In most circuits for amplifiers there is an input capacitor.
I have seen several variations including 2.2uF tantalum, 1uF bipolar, 1 uF electrolytic.100n,4.7uF, and some such as the digi125 don,t show any input cap.
Is there a preferred or better input capacitor?
Would it be possible to replace a 1uF bipolar which is shown in a circuit diagram as the input cap, with say any 1uF cap?
If I use a standard electrolytic I understand the positive goes to the input source.?
I have seen several variations including 2.2uF tantalum, 1uF bipolar, 1 uF electrolytic.100n,4.7uF, and some such as the digi125 don,t show any input cap.
Is there a preferred or better input capacitor?
Would it be possible to replace a 1uF bipolar which is shown in a circuit diagram as the input cap, with say any 1uF cap?
If I use a standard electrolytic I understand the positive goes to the input source.?
The value of the cap is dependent on the input impedance of the circuit. As for the type, that depends on preference I guess.
It may not be a good idea to replace a non-polarised (bipolar) cap with a polarised one as it may fail. The orientation of a polarised cap depends upon the offset voltages around it and whether it will be connected to a single-rail circuit or not. In my own designs any input or output cap is used -ve as the 'free' end.
It may not be a good idea to replace a non-polarised (bipolar) cap with a polarised one as it may fail. The orientation of a polarised cap depends upon the offset voltages around it and whether it will be connected to a single-rail circuit or not. In my own designs any input or output cap is used -ve as the 'free' end.
Hi,
Usually the input cap is either bipolar or back to back polar. In both these cases there is no polarity problem.
The longtail pair, if used as the input stage, will have a polarising voltage of about 700mV and should have a resistor tying this to the clean ground to bias the base current. The polarity of a polar input cap must be oriented to suit either the PNP or NPN used for the LTP. You can adjust the time constant of the input cap to give the bass response/roll off you require in conjunction with that bias resistor referred to earlier. Remember that large electrolytics have quite a bit of inductance and might roll of the treble hence the popularity of bypassing with a film or better. Also two back to back caps have half the effective capacitance and raise the roll off point one octave.
regards Andrew T.
Usually the input cap is either bipolar or back to back polar. In both these cases there is no polarity problem.
The longtail pair, if used as the input stage, will have a polarising voltage of about 700mV and should have a resistor tying this to the clean ground to bias the base current. The polarity of a polar input cap must be oriented to suit either the PNP or NPN used for the LTP. You can adjust the time constant of the input cap to give the bass response/roll off you require in conjunction with that bias resistor referred to earlier. Remember that large electrolytics have quite a bit of inductance and might roll of the treble hence the popularity of bypassing with a film or better. Also two back to back caps have half the effective capacitance and raise the roll off point one octave.
regards Andrew T.
Electrolytics are a poor choice, as their capacitance is directly related to their working voltage. My personal rule is to never use an electrolytic in AC applications.
The Standard Advice is to use a polypropylene or metallized polypropylene capacitor. Depending on the resistor to ground this could be anywhere in the range of 1µF-10µF. Be sure to calculate the right capacitance for the desired corner frequency.
Some amps are DC-coupled but you have to apply these carefully. If I have an AC-coupled output on the source then I don't also need an AC-coupled input on the amp. But if you build a DC-coupled amp you must think before you connect it to any given source component, whereas you can safely attach an AC-coupled amp to any source.
The Standard Advice is to use a polypropylene or metallized polypropylene capacitor. Depending on the resistor to ground this could be anywhere in the range of 1µF-10µF. Be sure to calculate the right capacitance for the desired corner frequency.
Some amps are DC-coupled but you have to apply these carefully. If I have an AC-coupled output on the source then I don't also need an AC-coupled input on the amp. But if you build a DC-coupled amp you must think before you connect it to any given source component, whereas you can safely attach an AC-coupled amp to any source.
Thanks for the assistance , I am building an amplifier, ETI466 "THE BRUTE" which does 300W into 4 ohm,so as I understand it, the input cap is 2.2uF tantalum 16V I would be better off to replace that with a 2uF metallized polypropylene .
I don,t know how to calculate the right capacitance for the desired corner frequency.
I can tell you that in the circuit diagram the cap goes through two 1k resistors to the base of a BC547, and also there is a 10K resistor after the 1st 1k on the input to ground.
There certainly is a lot more to solid state amplifiers than I thought.
Also as the poly cap is not polarised I guess either way inserted doesnt matter.
I don,t know how to calculate the right capacitance for the desired corner frequency.
I can tell you that in the circuit diagram the cap goes through two 1k resistors to the base of a BC547, and also there is a 10K resistor after the 1st 1k on the input to ground.
There certainly is a lot more to solid state amplifiers than I thought.
Also as the poly cap is not polarised I guess either way inserted doesnt matter.
The formula to use is the standard 1st order formula. C=1/(2x pi) X F X R, where C is the cap value in farads, pi is 3.14159265, F is -3db frequency, and R is resistance (input Z, in this case). You certainly dont want to use a bi-polar or tantalum, if you desire the best in sound quality. Use a film type, polyester (mylar) being the entry level. Polypropylene is better, and probably the most easily found.
Your input filter is a high-pass filter with a capacitor (size as yet undetermined) and 11K resistance to ground. The corner frequency f0 is equal to 1/(2pi*R*C), in Hz. If you wanted to pass frequencies higher than 20Hz, you could use a 1uF cap. If you used 2.2uF you'd pass everything higher than about 7Hz.
2.2uF is the practical limit of polypropylene capacitors. Anything larger than 2.2uF will tend to be physically very large.
I recommend you get a copy of Horowitz & Hill Art of Electronics, and maybe play around with this Java applet that explains high-pass filters: http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/experiment/highpass/hpf.html
2.2uF is the practical limit of polypropylene capacitors. Anything larger than 2.2uF will tend to be physically very large.
I recommend you get a copy of Horowitz & Hill Art of Electronics, and maybe play around with this Java applet that explains high-pass filters: http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/experiment/highpass/hpf.html
In my most recent project, I used a X5R, 100uF/6.3V MLCC ceramic from Nic Components for input DC blocking.
These things do have a bit of a voltage coefficient, but I've tested it on the AP System Two at work and it doesn't look any different from a piece of wire. Wether you can actually *hear* this cap is subject to a few debates/flames/etc though
These things do have a bit of a voltage coefficient, but I've tested it on the AP System Two at work and it doesn't look any different from a piece of wire. Wether you can actually *hear* this cap is subject to a few debates/flames/etc though
This is also the formula I know for Pi Filter.
But "not ordinary" cap value can be seen in tubes, like tube preamp output. Some tube preamp output just use 0.1uF. If the power amp has input impedance of 22khohm, it will form highpass at 72hz.
Is it different formula for RC filter in tube preamp output?
Part Out Tektronix Gear for Caps
Tek put some fine capacitors in its better scopes and amps, still does. Go to a surplus electronics store, pick up an old scope (dead preferably for this purpose) and amp/s, pull the larger caps, and take a listen. Some of those 10uf shielded metal input modules sound very smooth (to my ears).
Tek put some fine capacitors in its better scopes and amps, still does. Go to a surplus electronics store, pick up an old scope (dead preferably for this purpose) and amp/s, pull the larger caps, and take a listen. Some of those 10uf shielded metal input modules sound very smooth (to my ears).
Hi,
swapping the MKT for an electrolytic will probably make quite a difference to the sound, particularly in the mid and treble.
If you change the value to something lower you will change the filter turn-over frequency.
You should check the component values around the input and in your source. They all interact to set the frequency and may (will) affect the bass.
swapping the MKT for an electrolytic will probably make quite a difference to the sound, particularly in the mid and treble.
If you change the value to something lower you will change the filter turn-over frequency.
You should check the component values around the input and in your source. They all interact to set the frequency and may (will) affect the bass.
Triox said:
It is not the end of the world using some elctrolytic input capacitor.
But not an electrolyt of low quality.
I have used non-polar electrolytic 1 uF caps for inputs. With very good sound.
Such caps are usual in Loudspeaker crossovers.
But for best possible performance (read: lowest distortion)
I now wouldnt go for anything but MKP Polyproylene.
They are electrically real topclass! Low losses.
My rule of the thumb is using ~50 uF kohm.
uF x kOhm = 50
----------------------
Input impedance is 10 kohm, I will go for 50/10 = 4.7 uF MKP
Input resistor is 22 kohm, I will go for 50/22k = 2.2 uF
Input resistor is 47 kohm, I will go for 50/47k = 1.0 uF
Input impedance is >47 kohm, I will still go for = 1.0 uF MKP
----------------------
Lineup
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.