Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > sci.physics > #530234 > unrolled thread

Capacitor Question

Started byFabian Russell <root@localhost.localdomain>
First post2015-11-03 22:14 +0000
Last post2015-11-05 16:34 -0800
Articles 20 on this page of 53 — 7 participants

Back to article view | Back to sci.physics


Contents

  Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-03 22:14 +0000
    Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-03 16:35 -0600
      Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-03 22:59 +0000
        Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-03 18:10 -0600
    Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-03 22:47 +0000
      Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-03 23:07 +0000
        Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-03 23:20 +0000
          Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-03 23:42 +0000
          Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 02:49 +0000
            Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-03 21:40 -0600
              Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 04:01 +0000
                Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-04 06:03 +0000
                  Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 07:07 +0000
                    Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-04 20:29 +0000
                      Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 20:51 +0000
                        Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-04 21:08 +0000
                          Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 21:53 +0000
                            Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-04 22:20 +0000
                              Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 22:42 +0000
                                Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-04 23:07 +0000
                                  Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 00:28 +0000
                                    Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-05 00:46 +0000
                                  Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 00:39 +0000
                                    Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-05 00:58 +0000
                                Re: Capacitor Question benj <none@gmail.com> - 2015-11-04 18:25 -0500
                              Re: Capacitor Question benj <none@gmail.com> - 2015-11-04 18:37 -0500
                                Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 00:09 +0000
                                  Re: Capacitor Question benj <nobody@gmail.com> - 2015-11-04 21:35 -0500
                                    Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 21:51 -0600
                                    Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 04:19 +0000
                                      Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-05 11:10 -0600
                                        Re: Capacitor Question "reber g=emc^2" <herbertglazier0@gmail.com> - 2015-11-05 09:15 -0800
                                        Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 18:08 +0000
                                      Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-05 19:26 +0000
                                        Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 19:44 +0000
                                          Re: Capacitor Question jimp@specsol.spam.sux.com - 2015-11-05 19:56 +0000
                                            Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 20:53 +0000
                                            Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 21:52 +0000
                        Re: Capacitor Question benj <none@gmail.com> - 2015-11-04 18:41 -0500
                          Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 21:54 -0600
            Re: Capacitor Question benj <nobody@gmail.com> - 2015-11-04 02:17 -0500
              Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 07:30 +0000
                Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 07:48 -0600
                Re: Capacitor Question benj <nobody@gmail.com> - 2015-11-04 14:04 -0500
                  Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-04 19:28 +0000
                    Re: Capacitor Question benj <none@gmail.com> - 2015-11-04 18:49 -0500
                      Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 20:41 -0600
                        Re: Capacitor Question Fabian Russell <root@localhost.localdomain> - 2015-11-05 03:16 +0000
                          Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 21:58 -0600
            Re: Capacitor Question "reber g=emc^2" <herbertglazier0@gmail.com> - 2015-11-04 12:14 -0800
              Re: Capacitor Question benj <none@gmail.com> - 2015-11-04 18:44 -0500
                Re: Capacitor Question gilber34 <fafa@invalid.com> - 2015-11-04 20:41 -0600
    Re: Capacitor Question gyansorova@gmail.com - 2015-11-05 16:34 -0800

Page 1 of 3  [1] 2 3  Next page →


#530234 — Capacitor Question

FromFabian Russell <root@localhost.localdomain>
Date2015-11-03 22:14 +0000
SubjectCapacitor Question
Message-ID<pan.2015.11.03.22.14.56@localhost.localdomain>
Electrolytic Capacitor Question
--------------------------------

A) Prologue of logic:

This question either is physics or is not physics.

This question is not physics.

Therefore it is.

B) Question

I have several home-brew circuits which were built several
years ago.  These circuits contain a few large capacity and other
electrolytic capacitors.

What is the lifetime of an electrolytic cap before it needs
replacing?

Because the very large caps are involved in power supply ripple
filtering, I have placed these into a capacitor short circuit
that I can can switch on after power off.  Does having these
caps on s short circuit improve lifespan?

[toc] | [next] | [standalone]


#530242

Fromgilber34 <fafa@invalid.com>
Date2015-11-03 16:35 -0600
Message-ID<n1bcsj$fla$1@speranza.aioe.org>
In reply to#530234
On 11/3/2015 4:14 PM, Fabian Russell wrote:
> Electrolytic Capacitor Question
> --------------------------------
>
> A) Prologue of logic:
>
> This question either is physics or is not physics.
>
> This question is not physics.
>
> Therefore it is.
>
> B) Question
>
> I have several home-brew circuits which were built several
> years ago.  These circuits contain a few large capacity and other
> electrolytic capacitors.
>
> What is the lifetime of an electrolytic cap before it needs
> replacing?

10 to 40 years, depends upon construction and if it had power on it most 
of the time. Also a lot of bad electrolytics were build about 10 to 15 
years ago, and there is a lot of computer stuff out there now, that have 
expanding electrolitics in them now, failing. (like viewsonic monitors) 
  if paper, get rid of them.


>
> Because the very large caps are involved in power supply ripple
> filtering, I have placed these into a capacitor short circuit
> that I can can switch on after power off.  Does having these
> caps on s short circuit improve lifespan?
>

no,
you can replace that switch with a resistor across the cap, which will 
drain power off cap.  something like a 1/4 watt, 20k ohm, depends upon 
voltage and a few other things.



[toc] | [prev] | [next] | [standalone]


#530251

FromFabian Russell <root@localhost.localdomain>
Date2015-11-03 22:59 +0000
Message-ID<pan.2015.11.03.22.57.30@localhost.localdomain>
In reply to#530242
On Tue, 03 Nov 2015 16:35:59 -0600, gilber34 wrote:

>
> you can replace that switch with a resistor across the cap, which will 
> drain power off cap.
>

That's what I have.  After I power off, I close a switch that connects
the two ends of the cap though a resistor.  This neutralizes any remaining
charge on the cap.

This action will not improve life?

[toc] | [prev] | [next] | [standalone]


#530268

Fromgilber34 <fafa@invalid.com>
Date2015-11-03 18:10 -0600
Message-ID<n1bidf$pum$1@speranza.aioe.org>
In reply to#530251
On 11/3/2015 4:59 PM, Fabian Russell wrote:
> On Tue, 03 Nov 2015 16:35:59 -0600, gilber34 wrote:
>
>>
>> you can replace that switch with a resistor across the cap, which will
>> drain power off cap.
>>
>
> That's what I have.  After I power off, I close a switch that connects
> the two ends of the cap though a resistor.  This neutralizes any remaining
> charge on the cap.
>
> This action will not improve life?
>

it should,  depends upon the type of electrolitic,
after a long time with a voltage on it, some electrolitics behave like 
weak batteries at 1/2 the voltage applied,

In your above design, you can size the resistor such you do not need the 
switch in most cases

[toc] | [prev] | [next] | [standalone]


#530254

Fromjimp@specsol.spam.sux.com
Date2015-11-03 22:47 +0000
Message-ID<q5qngc-nmb.ln1@mail.specsol.com>
In reply to#530234
Fabian Russell <root@localhost.localdomain> wrote:
> Electrolytic Capacitor Question
> --------------------------------
> 
> A) Prologue of logic:
> 
> This question either is physics or is not physics.
> 
> This question is not physics.
> 
> Therefore it is.
> 
> B) Question
> 
> I have several home-brew circuits which were built several
> years ago.  These circuits contain a few large capacity and other
> electrolytic capacitors.
> 
> What is the lifetime of an electrolytic cap before it needs
> replacing?
> 
> Because the very large caps are involved in power supply ripple
> filtering, I have placed these into a capacitor short circuit
> that I can can switch on after power off.  Does having these
> caps on s short circuit improve lifespan?

The usual way is to put a large resistor across the capacitors. 

Assuming you are talking about modern, i.e. well after the WWII era,
capacitors, there are essentially three different types of electrolytic
capacitors: aluminum, tantalum, and niobium.

Each has different expected operating life spans which are usually given as
curves of operating temperature at specified rating and range from the
low thousand hours to tens of thousand of hours.

Non-solid aluminum types have issues with storage and long unpowered
periods, so if your capacitors are non-solid aluminum you are decreasing
their life.

Read this: https://en.wikipedia.org/wiki/Electrolytic_capacitor

If your capacitors have manufacturers info on them, try the manufacturer
site for spec sheets.


-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530255

FromFabian Russell <root@localhost.localdomain>
Date2015-11-03 23:07 +0000
Message-ID<pan.2015.11.03.23.06.12@localhost.localdomain>
In reply to#530254
On Tue, 03 Nov 2015 22:47:54 +0000, jimp wrote:

> 
> The usual way is to put a large resistor across the capacitors. 
> 

That's what I have.  After I power off, I close a switch that connects
the two ends of the cap though a resistor.  This neutralizes any remaining
charge on the cap.

This does improve life, correct?

[toc] | [prev] | [next] | [standalone]


#530265

Fromjimp@specsol.spam.sux.com
Date2015-11-03 23:20 +0000
Message-ID<p2sngc-esb.ln1@mail.specsol.com>
In reply to#530255
Fabian Russell <root@localhost.localdomain> wrote:
> On Tue, 03 Nov 2015 22:47:54 +0000, jimp wrote:
> 
>> 
>> The usual way is to put a large resistor across the capacitors. 
>> 
> 
> That's what I have.  After I power off, I close a switch that connects
> the two ends of the cap though a resistor.  This neutralizes any remaining
> charge on the cap.
> 
> This does improve life, correct?
 
Not if it is a non-solid aluminum; it shortens the life.

Read the link.


-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530266

FromFabian Russell <root@localhost.localdomain>
Date2015-11-03 23:42 +0000
Message-ID<pan.2015.11.03.23.42.44@localhost.localdomain>
In reply to#530265
On Tue, 03 Nov 2015 23:20:25 +0000, jimp wrote:

>  
> Not if it is a non-solid aluminum; it shortens the life.
> 

Oh sh!t!  I got problems.  These are the common non-solid types.

I was wondering why the LED indicators were fading so quickly
when I power down.  They used to glow for 10 seconds but now
they're quenched within 2 seconds.

The LEDs are placed across the power rails to indicate that the
circuit is energized.  If they fade quickly after power off it
probably means that the power supply caps are not holding charge. 

Fortunately, they will be easy to cut out and re-solder in new
ones -- solid this time.

[toc] | [prev] | [next] | [standalone]


#530281

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 02:49 +0000
Message-ID<pan.2015.11.04.02.49.26@localhost.localdomain>
In reply to#530265
On Tue, 03 Nov 2015 23:20:25 +0000, jimp wrote:

>  
> Not if it is a non-solid aluminum; it shortens the life.
> 

Solid electrolyte aluminum capacitors seem only available with
a polymer electrolyte.  I need at least 4700uF caps at a 50 V
rating.  A check of both DigiKey and Mouser supply houses show
that 4700uF is available only at a max of 2.5 V rating.  I assume
this represents the state of the art.

This is way too low.  I'll have to get the wet, non-solid, electrolyte
and just replace when needed.

[toc] | [prev] | [next] | [standalone]


#530285

Fromgilber34 <fafa@invalid.com>
Date2015-11-03 21:40 -0600
Message-ID<n1bumm$db0$1@speranza.aioe.org>
In reply to#530281
On 11/3/2015 8:49 PM, Fabian Russell wrote:
> On Tue, 03 Nov 2015 23:20:25 +0000, jimp wrote:
>
>>
>> Not if it is a non-solid aluminum; it shortens the life.
>>
>
> Solid electrolyte aluminum capacitors seem only available with
> a polymer electrolyte.  I need at least 4700uF caps at a 50 V
> rating.  A check of both DigiKey and Mouser supply houses show
> that 4700uF is available only at a max of 2.5 V rating.  I assume
> this represents the state of the art.
>
> This is way too low.  I'll have to get the wet, non-solid, electrolyte
> and just replace when needed.
>

Ive had good sucess with surplus 20 year old big caps like that.
IF your current is low, you can use an active device to filter much more 
effectively, using an adjustable regulator

[toc] | [prev] | [next] | [standalone]


#530287

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 04:01 +0000
Message-ID<pan.2015.11.04.04.01.27@localhost.localdomain>
In reply to#530285
On Tue, 03 Nov 2015 21:40:01 -0600, gilber34 wrote:

>
> IF your current is low, you can use an active device to filter much more 
> effectively, using an adjustable regulator
>

Current is not necessarily low, although it could be.

I do use a fixed regulator.  The caps are between a diode bridge rectifier
and the regulator.

Brand new manufactured caps are supposed to have the bugs worked out,
but I'll plan on replacing them every few years anyway.

[toc] | [prev] | [next] | [standalone]


#530289

Fromjimp@specsol.spam.sux.com
Date2015-11-04 06:03 +0000
Message-ID<smjogc-c7e.ln1@mail.specsol.com>
In reply to#530287
Fabian Russell <root@localhost.localdomain> wrote:
> On Tue, 03 Nov 2015 21:40:01 -0600, gilber34 wrote:
> 
>>
>> IF your current is low, you can use an active device to filter much more 
>> effectively, using an adjustable regulator
>>
> 
> Current is not necessarily low, although it could be.
> 
> I do use a fixed regulator.  The caps are between a diode bridge rectifier
> and the regulator.
> 
> Brand new manufactured caps are supposed to have the bugs worked out,
> but I'll plan on replacing them every few years anyway.

Where's the inductor?

A series inductor goes a long way in reducing ripple and the size of
the capacitor to get a specified ripple.
 

-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530295

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 07:07 +0000
Message-ID<pan.2015.11.04.07.08.15@localhost.localdomain>
In reply to#530289
On Wed, 04 Nov 2015 06:03:40 +0000, jimp wrote:

> 
> Where's the inductor?
> 

With a voltage regulator you don't really need an inductor
because the regulator greatly reduces the ripple.

Besides smoothing the ripple, a large cap also insures
that any current surges drawn by the circuit can be
easily handled.

So you have AC --> transformer --> rectifier --> cap filter
--> regulator --> circuit.

Some DC circuits don't require any filter or regulator at all.
They can be fed the output from the rectifier directly.

The needs have to be determined beforehand and a suitable
power source implemented.

My power supplies are general purpose and so I tend to over
design them to be able to meet any needs.

Variable regulators can output a wide voltage range, but I never
use them.  The problem is that once the voltage difference
between the transformer output and the regulator output becomes
too large then way too much heat is produced within the regulator.
I always used a fixed regulator with a 5 Volt difference (recommended
by the specs).

So I have to have lots of power supplies to meet varying voltage requirements.

[toc] | [prev] | [next] | [standalone]


#530387

Fromjimp@specsol.spam.sux.com
Date2015-11-04 20:29 +0000
Message-ID<oe6qgc-b2k.ln1@mail.specsol.com>
In reply to#530295
Fabian Russell <root@localhost.localdomain> wrote:
> On Wed, 04 Nov 2015 06:03:40 +0000, jimp wrote:
> 
>> 
>> Where's the inductor?
>> 
> 
> With a voltage regulator you don't really need an inductor
> because the regulator greatly reduces the ripple.
> 
> Besides smoothing the ripple, a large cap also insures
> that any current surges drawn by the circuit can be
> easily handled.
> 
> So you have AC --> transformer --> rectifier --> cap filter
> --> regulator --> circuit.
> 
> Some DC circuits don't require any filter or regulator at all.
> They can be fed the output from the rectifier directly.
> 
> The needs have to be determined beforehand and a suitable
> power source implemented.
> 
> My power supplies are general purpose and so I tend to over
> design them to be able to meet any needs.
> 
> Variable regulators can output a wide voltage range, but I never
> use them.  The problem is that once the voltage difference
> between the transformer output and the regulator output becomes
> too large then way too much heat is produced within the regulator.
> I always used a fixed regulator with a 5 Volt difference (recommended
> by the specs).
> 
> So I have to have lots of power supplies to meet varying voltage requirements.

It seems you are a staunch believer in the BAC (Big Ass Capacitor) method
of filter design.

This is OK for most applications as it eliminates two components and
having to put any thought into the design.

However, if your power supply is greater than 5 V where your choices
for BAC's is more limited and much more expensive, it starts to become
smarter to do a little analysis if ripple is indeed important to you.

If you do the analysis, you will find that a pi filter will provide
the same ripple rejection as your single 4700 uf capacitor with capacitor
values far less than 4700 uf. Depending on the system requirements,
it is not unusual to find the cost of two smaller capacitors and an
inductor to be quite a bit less than that of one BAC.

Also, your statement that "a large cap also insures that any current
surges drawn by the circuit can be easily handled" tells me you have
never actually done such an analysis and are just depending on the BAC
to take care of everything, which in many cases it won't.
 
In reality the surge current to the supplied device is limited by
the wiring inductance from the BAC to the regulator, the regulator,
and the wiring from the regulator to the supplied device.

That is why you put a capacitor on the output of the regulator and
a local capacitor at the supplied device.


-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530394

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 20:51 +0000
Message-ID<pan.2015.11.04.20.51.23@localhost.localdomain>
In reply to#530387
On Wed, 04 Nov 2015 20:29:44 +0000, jimp wrote:

>
> It seems you are a staunch believer in the BAC (Big Ass Capacitor) method
> of filter design.
>

Nope.  Just big enough to do the job.  For my power supplies I use 2200uF @ 25V,
1000uF @20V, etc. etc.

In fact, when all specs for a linear supply are considered, my caps follow
the Goldilocks principle: not too big, not too small, but just right.


> 
> If you do the analysis,
>

I don't need no analysis.  I just put an oscilloscope on the output
and watch as everything stays perfectly flat with various loads.

Power supplies are quite trivial and not worth diddling over.

I save my in-depth analysis for the circuits that are powered by them.

[toc] | [prev] | [next] | [standalone]


#530398

Fromjimp@specsol.spam.sux.com
Date2015-11-04 21:08 +0000
Message-ID<5o8qgc-29k.ln1@mail.specsol.com>
In reply to#530394
Fabian Russell <root@localhost.localdomain> wrote:
> On Wed, 04 Nov 2015 20:29:44 +0000, jimp wrote:
> 
>>
>> It seems you are a staunch believer in the BAC (Big Ass Capacitor) method
>> of filter design.
>>
> 
> Nope.  Just big enough to do the job.  For my power supplies I use 2200uF @ 25V,
> 1000uF @20V, etc. etc.
> 
> In fact, when all specs for a linear supply are considered, my caps follow
> the Goldilocks principle: not too big, not too small, but just right.
> 
> 
>> 
>> If you do the analysis,
>>
> 
> I don't need no analysis.  I just put an oscilloscope on the output
> and watch as everything stays perfectly flat with various loads.
> 
> Power supplies are quite trivial and not worth diddling over.
> 
> I save my in-depth analysis for the circuits that are powered by them.


Right, which is why you spend close to $100 for one capacitor.

 

-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530402

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 21:53 +0000
Message-ID<pan.2015.11.04.21.53.14@localhost.localdomain>
In reply to#530398
On Wed, 04 Nov 2015 21:08:53 +0000, jimp wrote:

> 
> Right, which is why you spend close to $100 for one capacitor.
>

That's not me.  That's benj the thrifty shopper.  He couldn't rival
the average housewife in terms of financial prudence.

But maybe you help with this one.

I have an audio power amplifier built with discrete PNP (or is it NPN?)
transistors. (Yes, they still make them in droves.)

The design uses two transistors in a "mirror image" configuration with
one to handle the positive swings and the other to handle the negative
swings (there is a name for this config but I can't recall what it is).
There are some extra parts to cancel the bias currents and other things
but the design is nothing more than that.

It performs very well with no distortion whatsoever.

Now I use this amp to power my headphones, although it can kick out
a lot more power if needed.
 
I need music to help me with my workouts on my exercise machines,
thus I have the amp right next to my machine in the basement.
The headphones are attached to the amp with about 10 feet of
audio cable.

I have another exercise machine on the first floor.  Rather than
carry the amp up from basement to first floor and back again, I
decided to run about 50 feet of audio cable from the basement amp
to the headphones on the first floor (the distance is not quite 50 feet
but it always better to have a little more than less).

So everything is the same, except for an extra 40 feet of audio
cable between amp and headphones (1K ohm input impedance).

However, when I switched on the power and output, a fuse on one
rail of the power supply blew out after a few seconds -- and these
are 2 amp fuses!

What is happening?

I could put an oscilloscope on the output and observe what happens
but I don't want to blow another fuse -- and possibly mess up my
circuit.

I strongly suspect that the extra cable length is somehow inducing
oscillations at some supra-audio frequency and these oscillations
are drawing a lot of current from the supply.

So how can I fix it?

How about a nice choke (inductor) on the output?

[toc] | [prev] | [next] | [standalone]


#530414

Fromjimp@specsol.spam.sux.com
Date2015-11-04 22:20 +0000
Message-ID<gtcqgc-mok.ln1@mail.specsol.com>
In reply to#530402
Fabian Russell <root@localhost.localdomain> wrote:
> On Wed, 04 Nov 2015 21:08:53 +0000, jimp wrote:
> 
>> 
>> Right, which is why you spend close to $100 for one capacitor.
>>
> 
> That's not me.  That's benj the thrifty shopper.  He couldn't rival
> the average housewife in terms of financial prudence.
> 
> But maybe you help with this one.
> 
> I have an audio power amplifier built with discrete PNP (or is it NPN?)
> transistors. (Yes, they still make them in droves.)
> 
> The design uses two transistors in a "mirror image" configuration with
> one to handle the positive swings and the other to handle the negative
> swings (there is a name for this config but I can't recall what it is).
> There are some extra parts to cancel the bias currents and other things
> but the design is nothing more than that.
> 
> It performs very well with no distortion whatsoever.
> 
> Now I use this amp to power my headphones, although it can kick out
> a lot more power if needed.
>  
> I need music to help me with my workouts on my exercise machines,
> thus I have the amp right next to my machine in the basement.
> The headphones are attached to the amp with about 10 feet of
> audio cable.
> 
> I have another exercise machine on the first floor.  Rather than
> carry the amp up from basement to first floor and back again, I
> decided to run about 50 feet of audio cable from the basement amp
> to the headphones on the first floor (the distance is not quite 50 feet
> but it always better to have a little more than less).
> 
> So everything is the same, except for an extra 40 feet of audio
> cable between amp and headphones (1K ohm input impedance).
> 
> However, when I switched on the power and output, a fuse on one
> rail of the power supply blew out after a few seconds -- and these
> are 2 amp fuses!
> 
> What is happening?
> 
> I could put an oscilloscope on the output and observe what happens
> but I don't want to blow another fuse -- and possibly mess up my
> circuit.
> 
> I strongly suspect that the extra cable length is somehow inducing
> oscillations at some supra-audio frequency and these oscillations
> are drawing a lot of current from the supply.
> 
> So how can I fix it?
> 
> How about a nice choke (inductor) on the output?
 
If it is oscillating it is likely that your amplifier does not like
inductive loads, i.e. all that cable, so adding an inductor will probably
just make it worse.

If it were me, I would shit can all that wire and buy a set of wireless
stereo headphones, which can be had for well under $50 these days.


-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


#530418

FromFabian Russell <root@localhost.localdomain>
Date2015-11-04 22:42 +0000
Message-ID<pan.2015.11.04.22.43.07@localhost.localdomain>
In reply to#530414
On Wed, 04 Nov 2015 22:20:00 +0000, jimp wrote:

> 
> buy a set of wireless
> stereo headphones
>

Not me.  I am philosophically opposed to wireless devices.

I have no wireless mouse, no wireless keyboard, no wireless router,
etc., etc.

For me, the signal has got to be in the cable and only in the cable.

It's called a human factor.  It may be irrational, but it is also
incontrovertible.

But as regards the problem, I'll fix it, eventually.

[toc] | [prev] | [next] | [standalone]


#530431

Fromjimp@specsol.spam.sux.com
Date2015-11-04 23:07 +0000
Message-ID<imfqgc-l0l.ln1@mail.specsol.com>
In reply to#530418
Fabian Russell <root@localhost.localdomain> wrote:
> On Wed, 04 Nov 2015 22:20:00 +0000, jimp wrote:
> 
>> 
>> buy a set of wireless
>> stereo headphones
>>
> 
> Not me.  I am philosophically opposed to wireless devices.
> 
> I have no wireless mouse, no wireless keyboard, no wireless router,
> etc., etc.
> 
> For me, the signal has got to be in the cable and only in the cable.
> 
> It's called a human factor.  It may be irrational, but it is also
> incontrovertible.

Sounds like an irrational fear of RF to me.

> But as regards the problem, I'll fix it, eventually.

Maybe.

The way to fix the problem is to first measure what the load, i.e. the
headphone and all the cable, presents to the amplifier.

Then build a spice model of the amplifier with the load and run a
transit analysis.

That should tell you what you need to do.

Decades ago I was involved in the development of wire guided torpedoes.

When tested, guidance always failed.

Long story short, the actual impedance of the load as the wire spooled
out was measured and found to be a complex spiral when plotted on a
Smith chart. The driver amplifier had to be totally redesigned to drive
that load.



-- 
Jim Pennino

[toc] | [prev] | [next] | [standalone]


Page 1 of 3  [1] 2 3  Next page →

Back to top | Article view | sci.physics


csiph-web