oscillation in a preamp
- Thread starter Gustav
- Start date
Help Support GroupDIY Audio Forum:
kruz
Well-known member
Could be a grounding issue, check that you have connected all the grounds correctly
owel
Well-known member
Is this opamp based?
Check those bypass capacitors.
This is what my oscillation look like without the bypass capacitors. So I added some bypass capacitors...
and the oscillation was gone.
http://www.fivefish.net/diy/Altec1599A/page6.htm
Check those bypass capacitors.
This is what my oscillation look like without the bypass capacitors. So I added some bypass capacitors...
and the oscillation was gone.
http://www.fivefish.net/diy/Altec1599A/page6.htm
[quote author="owel"]Is this opamp based?
Check those bypass capacitors.
[/quote]
Which bypass capacitors?
Do you mean the ones in the feedback network for reducing gain at high frequencies.
I have an audible oscillation problem with a preamp, but it only happens when I try to drive very low-impedance loads (eg. 50 Ohms). Is this normal? Any solutions?
Check those bypass capacitors.
[/quote]
Which bypass capacitors?
Do you mean the ones in the feedback network for reducing gain at high frequencies.
I have an audible oscillation problem with a preamp, but it only happens when I try to drive very low-impedance loads (eg. 50 Ohms). Is this normal? Any solutions?
Bo Hansen
Well-known member
Some op-amp oscillations hints,
First of all, check if the op-amp only oscillate when you connect the scope probe direct on the output.
Some scope cable/probe that not have any "X10" 10:1 attenuation can stress the op-amp to oscillation.
In this case you can connect a 1 k resistor in series with the scope probe.
If it is a new construction that you "pilot testing" you can try following as normal proceedings:
Small capacitor parallel with the feedback resistor. (for 100 k, approx. 10pF, then you can count up/down with different resistors)
De-coupling 0,1 capasitors direct to op-amp voltage supply pins to ground, especially if you use fast op-amps with high slew rate.
A serial resistor around 47 to 100 ohms direct on the op-amp output.
If the op-amp drive lo-Z loads as a transformer, it is not good with a serial resistor, it is better with a inductor around some uH. (resistor with copper wire around the body)
If you use the single NE-5534, connect the usual pin 5 and 8 compensation capacitor approx. 10 pF.
--Bo
First of all, check if the op-amp only oscillate when you connect the scope probe direct on the output.
Some scope cable/probe that not have any "X10" 10:1 attenuation can stress the op-amp to oscillation.
In this case you can connect a 1 k resistor in series with the scope probe.
If it is a new construction that you "pilot testing" you can try following as normal proceedings:
Small capacitor parallel with the feedback resistor. (for 100 k, approx. 10pF, then you can count up/down with different resistors)
De-coupling 0,1 capasitors direct to op-amp voltage supply pins to ground, especially if you use fast op-amps with high slew rate.
A serial resistor around 47 to 100 ohms direct on the op-amp output.
If the op-amp drive lo-Z loads as a transformer, it is not good with a serial resistor, it is better with a inductor around some uH. (resistor with copper wire around the body)
If you use the single NE-5534, connect the usual pin 5 and 8 compensation capacitor approx. 10 pF.
--Bo
Samuel Groner
Well-known member
No, driving 50 ohm is not normal. An IC opamp may go into current limit. IC based output? Schemo?
What do you mean by "audible oscillation problem"? Do you have a scope and you see and hear the osciallation (or at better say the side effects) or you don't have a scope and you believe the problem you're hearing is caused by instability?
Samuel
This is a commercial product, so its not a DIY of my own Im trying to get to work.
It has audible oscillation - I didnt measure it since I dont have a scope avaliable.
I was gonna pop the lid next week to see if I could figure out what might be wrong, and just wanted a pointer or two about where to start , so the opamps are a good place to start?
Gustav
It has audible oscillation - I didnt measure it since I dont have a scope avaliable.
I was gonna pop the lid next week to see if I could figure out what might be wrong, and just wanted a pointer or two about where to start , so the opamps are a good place to start?
Gustav
Samuel Groner
Well-known member
Or perhaps a broken ground connection?
Sorry but I really don't understand what "audible oscillation" means; if an amp goes mad it very likely osciallates well above 100 kHz. Of course audio gets noisy and distorted because of this but unless you've seen the oscillation on a scope you are not sure if not another problem may have caused this.
If you really get an audible sinwave out of your equipment somthing is really wrong unless it is supposed to do so. Probably a broken connection or short-circuit.
Samuel
Sorry but I really don't understand what "audible oscillation" means; if an amp goes mad it very likely osciallates well above 100 kHz. Of course audio gets noisy and distorted because of this but unless you've seen the oscillation on a scope you are not sure if not another problem may have caused this.
If you really get an audible sinwave out of your equipment somthing is really wrong unless it is supposed to do so. Probably a broken connection or short-circuit.
Samuel
[quote author="Samuel Groner"]Or perhaps a broken ground connection?
Sorry but I really don't understand what "audible oscillation" means; if an amp goes mad it very likely osciallates well above 100 kHz. Of course audio gets noisy and distorted because of this but unless you've seen the oscillation on a scope you are not sure if not another problem may have caused this.
If you really get an audible sinwave out of your equipment somthing is really wrong unless it is supposed to do so. Probably a broken connection or short-circuit.
Samuel[/quote]
Im probably using the wrong word.
Im getting a pretty high freq sweeping tone (this is what I called oscillation), crosstalk between channels (two channels), and after a while - distortion to boot.
Ill record a little of it if possible, but it has a habit of coming and going. The distortion doesnt kick in till after the units been on for a while.
Gustav
Sorry but I really don't understand what "audible oscillation" means; if an amp goes mad it very likely osciallates well above 100 kHz. Of course audio gets noisy and distorted because of this but unless you've seen the oscillation on a scope you are not sure if not another problem may have caused this.
If you really get an audible sinwave out of your equipment somthing is really wrong unless it is supposed to do so. Probably a broken connection or short-circuit.
Samuel[/quote]
Im probably using the wrong word.
Im getting a pretty high freq sweeping tone (this is what I called oscillation), crosstalk between channels (two channels), and after a while - distortion to boot.
Ill record a little of it if possible, but it has a habit of coming and going. The distortion doesnt kick in till after the units been on for a while.
Gustav
Bo Hansen
Well-known member
Gustav,
Tell us a bit more about the unit you have problem with, type of gear, brand, etc.
For example, some gear can oscillate with open input, if you have maximum gain or treble boost.
There are also gear with electronic balansed output, that can start to oscillate if you connect one of the phase to ground to get a un-balanced output.
--Bo
Tell us a bit more about the unit you have problem with, type of gear, brand, etc.
For example, some gear can oscillate with open input, if you have maximum gain or treble boost.
There are also gear with electronic balansed output, that can start to oscillate if you connect one of the phase to ground to get a un-balanced output.
--Bo
[quote author="Samuel Groner"]
What do you mean by "audible oscillation problem"? Do you have a scope and you see and hear the osciallation (or at better say the side effects) or you don't have a scope and you believe the problem you're hearing is caused by instability?
Samuel[/quote]
[Gustav, sorry for hijacking the thread - maybe my problem can help you though...]
Samuel, by "audible oscillation" I mean that this thing sounds like a theramin when loaded with a very low load! It is in the audio range - eg. maybe 5K to 20K and the pitch of the tone changes when you change the gain.
How could Current limiting cause this problem?
No, driving 50 ohm is not normal. An IC opamp may go into current limit. IC based output? Schemo?
What do you mean by "audible oscillation problem"? Do you have a scope and you see and hear the osciallation (or at better say the side effects) or you don't have a scope and you believe the problem you're hearing is caused by instability?
Samuel[/quote]
[Gustav, sorry for hijacking the thread - maybe my problem can help you though...]
Samuel, by "audible oscillation" I mean that this thing sounds like a theramin when loaded with a very low load! It is in the audio range - eg. maybe 5K to 20K and the pitch of the tone changes when you change the gain.
How could Current limiting cause this problem?
Samuel Groner
Well-known member
There are two strong mechanisms working against each other--the feedback wants to supply more current and the output protection wants to supply less current. That makes the system somewhat unstable.
I guess your 50 ohm load are headphones? Then it is more likely the reactive load that causes the instability. Are there small resistors (around 50 ohm) in series with the output? You could try adding some externally and see what happens.
I was wrong when saying that HF-oscillation is not audible because usually there are some frequency products folded down into the audible range--that's probably what you are hearing. But the main oscillation is outside our hearing range, of course.
Samuel
[quote author="Samuel Groner"]
I was wrong when saying that HF-oscillation is not audible because usually there are some frequency products folded down into the audible range--that's probably what you are hearing. But the main oscillation is outside our hearing range, of course.
[/quote]
Ahh, that makes sense!
I also see your point about the audible oscillation "folding down" I wonder if this is simply a harmonic or an alias product....
Thanks!
There are two strong mechanisms working against each other--the feedback wants to supply more current and the output protection wants to supply less current. That makes the system somewhat unstable.
I was wrong when saying that HF-oscillation is not audible because usually there are some frequency products folded down into the audible range--that's probably what you are hearing. But the main oscillation is outside our hearing range, of course.
[/quote]
Ahh, that makes sense!
I also see your point about the audible oscillation "folding down" I wonder if this is simply a harmonic or an alias product....
Thanks!
Samuel Groner
Well-known member
Well harmonics are always of higher frequency than the fundamental and aliasing happens only in time-discrete systems (e.g. digital).
I don't pretend to be able to give you a full explanation of what happens when an amp is oscillating but I see mainly two ways to get a low frequency: 1. intermodulation distortion (if two sinwaves pass a nonlinear amplifier (and an oscillating one is surely not very linear) the output may contain low frequencies) and 2. demodulation (by BJT emitter diodes, similar to a radio).
Samuel
[quote author="Samuel Groner"]
...I see mainly two ways to get a low frequency: 1. intermodulation distortion (if two sinwaves pass a nonlinear amplifier (and an oscillating one is surely not very linear) the output may contain low frequencies) and 2. demodulation (by BJT emitter diodes, similar to a radio).
[/quote]
Good point - I could hear the oscillation sound before digital conversion therefore it is not aliasing (this would probably be lp filtered anyway).
I think you are probably right about the interodulation distortion (I did not know this term and for some reason thought it was also called a "harmonic" which is wrong!)
Thanks again,
Roddy
Well harmonics are always of higher frequency than the fundamental and aliasing happens only in time-discrete systems (e.g. digital).
...I see mainly two ways to get a low frequency: 1. intermodulation distortion (if two sinwaves pass a nonlinear amplifier (and an oscillating one is surely not very linear) the output may contain low frequencies) and 2. demodulation (by BJT emitter diodes, similar to a radio).
[/quote]
Good point - I could hear the oscillation sound before digital conversion therefore it is not aliasing (this would probably be lp filtered anyway).
I think you are probably right about the interodulation distortion (I did not know this term and for some reason thought it was also called a "harmonic" which is wrong!)
Thanks again,
Roddy
dale116dot7
Well-known member
If one op-amp is oscillating then another might be trying to - either from coupling on the power supply rails, or just proximity. Or you could have the voltage regulator oscillating too. If you have two amps or anything like that going at once, the products will mix and will appear in the audio band. Or sometimes an op-amp tries to oscillate at two different frequencies, and as it clips and mixes the two signals together, you get exactly what you describe - a theremin sound.
An amplifier driving a capacitive load may do this - or a load that's a lower impedance than it was designed for, especially through a transformer.
-Dale
An amplifier driving a capacitive load may do this - or a load that's a lower impedance than it was designed for, especially through a transformer.
-Dale
henk
Well-known member
My first question is , is that ic on a socket?
If yes , what kind of socket?
If it is a cheap socket , replace socket for a real one!!!!!
If yes , what kind of socket?
If it is a cheap socket , replace socket for a real one!!!!!
Im not sure, but I think Ive made somewhat of a mistake.
I encounterred this problem with my old rhodes. (MK1 73), and didnt really check with much else.
Tried mic lines today, didnt seem to cause the problem. Then a bass, just gave my usual noise - plugged in the rhodes again, and presto - Chinese radio at specific gain settings, its almost like you can tune in the frequency with the gain button on the DI of the pre.
Its probably been RFI all along...Ive never been able to plug instruments in directly at my control room because of noise/rumble, which I assumed was electrical, but the rhodes impendence must be loading it so I can hear the actual problem more than on other sources..? (an unqualified guess)
This is bad. Anyone got hints on where to start?
Gustav
I encounterred this problem with my old rhodes. (MK1 73), and didnt really check with much else.
Tried mic lines today, didnt seem to cause the problem. Then a bass, just gave my usual noise - plugged in the rhodes again, and presto - Chinese radio at specific gain settings, its almost like you can tune in the frequency with the gain button on the DI of the pre.
Its probably been RFI all along...Ive never been able to plug instruments in directly at my control room because of noise/rumble, which I assumed was electrical, but the rhodes impendence must be loading it so I can hear the actual problem more than on other sources..? (an unqualified guess)
This is bad. Anyone got hints on where to start?
Gustav
