Oscillation - Page 3

davidallancole · 18th February 2006, 01:56 AM

Remember I said before I put a capacitor from the output to ground. I meant the speaker output which you probably all guessed. I didn't notice for a bit that I still had oscillations until I put a test signal into it and I seen a waveform that looked like AM. So I looked closer and found I had only reduced the amplitude by a lot as well as the frequency. I decided to play capacitor roulette for the last quarter of the day and found that if I put the 0.1uF from the op-amp output to ground then the oscillations quit completely. It was time to go home when I figured this out so I will go to work tomorrow and run some tests on it with Pantera, Metallica and some speakers.

davidallancole · 18th February 2006, 02:37 AM

Heres what I figure the final amp will be tomorrow.

clem_o · 18th February 2006, 03:27 AM

Unless you are using an exotic op-amp that loves driving highly capacitive loads, I think that amp is going to sound really slow!

Suggest you try removing the 0.1uF from where it is now, and instead put it at the output as before, but with a 4.7 or 10ohm resistor first in series, as Chris (anatech) has suggested.

Then, add a small amount of capacitance across (in parallel with) the the 23K resistor (R14). 47pF or 22pF would be what I'd try first...

You could also put a capacitor to ground at the input (i.e. parallel with R62), just in case the AM signals are sneaking in at the input.

Cheers!

davidallancole · 18th February 2006, 02:00 PM

Good point. I ran it in simulatioion and the op-amp had to put out 10 ma or so at 1 kHz and then started increasing rapidly after that. One of the main problems is I don't have to many parts, like any resistors below 100 ohms. I have a 18 pF capacitor, so I will try that in the feedback path and also see if I can't get a better connection with the capacitor from speaker output to ground.

anatech · 18th February 2006, 03:19 PM

Hi Clem,
I was about to suggest the same treatment. Took the words right out of my mouth!

David, in general, amplifiers and op amps do not like to see a pure capacitance to ground (or many other places). It's always good practice to scale the resistor in series with the cap to the impedance of the circuit location. There are exeptions to this.

-Chris

Eva · 18th February 2006, 03:35 PM

I would put a resistor (100ohm to 1K) between the output of the op-amp and the input of the darlington, and a small capacitor (10 to 100pF) from the output of the op-amp to the non-inverting input. This will create a split-feedback system and will prevent the op-amp from trying to drive the darlington at RF and oscillate.

Standard op-amps may be expected to drive a small-signal emitter-follower without additional compensation, but a power darlington usually going to produce excessive phase shift when approaching its Ft.

davidallancole · 18th February 2006, 04:11 PM

If I remember correctly, the two requirements for oscillation is a phase shift greater than 180 degrees and a gain greater than 1, is this correct? So as the frequency gets higher, the phase shift gets greater and the gain is above unity in this circuit so it oscillates? Is the phase shift produced because of the capacitance between the base and the emitter of the darlington?

Putting a capacitor from the output of the op-amp to the inverting input will reduce the gain below unity at these higher frequencys and putting a resistor from the output of the op-amp to the base of the transistor in conjunction with the capacitance between the base/emitter of the transistor, they will act as a lowpass filter, helping to attenuate the higher frequency signals.

Is this all correct?

Eva · 18th February 2006, 04:30 PM

Yes. As frequency increases, darlington gain is lower (until unity). This means that more and more output current has to be provided by the op-amp and less is taken from the collector. This causes more voltage drop across the series base resistor, and more local op-amp negative feedback through the capacitor that I mentioned and through its own internal compensation.

davidallancole · 18th February 2006, 06:18 PM

Well, here is the final amp. I didn't have a low enough capacitor to through in from the op-amp output to the inverting input, so I just put in what I had. The 180p cap seems to limit the response to around 40 kHz so I will have to try and get a smaller cap. I put the 10 ohm and .1u cap on the breadboard first, and this dropped the oscillations voltage to around 200 mV. Then I put the cap in across the output to the inverting input, and it totally got rid of all oscillation.

Thank you for all of your help everyone.

anatech · 18th February 2006, 06:21 PM

Hi,
Circuit? Glad you got it working,

-Chris

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
807 AB2 oscillation maybe?	Craig405	Tubes / Valves	27	25th March 2008 10:01 PM
CDP Oscillation	emmasdad	Digital Source	0	4th January 2006 03:54 PM
Oscillation	raveenvijendren	Solid State	15	18th December 2005 11:16 AM
oscillation??	wintermute	Solid State	8	13th July 2005 01:45 PM
Oscillation	udip	Solid State	5	31st January 2005 06:43 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

18th February 2006, 01:56 AM	#21
davidallancole diyAudio Member Join Date: Nov 2005	Remember I said before I put a capacitor from the output to ground. I meant the speaker output which you probably all guessed. I didn't notice for a bit that I still had oscillations until I put a test signal into it and I seen a waveform that looked like AM. So I looked closer and found I had only reduced the amplitude by a lot as well as the frequency. I decided to play capacitor roulette for the last quarter of the day and found that if I put the 0.1uF from the op-amp output to ground then the oscillations quit completely. It was time to go home when I figured this out so I will go to work tomorrow and run some tests on it with Pantera, Metallica and some speakers.

18th February 2006, 03:27 AM	#23
clem_o diyAudio Member Join Date: Jun 2005 Location: Manila	Unless you are using an exotic op-amp that loves driving highly capacitive loads, I think that amp is going to sound really slow! Suggest you try removing the 0.1uF from where it is now, and instead put it at the output as before, but with a 4.7 or 10ohm resistor first in series, as Chris (anatech) has suggested. Then, add a small amount of capacitance across (in parallel with) the the 23K resistor (R14). 47pF or 22pF would be what I'd try first... You could also put a capacitor to ground at the input (i.e. parallel with R62), just in case the AM signals are sneaking in at the input. Cheers!

18th February 2006, 02:00 PM	#24
davidallancole diyAudio Member Join Date: Nov 2005	Good point. I ran it in simulatioion and the op-amp had to put out 10 ma or so at 1 kHz and then started increasing rapidly after that. One of the main problems is I don't have to many parts, like any resistors below 100 ohms. I have a 18 pF capacitor, so I will try that in the feedback path and also see if I can't get a better connection with the capacitor from speaker output to ground.

18th February 2006, 03:19 PM	#25
anatech diyAudio Moderator Join Date: Jun 2004 Location: Georgetown, On	Hi Clem, I was about to suggest the same treatment. Took the words right out of my mouth! David, in general, amplifiers and op amps do not like to see a pure capacitance to ground (or many other places). It's always good practice to scale the resistor in series with the cap to the impedance of the circuit location. There are exeptions to this. -Chris

18th February 2006, 03:35 PM	#26
Eva diyAudio Member Join Date: Oct 2003 Location: Near the sea	I would put a resistor (100ohm to 1K) between the output of the op-amp and the input of the darlington, and a small capacitor (10 to 100pF) from the output of the op-amp to the non-inverting input. This will create a split-feedback system and will prevent the op-amp from trying to drive the darlington at RF and oscillate. Standard op-amps may be expected to drive a small-signal emitter-follower without additional compensation, but a power darlington usually going to produce excessive phase shift when approaching its Ft.

18th February 2006, 04:11 PM	#27
davidallancole diyAudio Member Join Date: Nov 2005	If I remember correctly, the two requirements for oscillation is a phase shift greater than 180 degrees and a gain greater than 1, is this correct? So as the frequency gets higher, the phase shift gets greater and the gain is above unity in this circuit so it oscillates? Is the phase shift produced because of the capacitance between the base and the emitter of the darlington? Putting a capacitor from the output of the op-amp to the inverting input will reduce the gain below unity at these higher frequencys and putting a resistor from the output of the op-amp to the base of the transistor in conjunction with the capacitance between the base/emitter of the transistor, they will act as a lowpass filter, helping to attenuate the higher frequency signals. Is this all correct?

18th February 2006, 04:30 PM	#28
Eva diyAudio Member Join Date: Oct 2003 Location: Near the sea	Yes. As frequency increases, darlington gain is lower (until unity). This means that more and more output current has to be provided by the op-amp and less is taken from the collector. This causes more voltage drop across the series base resistor, and more local op-amp negative feedback through the capacitor that I mentioned and through its own internal compensation.

18th February 2006, 06:21 PM	#30
anatech diyAudio Moderator Join Date: Jun 2004 Location: Georgetown, On	Hi, Circuit? Glad you got it working, -Chris

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)