| lt cdr data |
I am replacing a few caps, on the output stage of a line amplifier.
It has non polarized caps., I have some suitable sized polarized ones, can I drop them in here, I am sure I have seen amps with coupling with polarized, and does the orientation matter?
thanks |
|
|
| destroyer X |
Interconnect...join the positive and you will have a unt that will work fine in that position.
2 capacitors to replace one...it is ugly but i am using those last 46 years and had no problem.
There are people that say that do not works....wrong!...i am using those last 46 years..for sure it works...even not making any sense.
regards,
Carlos |
|
|
| myhrrhleine |
Two caps in series + to + makes non-polar
It also tends to cancel internal distortions.
Use double (or more) the capacitance and the highest voltage available for lowest distortion |
|
|
| max29 |
Interesting to try...
Two caps in series you connect the + and + with each other
But why not the - and - . Or does it then not work as a non polar ? |
|
|
| Christer |
It doesn't matter if you connect them at the positive ends or at the negative ends. Both ways work.
Edit: Maybe somebody has a reason why one of the alternatives should be preferred, but technically both variants do the job. |
|
|
| AndrewT |
Hi,
electrolytics are usually considered as poor quality substitutes.
Why do you want to fit them in lieu of non polar? |
|
|
| jmagister |
Non polarised are electrolitic and often of high value (>10uF).
If you don't found them, you need to replace them with self-made non-polarised :D
Regards |
|
|
| AndrewT |
Hi,
Ltcdr has not said he wants to replace non polar electrolytics with polar electrolytics.
He is referring to aand these can usually manage with a few uF and should not need to resort to using electrolytics. |
|
|
| pooge |
maybe a little off topic, but has anyone ever tried charge coupling the series pair of electrolytics in an amplifier, such as in the feedback DC blocking position? JBL uses charge coupling of series resistors in their top-of-the-line speakers. They use this for film caps. I was wondering if it could be applied to series electrolytics, with due consideration for the voltage rating of the electrolytics.
"It is referred to as "Charge-Coupled Linear Definition Dividing Network". It utilized a biasing voltage in the crossover network to keep the music signal from crossing the dielectric zero-point of the capacitors. The result is more linear operation of the network with a significant improvement in sonic performance."
"The Charge-Coupled crossover network includes a bias battery that supplies constant DC bias to the capacitor, so that the network is always "charged", and avoids generating distortion while switching between the on and off states."
The positive of nine volt battery is connected through a 2megohm resistor to the junction point of the series caps. The negative is connected to ground. In a feedback network, 9 volts may be too much, depending on the voltage rating of the electrolytics. The charge is said to provide a bias to "squeeze" the dielectric to reduce microphonics, and also to prevent voltage reversals in the dielectric so that the cap's DA doesn't come in to play as much. |
|
|
| auplater |
What's the difference?
Polarized caps use tanatalum/niobium/aluminum foil for the plates and an anodically produced oxide layer for the dielectric. This oxide layer is produced by electrochemical means, applying an ever increasing voltage to the foil surface to a predetermined level to build up the oxide. This is produced by making the foil an anode (+) in an electrochemical circuit. maybe polarized to 40/50/60 volts until little current flows.
Non polarized caps use some sort of conductive foil with a plastic insulating film as the dielectric.
Why the distinction? In an electrical circuit, polarized caps rely on a constant (+) polarity on the anodized side of the cap to maintain the integrity of the dielectric. If one were to imposed a negative (-) current at sufficient voltage to this coating, it will start to de-oxidize and ultimately fail. I'm not sure what a continuos ac signal will do to an electrolytic over time, but I suppose it depends on the electrolyte, applied voltage, and construction material. I've used back to back eletrolytics in xovers, etc... and while they seem to work, I'd operate them well below they're rated voltage and not expect a loonngg lifespan.
Of course, ymmv.
John L. |
|
|
| K-amps |
| quote: | Originally posted by myhrrhleine
Two caps in series + to + makes non-polar
It also tends to cancel internal distortions.
Use double (or more) the capacitance and the highest voltage available for lowest distortion |
;) I have been connecting (-) to (-) and thats been working for me fo rthe last 20 years... |
|
|
| Christer |
Pooge,
that is an interesting question you raise, and I have been thinking about the same thing. However, Self reports from his distorsion measurements of electrolytics that the distorsion increases slightly with the bias voltage. So unless the signal voltage over the cap is large enough to cause too much reverse voltage (most source claim the limit is somewhere around 1.5 V), it seems that no biasing at all should be best. |
|
|
| lineup |
| quote: | Originally posted by lt cdr data
I am replacing a few caps, on the output stage of a line amplifier.
|
A little more of this line amplifier would be nice.
What will work in one case would not be good in another.
In audio the choice of a component is very often depending of the actual circuit in question.
Best would be if you could post a schema or a drawing
or maybe post or link to some other similar amplifier schematic.
To get a more specific answer, we need to be detailed and specific when describing the situation.
General and not specific questions - will cause general answers.
Answers with a lot of if and in case statements,
and many times even some really off topic speculations and ideas.
......We do not even know if is dual +- power supply or a single ended voltage supply.
This really matters a whole lot!
regards
lineup |
|
|
| pooge |
| quote: | Originally posted by Christer
Pooge,
that is an interesting question you raise, and I have been thinking about the same thing. However, Self reports from his distorsion measurements of electrolytics that the distorsion increases slightly with the bias voltage. So unless the signal voltage over the cap is large enough to cause too much reverse voltage (most source claim the limit is somewhere around 1.5 V), it seems that no biasing at all should be best. |
Still seems like worth a try, since in many cases, back to back electrolytics are used in a feedback shunt. But the voltage value of these electrolytics are usually only 16-35v, so maybe 9 volts would be too much bias. I also wonder if Self's test of distortion would be applicable to biasing the midpoint of a series connected pair; or if the distortion he measured my not be as significant as any possible advantages that might be had from the biasing. I guess the only thing to do is try it. Anyone have any opinion on how much voltage to apply? |
|
|
|
