectric layer without
scraching it.
If you reverse the power, flaws in the oxide layer will allow DC to flow
and the oxide layer will corrode off. The corrosion destroys the
capacitor and will cause it to explode from built up gasses.
Caps put back-to-back ( ----|(---)|---- ) will each bias themselves
to always have a positive voltage before any significant damage is done.
Richard Steven Walz
Charles W. Hubbard <chub...@oneworld.owt.com> wrote:
>bh72...@omega.ntu.ac.sg (PAUL WU -- NEW YEAR IS COMMING.) wrote:
>
>> In article <4b94tm$g...@ixnews2.ix.netcom.com>, je...@ix.netcom.com(Jeff Anderson) writes:
>> >
>> > A common technique when using an electrolytic cap in a circuit where
>> > the polarity of the DC voltage is unknown is to connect two
>> > back-to-back. I've used this technique for years, and so has every
>> > other engineer that I've worked with, but I've no idea *why* it works.
>> > Does anyone have a good explanation?
>
>> Jeff,
>> This method is to create a non-polarised electrolytic cap. ie. one
>> cap will take the positive cycle, and one takes the negative one.
>
>As long as we are on the subject, could somebody explain to me what it
>is about a polarized capacitor that makes it polarized? I got the
Electrolytics are a cheap shit way to make a small cap have a BIG
capacitance, but the rub IS that it only has it in ONE direction of
charge! An ideal cap is a non-polarized cap. Sticking them back to
back is a cheap-shit method of trying to cheat the electrolytic
property which doesn't work really very well, or for very long. It is
not good design, actually, because the electrolytics go to hell after
a while, as they might if used in an inverse voltage.
--
-Steve Walz rst...@armory.com ftp.armory.com/pub/user/rstevew
also mirrored at: ieee.cas.uc.edu/pub/electronics/faq/ftp.armory.com
and in Europe at: ftp.cised.unina.it/pub/electronics/ftp.armory.com
Michael A. Wyatt
writes:
Steve,
By placing diodes across each capacitor in the proper direction
(diode conducts when the capacitor's applied voltage is reversed from
the correct polarity), the reverse voltage across each capacitor is
limited to below one volt and the electrolytic capacitor's life is
extended to it's nora\mal expected life.
Hope this helps,
Mike Wyatt
Jurek
Is there any difference between back-to-back and head-to-head
connection? Probably not, but I never tried it myself.
Jurek
John Lundgren
Not if it's done as below. The capacitance is 5 uF, and both caps have a
DC voltage across them all the time, assuming that the circuits on the
input and output are running off a positive supply. The 1M lets the caps
charge up.
+ 10 uF 10 uF +
in O------|(-------+--------)|------O out
|
\
/ 1Meg
\
/
|
---
GND
--
#==========H==A==P==P==Y=====H==O==L==I==D==A==Y==S==!==!===========#
| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
| Rancho Santiago Community College | .deltanet.com |
| 17th St @ Bristol \ Santa Ana, CA 92706 | http://rsc.rancho|
| My opinions are my own, and not my employer's. | .cc.ca.us |
| "You can flame your brains out -- it won't take long." |
| I have gone out to look for myself.. If I should |
| return before I get back, hold me until I get here. |
#======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#
John Lundgren
: In <4bm8ah$p...@news.scruz.net> rst...@armory.com (Richard Steven Walz)
: writes:
: >
: Steve,
: By placing diodes across each capacitor in the proper direction
: (diode conducts when the capacitor's applied voltage is reversed from
: the correct polarity), the reverse voltage across each capacitor is
: limited to below one volt and the electrolytic capacitor's life is
: extended to it's nora\mal expected life.
: Hope this helps,
: Mike Wyatt
I've never, ever seen that done. Probably because the diodes introduce
nonlinearity into the path. And the diodes are more expensive and more
complex than a simple 1 meg resistor from the two caps to ground, as I
said in the other followup.
Marc Andelman
aluminum or tantulum metal,
with a non conductive oxide, and an electroytlyric solution.
The cathode would have to be an equal surface are material
such as a platimum black coated metal, or etched aluminum. The
reason these are polar is that the oxide forms only in one
direction. In the other direction it breaks down electrochemically,
causing leakage current to flow.
Marc Andelman
Marc Andelman
Michael A. Wyatt
John,
I have used this technique a few times. I can't claim I came up
with the idea, as I saw it in a design about 25 years ago. The diodes
will intoduce some non-linearity, but probabily not much more than
without them. If back to back electrolytics (diodes or no diodes) are
investigated from a linearity standpoint, you will discover they are
highly non-linear and should be utilized only in low precision circuits
as I mentioned in an earlier post. In fact, all electrolytic capacitors
are very non-linear due to the large capacitance change vs applied
voltage.High dielectric capacitors fall into this catagory also.
Arsenio Novo
Re: Why electrolytics back to back? (25 Dec 95 08:18:09)
rs>
rs> In article <4bg92m$k...@krel.iea.com>,
rs> Charles W. Hubbard <chub...@oneworld.owt.com> wrote:
>bh72...@omega.ntu.ac.sg (PAUL WU -- NEW YEAR IS COMMING.) wrote:
>
>> In article <4b94tm$g...@ixnews2.ix.netcom.com>,
je...@ix.netcom.com(Jeff
rs> Anderson) writes: >> >
>> > A common technique when using an electrolytic cap in a circuit where
>> > the polarity of the DC voltage is unknown is to connect two
>> > back-to-back. >
>> Jeff,
>> This method is to create a non-polarised electrolytic cap. ie. one
>> cap will take the positive cycle, and one takes the negative one.
rs> Electrolytics are a cheap shit way to make a small cap have a BIG
rs> capacitance, but the rub IS that it only has it in ONE direction of
rs> charge! An ideal cap is a non-polarized cap.
rs> -Steve Walz rst...@armory.com ftp.armory.com/pub/user/rstevew
It's fine for small signals. An electrolytic cap will tolerate a small
reverse voltage without damage to the oxide layer. Electrolytic caps are
made to be self-healing in the event of an oxide layer puncture. However
this self-healing mechanism is polarity dependant and only works with the
original "forming" polarity.
If you use an electrolytic capacitor at very low signal levels compared to
its working voltage rating, the danger of puncture is minimal hence the
self-healing mechanism should never be required. So it's quite alright to
use back to back electrolytics with small signals except of course for the
higher self inductance.
btw: I never needed to refer to feces in my explanation. If you can't
control yourself then at least don't give the rest of the net a bad name.
The occasional expletive can be useful though. <lecture mode off>
Arsenio
... Hey! +++*r*'s ++ p++b!C ++++!r !*\* *\/*y Co*<*!!
mi...@bangate.compaq.com said the following to All on the subject of
Richard Steven Walz
John Lundgren <jlun...@delta1.deltanet.com> wrote:
>Michael A. Wyatt (maw...@ix.netcom.com) penned:
>: In <4bm8ah$p...@news.scruz.net> rst...@armory.com (Richard Steven Walz)
>: writes:
>
>: >In article <4bg92m$k...@krel.iea.com>,
>: >>bh72...@omega.ntu.ac.sg (PAUL WU -- NEW YEAR IS COMMING.) wrote:
>
>: >>> In article <4b94tm$g...@ixnews2.ix.netcom.com>,
>: >>> >
>: >>> > A common technique when using an electrolytic cap in a circuit
>: where
>: >>> > the polarity of the DC voltage is unknown is to connect two
>: every
>: >>> > other engineer that I've worked with, but I've no idea *why* it
>: works.
>: >>> > Does anyone have a good explanation?
>: >>
>: >>> This method is to create a non-polarised electrolytic cap. ie.
>: one
>: >>> cap will take the positive cycle, and one takes the negative one.
>| Rancho Santiago Community College | .deltanet.com |
John, a query: Doesn't your 1 Megohm resistor produce signal
distortion with small signals as well? I haven't found a scheme to do
this yet that doesn't seem to introduce a distortion. I have tried a
pull up 1 Meohm, and with a 1V p-p sine I can see see weirdness with my
dual trace scope that I don't see with a ceramic or non-polarized. It's
not a huge distortion, but enough for annoyance.
--
-Steve Walz rst...@armory.com ftp://ftp.armory.com/pub/user/rstevew
mirrored: ftp://ieee.cas.uc.edu/pub/electronics/faqs/ftp.armory.com
and Europe: ftp://ftp.cised.unina.it/pub/electronics/ftp.armory.com
and Oz: ftp://ftp.peninsula.apana.org.au/pub/electronics/ftp.armory.com
John Fields
Steve,
An electrolytic capacitor is _designed_ to accept charge in one direction
only. If it is used as intended, an electrolytic capacitor can have a
very long, reliable life. I own some instrumentation which which is over
forty years old, contains the original electrolytics, and continues to
provide reliable service.
Far from being a "cheap shit" means of achieving a large capacitance in a
small volume, the amount of materials and processing research expended on
developing electrolytic capacitors has been substantial, to say nothing
of the cleverness of the idea of using an oxide film as a dielectric and
an electrolyte as a closely coupled conductor in the first place.
As far as connecting electrolytics "back to back" for AC use goes,
there is nothing wrong with the practice as long as the capacitors
are equal in capacitance, since while one is charging the other is
discharging. Capacitors _are_ allowed to discharge, no?
Assuring that they exhibit equal capacitances may require matching
them, but so what? I'd much rather do that than have to go out and
buy a huge (and much more expensive) bipolar cap to do the same job.
Finally, if you consider electrolytic caps inferior, do a "reducto ad
absurdium" re. their elimination and see what you come up with.
Starship
John Lundgren
When electrolytics are used for bypass caps, any change in capacitance
with voltage is of no consequence because the electrolytics are usually
ten times the value needed to get down the the low freq rolloff point.
Like a 10 uF cap has 10K reactance at 1.6 Hz, which is more than 10 times
lower than the 20 to 30 Hz rolloff point of HiFi gear. This is especially
helpful if the caps dry out or lose capacitance over the years.
If any bypass cap is within the feedback loop of the amplifier, then
there would be less problems with nonlinearity.
Michael A. Wyatt (maw...@ix.netcom.com) penned:
: In <4bt8ld$d...@news2.deltanet.com> jlun...@delta1.deltanet.com (John
: Lundgren) writes:
: >
: >Michael A. Wyatt (maw...@ix.netcom.com) penned:
: >: In <4bm8ah$p...@news.scruz.net> rst...@armory.com (Richard Steven
: Walz)
: >: writes:
: >
: >: >In article <4bg92m$k...@krel.iea.com>,
: >: >
: >
: >: Steve,
: >
: >: By placing diodes across each capacitor in the proper direction
: >: (diode conducts when the capacitor's applied voltage is reversed
: from
: >: the correct polarity), the reverse voltage across each capacitor is
: >: limited to below one volt and the electrolytic capacitor's life is
: >: extended to it's nora\mal expected life.
: >
: >: Hope this helps,
: >
: >: Mike Wyatt
: >
: >I've never, ever seen that done. Probably because the diodes
: introduce
: >nonlinearity into the path. And the diodes are more expensive and
: more
: >complex than a simple 1 meg resistor from the two caps to ground, as I
: >said in the other followup.
: >
: >--
: >#==========H==A==P==P==Y=====H==O==L==I==D==A==Y==S==!==!===========#
: >| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
: >| Rancho Santiago Community College | .deltanet.com |
: >| 17th St @ Bristol \ Santa Ana, CA 92706 | http://rsc.rancho|
: >| My opinions are my own, and not my employer's. | .cc.ca.us |
: >| "You can flame your brains out -- it won't take long." |
: >| I have gone out to look for myself.. If I should |
: >| return before I get back, hold me until I get here. |
: >#======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#
: John,
: I have used this technique a few times. I can't claim I came up
: with the idea, as I saw it in a design about 25 years ago. The diodes
: will intoduce some non-linearity, but probabily not much more than
: without them. If back to back electrolytics (diodes or no diodes) are
: investigated from a linearity standpoint, you will discover they are
: highly non-linear and should be utilized only in low precision circuits
: as I mentioned in an earlier post. In fact, all electrolytic capacitors
: are very non-linear due to the large capacitance change vs applied
: voltage.High dielectric capacitors fall into this catagory also.
: Hope this helps,
: Mike Wyatt
--
#======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#
| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
| Rancho Santiago Community College District | .deltanet.com |
| 17th St at Bristol \ Santa Ana, CA 92706 | http://rsc.rancho|
| My opinions are my own, and not my employer's. | .cc.ca.us |
| I have gone out to look for myself.. If I should |
| return before I get back, hold me until I get here. |
| "You can flame your brains out -- it won't take long." |
#===================================================================#
Michael A. Wyatt
Lundgren) writes:
>
>
>When electrolytics are used for bypass caps, any change in capacitance
>with voltage is of no consequence because the electrolytics are
usually
>ten times the value needed to get down the the low freq rolloff point.
>Like a 10 uF cap has 10K reactance at 1.6 Hz, which is more than 10
times
>lower than the 20 to 30 Hz rolloff point of HiFi gear. This is
especially
>helpful if the caps dry out or lose capacitance over the years.
>
>If any bypass cap is within the feedback loop of the amplifier, then
>there would be less problems with nonlinearity.
>
John,
This might be true in HiFi applications, which I consider low
precision circuits. In HiFi, most people don't care if the bandwidth
changes a little or the signal non-linearity degrades a little (unless
you one of those with the "Golden Ears"). But in high presision
applications such as instrumentation, an electrolytic in the wrong
place can produce a poor performing circuit.
Consider your case of the bypass capacitor; If the effective
impedance of the bypass capacitor, at the lower corner frequency of the
amplifier, was 50 times lower than the impedance being bypassed, and
the non-linearity of the bypass capacitor was only 1%, this system
would introduce an equivalent of 1/2 LSB full scale non-linearity in a
12 bit system (~80dB dynamic range)! Extend this to systems which
require full scale non-linearities of 1/2 LSB in 16 bits (~100dB
dynamic range) or more, and you can see the importance. We have worked
on filters which had requirements of greater than 22 bits full scale
linearity (dynamic range ~140dB, thats less than 100nV non-linearity in
a 1 volt signal), the non-linearity of standard plastic film type
capacitors was unacceptable.
As you mentioned, feedback can be an important technique for
reducing non-linearity. This requires an amplifier with enough GBW
product and internal linearity to reduce the non-linearity at the
frequency(ies) of intrest. In the above extreme case, most op-amps
can't provide the necessary open loop linearity to be useful (most
op-amps in a unity gain configuration at low frequencies aren't good
enough for even 18 to 20 bit systems) . Special amplifiers must be
employed and very carefully, it's a pain but can be done.
Thanks for the reply and Happy New Year to everyone,
Mike Wyatt
John Lundgren
+ +
Input O-------|(-----+-----)|-------O output
|
/
\ Any reasonable value resistor
/ that won't load the input or
\ output too much
|
GND
THe goal is to allow either input or output to be any voltage from 0 to
the max rating of the caps without worrying about the capacitors having a
reverse voltage on them. Assuming that the resistor is a linear device
(last time I checked, it was) then the circuit doesn't have any nonlinear
devices. But then an electrolytic may add distortion in the form of
noise or leakage currents that aren't wanted. The capacitors have to be
twice the overall capacitance. One would think that the two
capacitors being back to back would make any nonlinearity symmetrical.
If any distortion is visible on a scope, then it's pretty bad! I built
the Wein Bridge Oscillator from the "Wave Generating" article in
Electronics Now, Dec, 1995, P. 82, Fig. 7. I added an extra diode in
series with D1 and D2 to get a higher output voltage. It works OK, but
the distortion is visible on the scope. I believe that the only way to
generate low distortion with this circuit is to replace the clipping
circuit with a low voltage lamp, which doesn't cause clipping.
The only way to measure the true amount of distortion is with a
distortion meter.
Off the subject, I've wondered about those three legged caps that are on
the PC four layer motherboards for bypassing. I think that they're two
tantalums in one package. I put in a motherboard one day and fired it up,
and moved on to the next. A few minutes later, the people were telling me
to come over and find out why the PC was smelling awful. I opened it up,
and one of those caps between the slots had burned up, depositing a thick
layer of soot over the adjacent adapter cards and the top of the case.
Looked u-g-l-y. I had to replace the MB. I couldn't tell if the mfgr
had put it in backwards, because it was just a charred mess. Cleanup was
no fun.
Richard Steven Walz (rst...@armory.com) penned:
: In article <4bt8ld$d...@news2.deltanet.com>,
: >| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
: >| Rancho Santiago Community College | .deltanet.com |
: ----------------------------------------
: John, a query: Doesn't your 1 Megohm resistor produce signal
: distortion with small signals as well? I haven't found a scheme to do
: this yet that doesn't seem to introduce a distortion. I have tried a
: pull up 1 Meohm, and with a 1V p-p sine I can see see weirdness with my
: dual trace scope that I don't see with a ceramic or non-polarized. It's
: not a huge distortion, but enough for annoyance.
: --
: -Steve Walz rst...@armory.com ftp://ftp.armory.com/pub/user/rstevew
: mirrored: ftp://ieee.cas.uc.edu/pub/electronics/faqs/ftp.armory.com
: and Europe: ftp://ftp.cised.unina.it/pub/electronics/ftp.armory.com
: and Oz: ftp://ftp.peninsula.apana.org.au/pub/electronics/ftp.armory.com
--
#======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#
| John Lundgren - Elec Tech - Info Tech Svcs. | jlundgre@delta1 |
| Rancho Santiago Community College District | .deltanet.com |
| 17th St at Bristol \ Santa Ana, CA 92706 | http://rsc.rancho|
| My opinions are my own, and not my employer's. | .cc.ca.us |
| I have gone out to look for myself.. If I should |
| return before I get back, hold me until I get here. |
| "You can flame your brains out -- it won't take long." |
#===T=u=z=l=a==C=o=m=p=a=n=y=.=.===t=h=r=e=e='=s==L=e==C=r=o=w=d=!==#
John Lundgren
Theoretical circuit with +20 volts DC on the output of the first stage, and
zero volts DC on the input of the second stage. The two capacitors
back-to-back couple the two stages together. The capacitors are equal in
value. The negatives of the caps are connected together. The circuit is
powered off.
When the circuit is first powered on, the first stage is brought up to +20
VDC. THe two capacitors split the voltage equally, 10 volts DC across
each one. This is assuming that neither capacitor has any leakage.
The first or left capacitor has +20 VDC on the + and +10VDC on the -.
The second capacitor has +10 VDC on the - and 0 VDC on the + ...
OOPS! This cap has reversed polarity! But since it leaks away in a short
while and the first cap charges up to a full 20VDC, no damage is done.
Or has it?
John Fields (star...@freeside.fc.net) penned:
: ----------------------------------------------
: Steve,
: An electrolytic capacitor is _designed_ to accept charge in one direction
: only. If it is used as intended, an electrolytic capacitor can have a
: very long, reliable life. I own some instrumentation which which is over
: forty years old, contains the original electrolytics, and continues to
: provide reliable service.
: Far from being a "cheap shit" means of achieving a large capacitance in a
: small volume, the amount of materials and processing research expended on
: developing electrolytic capacitors has been substantial, to say nothing
: of the cleverness of the idea of using an oxide film as a dielectric and
: an electrolyte as a closely coupled conductor in the first place.
: As far as connecting electrolytics "back to back" for AC use goes,
: there is nothing wrong with the practice as long as the capacitors
: are equal in capacitance, since while one is charging the other is
: discharging. Capacitors _are_ allowed to discharge, no?
: Assuring that they exhibit equal capacitances may require matching
: them, but so what? I'd much rather do that than have to go out and
: buy a huge (and much more expensive) bipolar cap to do the same job.
: Finally, if you consider electrolytic caps inferior, do a "reducto ad
: absurdium" re. their elimination and see what you come up with.
: Starship
--
John Lundgren
: Lundgren) writes:
This is not the typical electronics gear that is found in common use. The
high precision D-A is going to have other problems as well. Like
regulation and noise in the power supply, and problems with thermal drift.
That's probably why everyone wants to digitize, process and transport the
signals, so that the analog design problems won't have to be dealt with.
The CDs have D-A converters that are supposed to be something like 18
bits or whatever it is when rounded off. But the straight analog gear
doesn't have to deal with that wide a dynamic range. I guess for audio
purposes, the ear can tolerate some distortion in the high freqs probably
because the distortion harmonics are too high to be audible.
I've read that back before digital signals came into use, the phone
system had to amplify the analog signals a hundred times between coasts,
and the tiny amount of distortion added at each repeater could cause
major problems on the other end. That's why the negative feedback was so
important.
: >Michael A. Wyatt (maw...@ix.netcom.com) penned:
: >: John,
#===T=u=z=l=a==C=o=m=p=a=n=y=.=.===t=h=r=e=e='=s==L=e==C=r=o=w=d=!==#
A. Shiekh
to 30% reverse voltage.
Anyone explain why?