I made few amps based on this schematics (call it clone Wink ) because I really liked the sound of the original. Every time i had nasty oscillations what were solved with playing with caps values. This time I have no luck. I think that this is Sziklai output (correct me if I'm wrong) and It's one of disadvantages of that design. I would like to hear about your idea... What can i do to get rid of problems?
BTW, I can't find Sanken's models so on the sim schemo are much slower Onsemi substitute Confused
BTW, I already posted this thread on some other forum but still have no real suggestion. Only thing I know is that I have no enough knowledge to modify this circuit. One solution is to skip Sziklai output and solve oscillations problem, but it will be great if I can keep the sonic
Shot at 2007-08-11
BTW, I can't find Sanken's models so on the sim schemo are much slower Onsemi substitute Confused
BTW, I already posted this thread on some other forum but still have no real suggestion. Only thing I know is that I have no enough knowledge to modify this circuit. One solution is to skip Sziklai output and solve oscillations problem, but it will be great if I can keep the sonic
An externally hosted image should be here but it was not working when we last tested it.
Shot at 2007-08-11
Hi Moby,
Try using a zobel network on your output. All leads and traces need to be short at your output transistors as well.
You may need some C-B caps on your outputs a well and the first transistors are too big and slow. Try using MJE1503x types there for less capacitance and more speed.
-Chris
Try using a zobel network on your output. All leads and traces need to be short at your output transistors as well.
You may need some C-B caps on your outputs a well and the first transistors are too big and slow. Try using MJE1503x types there for less capacitance and more speed.
-Chris
FWIW I've managed to get one of these working. The devices used are 2SA5171 / 2SC1930 (drivers) and Hitachi lateral mosfets for outputs. For best results the drivers need to be as fast as possible and you need to add gate resistors to control the gain roll off of the mosfets. 100 ohms worked well in my design and I'm using 2 devices in parallel per side. Without the gate resistors there was a very nice 300MHz oscillation but with the gate resistors added I can't see anything on the CRO within a 500MHz bandwidth driving resistive or speaker loads.
You may ask does it sound good? The answer is yes, very.
I built an amplifier with this output stage for my next door neighbour and he is over the moon. sounds smooth as silk like a class A but punches like a big AB and runs cold with just 100mA quiescent.
You may ask does it sound good? The answer is yes, very.
I built an amplifier with this output stage for my next door neighbour and he is over the moon. sounds smooth as silk like a class A but punches like a big AB and runs cold with just 100mA quiescent.
The biggest problem with the Sziklai is that making the design work well on paper/simulator is only half of the challenge. The Sziklai output stage is very sensitive to PCB layout... so you may have a design that works great in theory, but oscillates like crazy when you make it for real.
Personally, I prefer the triple darlington
Personally, I prefer the triple darlington
Only thing I don't know is how to treat BJT's since you described Mosfet situation. Yes, this topology sounds kind of A class. Only way I can do is to put cap over output to inv input of op-amp but I have an impression that I will ruined the sound that way. Some local compensation is needed but I don't know what. Oh I forget to mention, this is not just drawing this is real AMP I made that completely but have that issues stated beforeMoby said:
At first glance I thought you were using mosfets but the situation is same for BJT. Try first a small amount of resistance in series with each base of the output devices (ie 10 ohms). Alternatively you could use a single ferrite bead. And yes layout is very important. and it's important regardless of topology to keep supply impedance low. I tend to use an independant RF groundplane for this purpose to avoid stability problems.
If you have some more ideas don't hesistate to suggest
Hi i seem to remeber an old wireless world article which aluded to a problem with this type of output stage and it stated with bipolars that the stage could oscillate if the drivers had an ft of greater than 20 times that of the outputs the circiut would in effect become a phase shift oscillator maybe this might jog someone elses memory iwill try to dig up the article to verify this.
Regards P.Allan.
Regards P.Allan.
Hi,
the CFP has a strong tendency to oscillate mostly when global negative feedback is also used.
As pointed out above it is rather the bandwidth of the main output transistor that should be limited (which provides the local feedback to the driver).
Of course increasing its base-collector capacitance externally is an undesired and very ugly way but probably the most effective.
In this context I wonder if the emitter circuitry of the driver is designed optimally.
the CFP has a strong tendency to oscillate mostly when global negative feedback is also used.
As pointed out above it is rather the bandwidth of the main output transistor that should be limited (which provides the local feedback to the driver).
Of course increasing its base-collector capacitance externally is an undesired and very ugly way but probably the most effective.
In this context I wonder if the emitter circuitry of the driver is designed optimally.
But,hey, that's just a sim. I have to check that in reality.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.