Oscillation in hybrid amp, how to solve?

luixssj3 · 11th December 2010, 04:04 PM

Hi to all!

I'm an electronic eng. student and for the "tesi" (the final exam) I made an hybrid audio amplifier, here is the schematic:

http://img703.imageshack.us/img703/8273/circuitom2.gif

http://img651.imageshack.us/img651/5...entazione1.gif

the 10k resistor in VT1 and VT2 now is 2.2k to drop less voltage.

the problem is that this amp work properly without the capacitor in parallel with the feedback resistor, if I add this capacitor the amp start to oscillate at about 500kHz...no matter on the value of the capacitor, 10pF give the same frequency of 100pF.

Measuring the frequency responce it give GBW at about 100kHz!!! and above 30kHz the responce is not flat and goes up and down!!

What can I try to tune this amp?

Thanks to all and sorry for my bad English!

DF96 · 11th December 2010, 04:15 PM

If it is OK without the capacitor, then omit the capacitor! This capacitor has two roles:
1. it sets the HF roll-off on closd-loop gain,
2. it gives some phase advance on the feedback, which offsets phase lags elsewhere and so aids stability.

However, it sounds marginal without the capacitor as you have HF wiggles. Your amplifier forward path may have too high a bandwidth, with cascode input and cathode followers. The natural roll-off provided by Miller effect has been designed away, but you actually need some of this to cope with the phase shift created by the output transformer HF resonance.

As you are an EE student you will know about servo systems and stability (see Bode, Nyquist). The same rules apply to negative feedback.

luixssj3 · 12th December 2010, 12:26 PM

Hi DF96, thank in advance for your reply!

Yess I know theory of servo sistems but I need to measure the phase at GBW and with Lissajous figure is impossible cause the wave at such high frequency (ca. 100kHz) is not sinusoidal and have an awful shape... tomorrow I make some pictures... here you can find a document with measures I do misurazioni.rar

Another thing, with electro harmonix EL34 the amp reach only 34Wrms @0.8% THD before wild clipping, with JJ EL34 the amp reach about 43Wrms @0.9% before the same wild clipping.... is this a sign that internal impedance of this valve higher than declared?

DF96 · 12th December 2010, 02:15 PM

Instead of Lissajous try using a dual-trace scope, looking at input and output. You don't need an accurate measurement, estimates are good enough. If you can see 90 deg, 180 deg points and measure gain then you can do a rough plot and see where the trouble lies. A smaller amplitude might give a cleaner waveform.

Differences between valves could just be due to biasing - only 1dB difference between 34W and 43W anyway.

11th December 2010, 04:04 PM	#1
luixssj3 diyAudio Member Join Date: Aug 2004 Location: sulmona	Oscillation in hybrid amp, how to solve? Hi to all! I'm an electronic eng. student and for the "tesi" (the final exam) I made an hybrid audio amplifier, here is the schematic: http://img703.imageshack.us/img703/8273/circuitom2.gif http://img651.imageshack.us/img651/5...entazione1.gif the 10k resistor in VT1 and VT2 now is 2.2k to drop less voltage. the problem is that this amp work properly without the capacitor in parallel with the feedback resistor, if I add this capacitor the amp start to oscillate at about 500kHz...no matter on the value of the capacitor, 10pF give the same frequency of 100pF. Measuring the frequency responce it give GBW at about 100kHz!!! and above 30kHz the responce is not flat and goes up and down!! What can I try to tune this amp? Thanks to all and sorry for my bad English!

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
Input oscillation (Pin2) OPA2604 & OP275	grievousangel	Solid State	11	5th August 2008 06:36 AM
Amp oscillation	kubeek	Solid State	22	16th March 2007 04:50 PM
How to solve noise/storing/on car Tube Amp ? any Schematic ?	calico88	Tubes / Valves	3	13th January 2007 05:02 AM
T-Amp 5Hz oscillation	rudiloos	Class D	6	24th May 2006 08:27 PM
Hybrid amp phase shift oscillation?	Duo	Solid State	5	21st October 2004 08:16 AM

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

11th December 2010, 04:15 PM	#2
DF96 diyAudio Member Join Date: May 2007	If it is OK without the capacitor, then omit the capacitor! This capacitor has two roles: 1. it sets the HF roll-off on closd-loop gain, 2. it gives some phase advance on the feedback, which offsets phase lags elsewhere and so aids stability. However, it sounds marginal without the capacitor as you have HF wiggles. Your amplifier forward path may have too high a bandwidth, with cascode input and cathode followers. The natural roll-off provided by Miller effect has been designed away, but you actually need some of this to cope with the phase shift created by the output transformer HF resonance. As you are an EE student you will know about servo systems and stability (see Bode, Nyquist). The same rules apply to negative feedback.

12th December 2010, 12:26 PM	#3
luixssj3 diyAudio Member Join Date: Aug 2004 Location: sulmona	Hi DF96, thank in advance for your reply! Yess I know theory of servo sistems but I need to measure the phase at GBW and with Lissajous figure is impossible cause the wave at such high frequency (ca. 100kHz) is not sinusoidal and have an awful shape... tomorrow I make some pictures... here you can find a document with measures I do misurazioni.rar Another thing, with electro harmonix EL34 the amp reach only 34Wrms @0.8% THD before wild clipping, with JJ EL34 the amp reach about 43Wrms @0.9% before the same wild clipping.... is this a sign that internal impedance of this valve higher than declared?

12th December 2010, 02:15 PM	#4
DF96 diyAudio Member Join Date: May 2007	Instead of Lissajous try using a dual-trace scope, looking at input and output. You don't need an accurate measurement, estimates are good enough. If you can see 90 deg, 180 deg points and measure gain then you can do a rough plot and see where the trouble lies. A smaller amplitude might give a cleaner waveform. Differences between valves could just be due to biasing - only 1dB difference between 34W and 43W anyway.