When I raise the bias current of output transistors or bias input transistors more than 3mA oscillation occurs.
Balanced version supposed to be free from picking a noise by input cables, better sound stage, crosstalk performance, can be used with pro audio equipment. Also it's always can be used in SE configuration as well, so why not.
Balanced version supposed to be free from picking a noise by input cables, better sound stage, crosstalk performance, can be used with pro audio equipment. Also it's always can be used in SE configuration as well, so why not.
Member
Joined 2009
Paid Member
Balanced or not, there are many ways to mess up your wiring
What frequency are the oscillations, roughly speaking ?
>> 1MHz - usually points to an unstable power output device (esp. FETs by the way)
few 100's kHz - usually points to a wiring issue, often the ground is contaminated with current from the speaker return or similar and it gets into the signal ground.
few 100's kHz to a few MHz - usually points to unstable feedback loop, mostly inadequate compenstion or poor layout
Balanced or not, there are many ways to mess up your wiring
What frequency are the oscillations, roughly speaking ?
>> 1MHz - usually points to an unstable power output device (esp. FETs by the way)
few 100's kHz - usually points to a wiring issue, often the ground is contaminated with current from the speaker return or similar and it gets into the signal ground.
few 100's kHz to a few MHz - usually points to unstable feedback loop, mostly inadequate compenstion or poor layout
Thanks for your reply!
I don't know the oscillation frequency, but definitely it could be there. I will buy the scope in a month and will measure it.
OPS bias problem is oscillations related as well?
Speaker wires (GND return) is half meter long only. Could it be contaminated than?
The amp is all BJTs made.
Member
Joined 2009
Paid Member
Post#14
In post#19 Cortez suggested a method to test the zeners by flowing current through them using a 2k2 current-limiting resistor - did you do that ?
In post #14 you increased bias current - if this is a change from the published circuit then I suggest you roll-back all such changes and ensure you have the right parts per the original circuit published by Lazy Cat. Make sure it is the schematic from the working version, not some earlier schematic before Lazy Cat had it all tested.
In post #25 you listed out your transistor choices; they are different from the original schematic. Did you check that these replacement transistors have the same pin-out designations as those on the schematic and/or you installed them correctly ?
In post #27 you said that there was a fall in the transistor Hfe. I'm not sure how reliable your measurement meter is, but I wonder if this could indicate some damage to the devices. Do you remember if the Hfe was much lower than when they were new ?
Post#48
But if you build the simpler proven circuit you will still have a great amplifier. Then you can move to next level and build a more power version of it. The fun is learning all the things each of us learned.
To fix the current design will take some time and effort from several forum members - if they can indeed spend that time to help you that would be nice. In this case, I would follow advice given to do a good LTspice simulation, understand the operating points, vary the component values, look at the phase margin of the feedback loop to see if your feedback is stable. This takes a lot of time but is a good step before trying to build such a complex amplifier.
In post#1 you said you replaced the zeners so you know they are good. If the zener is good, the only cause for having a lower voltage drop is because there isn't enough current flowing through them - they are operating below the knee.
In post#19 Cortez suggested a method to test the zeners by flowing current through them using a 2k2 current-limiting resistor - did you do that ?
In post #14 you increased bias current - if this is a change from the published circuit then I suggest you roll-back all such changes and ensure you have the right parts per the original circuit published by Lazy Cat. Make sure it is the schematic from the working version, not some earlier schematic before Lazy Cat had it all tested.
In post #25 you listed out your transistor choices; they are different from the original schematic. Did you check that these replacement transistors have the same pin-out designations as those on the schematic and/or you installed them correctly ?
In post #27 you said that there was a fall in the transistor Hfe. I'm not sure how reliable your measurement meter is, but I wonder if this could indicate some damage to the devices. Do you remember if the Hfe was much lower than when they were new ?
Post#48
OK, this post catches my up to date with the situation - and you won't like my advice on this topic! Stop work on this amplifier, put it away until you are ready to tackle it with confidence. First go back and build a proven SSA schematic. It takes quite a lot of time and some smaller projects to establish the design knowledge, simulation skills and debug experience (and accumulate the test equipment) to make your current journey a pleasurable one ? Ščaslyvoji podoroži ...maybe you are too determined
But if you build the simpler proven circuit you will still have a great amplifier. Then you can move to next level and build a more power version of it. The fun is learning all the things each of us learned.
To fix the current design will take some time and effort from several forum members - if they can indeed spend that time to help you that would be nice. In this case, I would follow advice given to do a good LTspice simulation, understand the operating points, vary the component values, look at the phase margin of the feedback loop to see if your feedback is stable. This takes a lot of time but is a good step before trying to build such a complex amplifier.
Last edited:
Post 1.
Yes, I already know that. I changed 2.2k 2w current set resistors to 750R 5w ones. However I think it should be enough 1k 5w if the amp would be oscillation free.
Post 19.
Yes, zeners are ok. I also using tl431 instead.
Post 14.
I increased input bias current only to 3mA. LC suggested bias current for ssa input 4.5mA as optimal.
As I know, the only tested and working ssa schematics are BIBGT HP, VSSA, TSSA, Shaan SSA schematic, Nico's not published schematic for his private use, LC's CSA, First One not published as well and your full BJT version. Btw, bibgt hp published schematic differs from LC's real built photos of this amp.
My changes to the original schematic are only related to filtering and compensation. They are present in BIBGT HP schematic and Low Z schematic was promised to be updated from bibgt hp but it didn't happen.
Post 25
Yes, I have checked them. Replacement transistors have similar or better specs than in the original schematic.
If there would be a pinout fault the amp won't work at all. It happened at first start when I swapped npn pnp in wrong places.
Post 27.
There were only steady hfe not fall of hfe. But I don't bother with them already, I rematched all my input transistors with 4.5mA bias set with hfe tester schematic I posted earlier and they were measured and working fine the same as newly selected transistors for second channel.
Hfe wasn't much lower. When new hfe raised 10-20 points after a current applied faster under dmm test, used ones - slower.
Post 48
I wouldn't like to stop on this. Simpler version is too weak, waist of time and resources. I prefer at least 100w at 8R load. This schematic can provide such output and I have required components only for this one variant.
I will have test the equipment in end of November.
I expect to overcome the obstacles in building this amp with your's guys help, experience and knowledge of building such an amps. I've learned much from the day I start this project, including pcb layout).
I will very appreciate and will be very happy if those respectable forum members find a free time and put their efforts to help me to built this amplifier. I'll contribute from my side all what I can.
Could you please advise me how to do a proper LTspice simulation, where can i get the parts models?
With kind regards.
Andriy.
Yes, I already know that. I changed 2.2k 2w current set resistors to 750R 5w ones. However I think it should be enough 1k 5w if the amp would be oscillation free.
Post 19.
Yes, zeners are ok. I also using tl431 instead.
Post 14.
I increased input bias current only to 3mA. LC suggested bias current for ssa input 4.5mA as optimal.
As I know, the only tested and working ssa schematics are BIBGT HP, VSSA, TSSA, Shaan SSA schematic, Nico's not published schematic for his private use, LC's CSA, First One not published as well and your full BJT version. Btw, bibgt hp published schematic differs from LC's real built photos of this amp.
My changes to the original schematic are only related to filtering and compensation. They are present in BIBGT HP schematic and Low Z schematic was promised to be updated from bibgt hp but it didn't happen.
Post 25
Yes, I have checked them. Replacement transistors have similar or better specs than in the original schematic.
If there would be a pinout fault the amp won't work at all. It happened at first start when I swapped npn pnp in wrong places.
Post 27.
There were only steady hfe not fall of hfe. But I don't bother with them already, I rematched all my input transistors with 4.5mA bias set with hfe tester schematic I posted earlier and they were measured and working fine the same as newly selected transistors for second channel.
Hfe wasn't much lower. When new hfe raised 10-20 points after a current applied faster under dmm test, used ones - slower.
Post 48
I wouldn't like to stop on this. Simpler version is too weak, waist of time and resources. I prefer at least 100w at 8R load. This schematic can provide such output and I have required components only for this one variant.
I will have test the equipment in end of November.
I expect to overcome the obstacles in building this amp with your's guys help, experience and knowledge of building such an amps. I've learned much from the day I start this project, including pcb layout).
I will very appreciate and will be very happy if those respectable forum members find a free time and put their efforts to help me to built this amplifier. I'll contribute from my side all what I can.
Could you please advise me how to do a proper LTspice simulation, where can i get the parts models?
With kind regards.
Andriy.
Last edited:
Member
Joined 2009
Paid Member
For LTspice you should start with the simplest version of this topology, ideal voltage supplies, ideal current sources etc. with the goal to ensure you can get the basic circuit working in the simulator and can see how and why it works. Then work outwards in terms of complexity, replacing ideal parts with better models etc. I have spent countless hours with LTSpice, looking carefully at how things behave to understand how my circuits are operating and how the choice of currents and things are to be optimized.
There are device models on the forum published in various threads and there are models on some websites. I've collected them over the years, some of my favourite are from andyc, an old forum member, some from Ranchu32 who's still active and there are some on Bob Cordells website.
I'm very busy at work and at home and can not lead you through this project in detail but I'll pop in from time to time
It's good to see you have determination, although I wish you had chosen an amplifier that had already been built and proven, there will be satisfaction of the challenge here.
There are device models on the forum published in various threads and there are models on some websites. I've collected them over the years, some of my favourite are from andyc, an old forum member, some from Ranchu32 who's still active and there are some on Bob Cordells website.
I'm very busy at work and at home and can not lead you through this project in detail but I'll pop in from time to time
It's good to see you have determination, although I wish you had chosen an amplifier that had already been built and proven, there will be satisfaction of the challenge here.
Last edited:
AndriyOL, You simply changed input to differential pairs, but feedback from output remains to input transistor emitor, as with single ended input..But here are two transistors now, and feedback is so undefined, for added transistors it is positive feedback... It must go to -In, not to differental pair emitors. Try it in simulation first.
This is Lazy Cat balanced input design. it was published 6 years ago. As I have been told this is Lender's circuit the way the negative input is connected. Simple Symetrical Amplifier
Could you please show in the schematic what you are offering?
Could you please show in the schematic what you are offering?
Last edited:
Member
Joined 2009
Paid Member
Of course I would send the models to you if I had them but I didn’t use these parts, I tend to stick with what I have.
I don’t know Lazy Cat personally, you can try messaging him yourself.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.