Very lame question..
We have a bipolar supply Opamp module.
The signal through the opamp is audio signal.
The opamp output signal is AC, both negative and positive in regards with the ground.
Could I use electrolytic cap at the output?
If it is 50-100% more voltage rated than the signal peaks?
f.ex. R.C4() on the circuit diagram.
We have a bipolar supply Opamp module.
The signal through the opamp is audio signal.
The opamp output signal is AC, both negative and positive in regards with the ground.
Could I use electrolytic cap at the output?
If it is 50-100% more voltage rated than the signal peaks?
f.ex. R.C4() on the circuit diagram.
Normal polarized aluminium electrolytic capacitors work at reverse voltages up to 1.5 V ... 2 V, so if you are sure the reverse voltage across the capacitor (not to be confused with the output voltage) will stay below that, you can use a normal electrolytic capacitor. In fact that's done quite often in consumer electronics.
Not at all lame, use a bipolar (non-polar) type, rated at more than the supply (25V or higher).
They also have lower distortion than a polar type.
Add a large (100k or so) resistor from the output to ground to avoid noises when connecting.
https://hfc-fs.s3-eu-west-1.amazonaws.com/s3fs-public/nich_es_0.pdf
Last edited:
You can use 2 back to back electrolytics to make a non polar one. Althou the cap in your schematic is very low to be an electrolytic. Whenever use polypropilene if possible
What is the difference between:
- the voltage across the cap and
- the voltage at the output of opamp (lets use the circuit as a ref.)?
Isn't it the same ?
------------
I am literally putting up an Instrumental amp (bass guitar) out of diffetent modules.
The signal peaks could easily be over 2 volts.
There is an EQ module which could amplify certain frequencies even more (+12 db...).
Moreover, I will try to keep the line signal as high as possible to get better signal to noise ratio.
The supply voltage is +-15.
At the very end, just before the power amp, I will attenuate/ limit to avoid cliping and protect the speaker.
- the voltage across the cap and
- the voltage at the output of opamp (lets use the circuit as a ref.)?
Isn't it the same ?
------------
I am literally putting up an Instrumental amp (bass guitar) out of diffetent modules.
The signal peaks could easily be over 2 volts.
There is an EQ module which could amplify certain frequencies even more (+12 db...).
Moreover, I will try to keep the line signal as high as possible to get better signal to noise ratio.
The supply voltage is +-15.
At the very end, just before the power amp, I will attenuate/ limit to avoid cliping and protect the speaker.
The DC voltage across the capacitor is 0VDC, if the op amp has a bipolar power supply.
The DC voltage across the capacitor is half the power supply, if the op amp has a single polarity power supply.
The AC voltage across the capacitor is 0VAC (within the design bandwidth).
The DC voltage at the output of the op amp wrt ground is 0VDC, if the op amp has a bipolar power supply.
The DC voltage at the output of the op amp wrt ground is half the power supply voltage, if the op amp
has a single polarity power supply.
The AC voltage at the output of the op amp wrt ground is the input voltage times the gain.
If the signal frequency is greater than the corner frequency of the capacitor with the load resistance, most of the signal will drop across the load rather than the coupling capacitor.
Example: 10 uF coupling cap, 100 kohm from output to ground, next stage has 10 kohm input resistance:
Total resistance 10 kohm in parallel with 100 kohm -> 9.090909... kohm
Corner frequency 1/(2 pi R C) ~= 1.75 Hz
Example signal frequency: 20 Hz -> only about 8.72 % drops across the capacitor.
For professional use, I would advise you to use a capacitor that can handle at least 52 V, either a bipolar capacitor or a polarized capacitor with its positive side connected to the output connector. The capacitor then sort of protects your circuit if someone should accidentally connect the output to a microphone input with phantom supply.
Example: 10 uF coupling cap, 100 kohm from output to ground, next stage has 10 kohm input resistance:
Total resistance 10 kohm in parallel with 100 kohm -> 9.090909... kohm
Corner frequency 1/(2 pi R C) ~= 1.75 Hz
Example signal frequency: 20 Hz -> only about 8.72 % drops across the capacitor.
For professional use, I would advise you to use a capacitor that can handle at least 52 V, either a bipolar capacitor or a polarized capacitor with its positive side connected to the output connector. The capacitor then sort of protects your circuit if someone should accidentally connect the output to a microphone input with phantom supply.
Thank you for all replies and the corner frequency/voltage across cap calc !!!
Anyway, even if I could (eventually) use a single polar cap, it behaves differently when subjected to reverse voltage - there is a different current flow.
So, the distortions would be significantly more.
I will use bipolar electrolytics or a pair of reverse polarity in series.
---
Still another question.
In general, when we have dual rail opamp, the DC offset at the output should be 0 volts.
In reality, there would be some DC offset at the output.
Most of the coupling cap slots left on the PCB's (as in the circuit above), are to DE-couple the opamp stages from one another.
Should I worry about the opamp's DC offset at the output (dual rail supply) ?
What if we have, let's say 1 mV DC offset at the output?
- If next stage is unity gain (lots of opamps there are just buffers)
- If the next stage has some gain? (If the DC offset at the input is a problem and if the problem is dependent on the voltage gain).
In both cases I am not concern about adding up DC offset - it might be filtered out at just one place.
I am concerned about S/N ratio.
The DC offset would cause some current flow - f.ex. through the input to ground resistors.
Wouldn't there be a current flow noise? Or thermal/other noise?
Anyway, even if I could (eventually) use a single polar cap, it behaves differently when subjected to reverse voltage - there is a different current flow.
So, the distortions would be significantly more.
I will use bipolar electrolytics or a pair of reverse polarity in series.
---
Still another question.
In general, when we have dual rail opamp, the DC offset at the output should be 0 volts.
In reality, there would be some DC offset at the output.
Most of the coupling cap slots left on the PCB's (as in the circuit above), are to DE-couple the opamp stages from one another.
Should I worry about the opamp's DC offset at the output (dual rail supply) ?
What if we have, let's say 1 mV DC offset at the output?
- If next stage is unity gain (lots of opamps there are just buffers)
- If the next stage has some gain? (If the DC offset at the input is a problem and if the problem is dependent on the voltage gain).
In both cases I am not concern about adding up DC offset - it might be filtered out at just one place.
I am concerned about S/N ratio.
The DC offset would cause some current flow - f.ex. through the input to ground resistors.
Wouldn't there be a current flow noise? Or thermal/other noise?
Yes, that's why you should use a non-polar cap. I prefer 10uF polypropylene, but the Nichicon Muse could be an alternative. Or you can use a servo to null out the offset.
stop/block DC offset into the next stage. i f you use TL071, (or you make a null ckt or DC servo) you can null DC offset with a pot or the servo is automatic and not need a DC block cap. You can use a dual opamp, make one the gain stage and the other the servo.
design trade offs, use a DC blocking cap or have a null DC offset means, cap is usually cheaper so that is what many do. a lot of times the next stage has a DC block cap too, so it can be redundant.
design trade offs, use a DC blocking cap or have a null DC offset means, cap is usually cheaper so that is what many do. a lot of times the next stage has a DC block cap too, so it can be redundant.
Last edited:
First what you have to understand is that there are DC errors and AC errors, DC errors are for example DC offset. So regarding DC errors:
Even if you use a dual rail supply you will have a voltage offset at the output, now, the magnitude of this offset depends on:
- The opamp you are using
- The resistors you are using
- The gain
The opamp will have an offset voltage, bias current and offset current, this will influence the amount of DC you will have at the output, bias and offset currents are really only relevant when you are not using low value resistors, for higher value resistors a FET opamp will be better because it will have lower bias and offset currents, however most FET opamps will have higher offset voltages than bipolars, so pick your poison.
A DC offset will reduce headroom but not precisely S/N ratio, the thing you are mentioning about noise has to do with the following:
- Opamp voltage and current noise densities
- Resistor Thermal noise
- Source resistance (wihich has to do with the voltage and noise densities)
- Gain
Noise is in itself a very vast topic, and I will not go any further.. but I hope this helps!
I suggest to add a 1KΩ resitsor between output and GND at the rightest side of such cap, in this way the cap has no voltage difference to ground, avoiding pops and other noises when plugging this pre to other amp already powered on.
1 kohm is a bit low for an op-amp designed to drive 2 kohm or more. Rayma's 100 kohm (see post 3) sounds more reasonable.
If you worry about capacitor distortion, this is roughly the order from very good to very bad:
1. polypropylene (or polystyrene or class 1 ceramic, but those probably have too low values)
2. polyethylene terephtalate (MKT, polyester)
3. bipolar electrolytic
4. two polar electrolytics in antiseries
5. single polar electrolytic
6. tantalum (can only handle about 0.5 V in reverse, by the way)
7. class 2 ceramic
If you must use a cap in the output, besides using a big "non-polar" cap, you can use a T network with a high value resistor (~1meg) to one of the rails or a divider to keep the bias down for 6.3V caps. Then a small load resistor (~100K) to remove any leakage.
But I would simply NOT use a cap in the output.
But I would simply NOT use a cap in the output.
If possible, no cap is certainly preferable. In that case, increase the series output resistor
to 1k for short circuit protection.
I would use a bi polar electrolytic if possible.
If the input can be different items then some might be positive. negative or ground.
A cheap solution is to just put a couple of polar electrolytics back to back.
The DC threshold in my experience for electroyltics is about 1.5 volts, any more and the cap becomes a short circuit.
If the input can be different items then some might be positive. negative or ground.
A cheap solution is to just put a couple of polar electrolytics back to back.
The DC threshold in my experience for electroyltics is about 1.5 volts, any more and the cap becomes a short circuit.
- Status
- Not open for further replies.