We should definetaly need one SSA/TSSA/CSA impressions thread.LC, you should also think opening LC Facebook fanpage
Absolutely, utterly no way! FB is the scourge of those who do not want synthetic, make believe 'friendship'
!!A new thread here on DIYaudio would help keep the technical apart from the pure listening experience. It would also be a good idea to break this thread into 2 sections, 1 covering the simplest version and the other covering the more full blown versions.
While Waiting For The Smart Guys
I have been long waiting for good high power amp utilizing Hitachi Latfet, and I have been hoping for the circuit to appear from SSA/TSSA thread.
But from a few circuits utilizing the latfet output, they are not very interesting because:
1) Low power (It seems that in TSSA topology the latfet cannot be operated at high power, and I couldn't solve the stability issues, but what the heck, IRF device works wonder. Ain't hexfets are so good at class-A!).
2) Not sufficient mosfet drive (Nico's version is the best among others but I'm not comfortable with simulated performance).
3) Not so good simulated performance
4) Not stable enough
Currently LC is developing the CSA, I think it is not latfet output but the old trusted BIGBT output. So I cannot wait any longer, I have to take the big riskD) and design a latfet output SSA/TSSA
The reference/benchmark is a Goldmund Mimesis3 clone. So the SSA will also use 2 pairs of Latfet, powered from the same power supply and set with equal gain.
From preliminary design today, the SSA and Mimesis3 has equal simulated performance except:
1) SSA is much better at distortion. High order distortions have been reduced at the price of 2nd order (and overall) distortion, by increasing the output bias to around 400mA per mosfet.
2) Mimesis3 is much better at stability.
3) Mimesis3 is better at high frequency performance (especially my clone of the original). The best I have seen actually.
So for now I will take a break (have been sitting for hours) and post the distortion spectrum (Fourier chart) of Mimesis3 and the SSA:
I have been long waiting for good high power amp utilizing Hitachi Latfet, and I have been hoping for the circuit to appear from SSA/TSSA thread.
But from a few circuits utilizing the latfet output, they are not very interesting because:
1) Low power (It seems that in TSSA topology the latfet cannot be operated at high power, and I couldn't solve the stability issues, but what the heck, IRF device works wonder. Ain't hexfets are so good at class-A!).
2) Not sufficient mosfet drive (Nico's version is the best among others but I'm not comfortable with simulated performance).
3) Not so good simulated performance
4) Not stable enough
Currently LC is developing the CSA, I think it is not latfet output but the old trusted BIGBT output. So I cannot wait any longer, I have to take the big risk
D) and design a latfet output SSA/TSSAThe reference/benchmark is a Goldmund Mimesis3 clone. So the SSA will also use 2 pairs of Latfet, powered from the same power supply and set with equal gain.
From preliminary design today, the SSA and Mimesis3 has equal simulated performance except:
1) SSA is much better at distortion. High order distortions have been reduced at the price of 2nd order (and overall) distortion, by increasing the output bias to around 400mA per mosfet.
2) Mimesis3 is much better at stability.
3) Mimesis3 is better at high frequency performance (especially my clone of the original). The best I have seen actually.
So for now I will take a break (have been sitting for hours) and post the distortion spectrum (Fourier chart) of Mimesis3 and the SSA:
Attachments
Give me 2 days, and you will have a 3 stage Latfet 100W TSSA amp in the same size as the one i have shown with 2 stages.
1) There will be no drive issues.
2) Low power (150mA idle)
3) No stability issues.
I have an amp here running with similar circuit which i can modify to a TSSA very easy.
- Sonny
1) There will be no drive issues.
2) Low power (150mA idle)
3) No stability issues.
I have an amp here running with similar circuit which i can modify to a TSSA very easy.
- Sonny
Thanks Sonny. Looking forward to it.
I have a high standard with drive capability. To achieve that, three-stage TSSA is possible only with very high supply rail (I use 40V secondary to produce 56v56 for output and 75V for front end after voltage doubler/cap multiplier), and emiter follower of course.
They say 170mA gives the lowest distortion but in my experiment, the spectrum is better with higher bias (higher overall distortion, higher 2nd order, lower high order distortions). With sufficient drive, no problem with 150mA.
Cool.
I have a high standard with drive capability. To achieve that, three-stage TSSA is possible only with very high supply rail (I use 40V secondary to produce 56v56 for output and 75V for front end after voltage doubler/cap multiplier), and emiter follower of course.
They say 170mA gives the lowest distortion but in my experiment, the spectrum is better with higher bias (higher overall distortion, higher 2nd order, lower high order distortions). With sufficient drive, no problem with 150mA.
Cool.
Finally got this thermaltrak version in stereo and housed.
http://www.diyaudio.com/forums/atta...metrical-amplifier-ssa-simple-thermaltrak.jpg
Not as neat as LC's fabulous versions, but quite functional. The SSA amps are on the vero board at the rear of the chassis. The other two amps are symasyms - for a four channel amp (still need a set of binding posts and rca jacks for the symasyms). Early listening to the SSA shows no defects that I can hear. Clean, clear music, powering the range above 300Hz so far. I like it.
Technical: Gain of 23dB. No noise at all with 100dB+ speaker. Very slight low frequency blip at switch on (100mV offset, maximum). Not audible with my speakers. No audible blip at shut off. Offset wanders around for a minute or so, but settles down to less than 20mV. Nice square wave at 100kHz - slight, well damped overshoot with a load, a little more pronounced without.
Practical: It takes a while to tune in the bias and offset, as they are interdependent. Once done, it's stable so far. If I were building again, I'd adjust it close, then measure the 500R pots on the current sources. Then I'd replace with a resistor and 50R pot, as it's quite sensitive to small changes and takes some fiddling to get within 10mV. But I'm not highly motivated to order the pots and change now, as I've already done the fiddling. The input transistors are from the same batch and well matched within type, but no matching across types. The driver and outputs are not particularly well matched. They are what I had.
http://www.diyaudio.com/forums/atta...metrical-amplifier-ssa-simple-thermaltrak.jpg
Not as neat as LC's fabulous versions, but quite functional. The SSA amps are on the vero board at the rear of the chassis. The other two amps are symasyms - for a four channel amp (still need a set of binding posts and rca jacks for the symasyms). Early listening to the SSA shows no defects that I can hear. Clean, clear music, powering the range above 300Hz so far. I like it.
Technical: Gain of 23dB. No noise at all with 100dB+ speaker. Very slight low frequency blip at switch on (100mV offset, maximum). Not audible with my speakers. No audible blip at shut off. Offset wanders around for a minute or so, but settles down to less than 20mV. Nice square wave at 100kHz - slight, well damped overshoot with a load, a little more pronounced without.
Practical: It takes a while to tune in the bias and offset, as they are interdependent. Once done, it's stable so far. If I were building again, I'd adjust it close, then measure the 500R pots on the current sources. Then I'd replace with a resistor and 50R pot, as it's quite sensitive to small changes and takes some fiddling to get within 10mV. But I'm not highly motivated to order the pots and change now, as I've already done the fiddling. The input transistors are from the same batch and well matched within type, but no matching across types. The driver and outputs are not particularly well matched. They are what I had.
Attachments
SSA by Sheldon
Very happy for you, after many hours spent on SSA project, well-founded considerations, good work finally rewarded to you by enjoying the music.
Since you have four channels amplifier, you can also check to swap low-high amplification to see if vice versa combination is better and/or to go with same channels for both bandwidths.
Regards Andrej
Hi Sheldon
Very happy for you, after many hours spent on SSA project, well-founded considerations, good work finally rewarded to you by enjoying the music.
Since you have four channels amplifier, you can also check to swap low-high amplification to see if vice versa combination is better and/or to go with same channels for both bandwidths.
Regards Andrej
How many hours nowadays day counts in Denmark? Daylight is surely longer but still 24 hours/day should be or is it?
Thanks Andrej,
The mission for this amp is to power the horn for this speaker: http://www.diyaudio.com/forums/multi-way/69906-unity-horn-project-progress.html#post792333
Right now, the horn has a passive crossover internally, crossed with a DEQX to the midbass and low bass. I have a 4W SE tube amp that I've been using for the horn lately, and wanted something with a little more reserve to play with. The passive crossover in the horn is not ideal, but is corrected by the DEQX. I will either redesign the passive network, or more likely cross everything active. If I go completely active, the SSA/Symasym is a good candidate for the horn compression and mid drivers. My version of the SSA is well suited for this, as the low power requirement will not strain the VAS current capability of this three stage amp. The Symasym is probably better in higher current applications than this version of the SSA, so well suited to driving the mids, or midbass.
Right now, I have an 801 based SE parafeed amp driving the horn. It sounds good. The SSA sounds just as good.
Sheldon
The mission for this amp is to power the horn for this speaker: http://www.diyaudio.com/forums/multi-way/69906-unity-horn-project-progress.html#post792333
Right now, the horn has a passive crossover internally, crossed with a DEQX to the midbass and low bass. I have a 4W SE tube amp that I've been using for the horn lately, and wanted something with a little more reserve to play with. The passive crossover in the horn is not ideal, but is corrected by the DEQX. I will either redesign the passive network, or more likely cross everything active. If I go completely active, the SSA/Symasym is a good candidate for the horn compression and mid drivers. My version of the SSA is well suited for this, as the low power requirement will not strain the VAS current capability of this three stage amp. The Symasym is probably better in higher current applications than this version of the SSA, so well suited to driving the mids, or midbass.
Right now, I have an 801 based SE parafeed amp driving the horn. It sounds good. The SSA sounds just as good.
Sheldon
I have been working on the amps (testing).
But i am struggling to find the right solution for the TSSA AMP i promised. The following Will start a discussion.
I have three AMP here right now.
The TSSA , mirand modified, simple cfb AMP.
The mirand and simple cfb AMP shares the same topology and Sub -3dB point. The mirand is an bjt's output stage. The simple cfb AMP is an lateral fet output type. Both "EF" types.
basicly those two AMP should have the same bas reproduction. But that is not the case. The lateral AMP is a bit lower level in perception, and less detailed. There is information missing.
But the TSSA which is a Drain output type does not miss level.
I now a lot Will not agree with me. Some would say that this is not the case with pass amps. But they are actually Drain output types. (aleph)
We are now 4 Independent people saing the same here in Denmark. Sourcefollower lacks precision and level in the Sub region.
But Drain follower Will be tested wednesday by two more. But until now we are two saying that it does not lack precision and level.
Maybe i Will end up with a 100watt TSSA 2 stage AMP.
But i am struggling to find the right solution for the TSSA AMP i promised. The following Will start a discussion.
I have three AMP here right now.
The TSSA , mirand modified, simple cfb AMP.
The mirand and simple cfb AMP shares the same topology and Sub -3dB point. The mirand is an bjt's output stage. The simple cfb AMP is an lateral fet output type. Both "EF" types.
basicly those two AMP should have the same bas reproduction. But that is not the case. The lateral AMP is a bit lower level in perception, and less detailed. There is information missing.
But the TSSA which is a Drain output type does not miss level.
I now a lot Will not agree with me. Some would say that this is not the case with pass amps. But they are actually Drain output types. (aleph)
We are now 4 Independent people saing the same here in Denmark. Sourcefollower lacks precision and level in the Sub region.
But Drain follower Will be tested wednesday by two more. But until now we are two saying that it does not lack precision and level.
Maybe i Will end up with a 100watt TSSA 2 stage AMP.
When you use lateral Fets at the output it is a good idea to go out of the drains. They are fantastic constant current sources. On Gm they are quite poor against bipolars so i think bipolars are a better choice when you want to go out of the low impedance output, emitter in that case. Laterals could be configured driving bipolars from the drain and the bipolars driving the output through the emitters. One problem remains : an emitter or source follower does not like capacitive loading. When you want to drive an ELS with a source or emitter follower you need an output coil and a Zobel.
Last edited:
I think it is more the positive temperature coefficient that come into account than the poor Gm.
In Sub bas impulses it has time to heat up, and the Gm decreases.?!???
Maybe i should rephrase it.
The feedback and feed forward is always tracking the errors. They can only correct them if they exist.
So in time domain the correction through the feedback is always behind. => The bas impulses are in most cases the one with the highest energy. So if the mosfet heats up and starts to produce an lower Gm the feedback tries to fix it. In other words the first edge of those impulses will have a false value.
The perfect amp would be a non feedback amp with an output stage in high class A level. But that is impractical to make. There for we add feedback to compensate. That why we want an as high as possible Openloop gain frequency as possible, so that the delay of the correction through the feedback is as little as possible.
The feedback and feed forward is always tracking the errors. They can only correct them if they exist.
So in time domain the correction through the feedback is always behind. => The bas impulses are in most cases the one with the highest energy. So if the mosfet heats up and starts to produce an lower Gm the feedback tries to fix it. In other words the first edge of those impulses will have a false value.
The perfect amp would be a non feedback amp with an output stage in high class A level. But that is impractical to make. There for we add feedback to compensate. That why we want an as high as possible Openloop gain frequency as possible, so that the delay of the correction through the feedback is as little as possible.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.