I confirm It, new capacitors but not working. It drove me crazy.Salve, ho sostituito l'alimentatore con due pessimi condensatori da 10000uF e sembra che l'amplificatore voglia far crollare il pavimento. Forse un errore di montaggio sugli alimentatori? O cattivi condensatori?
View attachment 1052966 View attachment 1052967
Hello Andrew,
As you suggested on the kx2 homepage, I'm sending you the necessary information.
So, I have a working DIY dual mono amplifier in AB class with 250VA transformers each, that outputs +-27V on secondary winding,
and I'm hoping that I can "transform" it to dual mono kx2 amplifier running in AB class.
So, the construction basically follows your guidelines, there is a PSU PSB and two amplifier PCBs, as you can see on the photo.
Since the heatsinks are not as big as the ones suggested in the kx2 documentation, do you think that this is possible and do you have any suggestions ?
Thanks in advance!
Regards,
Filip.
As you suggested on the kx2 homepage, I'm sending you the necessary information.
So, I have a working DIY dual mono amplifier in AB class with 250VA transformers each, that outputs +-27V on secondary winding,
and I'm hoping that I can "transform" it to dual mono kx2 amplifier running in AB class.
So, the construction basically follows your guidelines, there is a PSU PSB and two amplifier PCBs, as you can see on the photo.
Since the heatsinks are not as big as the ones suggested in the kx2 documentation, do you think that this is possible and do you have any suggestions ?
Thanks in advance!
Regards,
Filip.
Attachments
Hey all KX2 lovers. Excuse me, I come from spending time on PASS amps (study) and considering a bipolar amp is new to me, but I like the outlook of the measured data show in the main PDF. What I find interesting is the triode harmonic behaviour found on page 59 in class AB, besides that, its an ultra quiet class A amplifier, absolutely class leading class A THD and noise floor. Class AAB show tendency to be a bit more revealing with dominant H3, a good blend of pure class A and AB.
I really like the option of switching class, this make it such that one can pair the amp with the driver or loudspeaker bandwidth. Interpret that as either bi amplification or active XO.
While I am learning about amplifiers, and especially this amplifier, I have a few questions.
I really like the option of switching class, this make it such that one can pair the amp with the driver or loudspeaker bandwidth. Interpret that as either bi amplification or active XO.
While I am learning about amplifiers, and especially this amplifier, I have a few questions.
- Converting the KX2 to full balanced operation (KX2B), is it best to stick with the symmetrical gain stage used or is it better to consider a FBS stage similar to the ones used in the Ovation 180/250 ?
- While looking at the Ovation 180W, I noticed you used a very nice protection board - would it be possible to move forward with a KX2B with such protection ?
Hello Oneminde,
The kx2-Amplifer is a fully balanced symmetrical design. This just refers to the fact that the amplifier topology is symmetrical and therefore notionally 'balanced'.
However, if you want to accept balanced input signals on a CFA amp (which the kx2 is) you will have to use a balanced to single ended converter and this is because the inverting input usually demands being driven from a low-Z source so you can't use it as the -input of a balanced input signal.
Some VFA designs that use JFET inputs, can take balanced signals in directly, but even with VFA's like the original Ovation 250 and the e-Amp, it is usual to use a balanced to single ended converter - you can do this with an opamp, or use a specialist chip like one of those from THAT Corporation which can offer very high CMRR.
The protection board on the Ovation 250 was designed in 2006/2007 and used a now discontinued 8-bit microcontroller (the world has moved on to 32-bit ARM and PIC devices in the meantime!). I might come back and look at this but it would almost certainly be an SMD package so the soldering skills of the builder would have to be very good - but no commitments to do this from my side at this stage - I really need to look at the practicality of doing this.
The kx2-Amplifer is a fully balanced symmetrical design. This just refers to the fact that the amplifier topology is symmetrical and therefore notionally 'balanced'.
However, if you want to accept balanced input signals on a CFA amp (which the kx2 is) you will have to use a balanced to single ended converter and this is because the inverting input usually demands being driven from a low-Z source so you can't use it as the -input of a balanced input signal.
Some VFA designs that use JFET inputs, can take balanced signals in directly, but even with VFA's like the original Ovation 250 and the e-Amp, it is usual to use a balanced to single ended converter - you can do this with an opamp, or use a specialist chip like one of those from THAT Corporation which can offer very high CMRR.
The protection board on the Ovation 250 was designed in 2006/2007 and used a now discontinued 8-bit microcontroller (the world has moved on to 32-bit ARM and PIC devices in the meantime!). I might come back and look at this but it would almost certainly be an SMD package so the soldering skills of the builder would have to be very good - but no commitments to do this from my side at this stage - I really need to look at the practicality of doing this.
Hello Bonsai
Just to be clear, maybe you misunderstood what I meant by balanced. I want an X amplifier, as in in balanced out vs single ended as it is atm.
So two mirrored sides like Pass usually does it. Each side takes care of half and opposite polarity. Symmetrical operation of the loudspeaker is utterly important to access the last transient gain. Super Symmetry should be relatively easy to inject but is not a must. An X amp allows for dual mono floating power supply, (not referenced to common ground). One of the marketing aspects used by Yamaha in their circlotron amplifiers
I am also reading the AFEC document and looks like an elegant Feed Forward solution, So while one are rearranging the KX2 to be fully balanced through and through, AFEC can be added.
And I am thinking about the protection board found in the e-amp 180W, if you can find a chip that will use the original software, that should be okay I think. No need to write a new program unless absolutely necessary.
I am all for SMD, that is what I want for most of the components anyway since the noise compared to TH is usually lower. And I checked, wire wound resistors does exist, but TH can be used as surface mount in a pinch.
And you mention THAT, I am considering those.... I stumbled upon it a few days ago investigating complimentary BJT's suitable for a SuSy gain stage and while doing so, a SuSy design by Kevin Gilmore popped up. the KG Dynali Mk2 or "The Multi Amp aka Dynalo Mk2" as it is also called. This amp is using the THAT340 which looks to have very nice specs compared to the 2SK170/SJ74 JFet everyone loves to throw at everything ... LOL.
I can also see that the Dynali Mk2 is using some constant currant source and there does exist another interesting solution here which interest me: A programmable CCS which accept a nice little shunt precision voltage source. "Current regulation is better than 10ppm/V from 1.5V to 40V" - The LT3092 can do 200mA.
These are early days for me and the KX2 amplifier platform, but I must say I like what I read and see in terms of performance. I am currently going over the AFEC document and need some time with it as well as the CFA paper you wrote. I know PASS LABS use CFA in their gain stages, but since I have not read the document yet, we might talk about two different CFA. And as I mentioned, this amp can be keep open to anyone who which to build one. Its not my IP anyway, I am merely looking at this as a smörgåsbord and considering what options are best for me and moving over to BJT from JFet, especially for the gain stage, could be a very nice one. Who wants to pay $10-12 per JFet and needing a bunch just so you can find the correct Id at the intended operating voltage, what a tiny nightmare.
I don't think one can hear the difference between a JFet and BJT if they have similar noise and capacitance values. If all else is similar, no. Humans does not have sophisticated enough hearing. In the end, you aren't just listening to the FET or Transistor, you are listening to the whole circuit, you are listening to the amps harmonic profile, its SNR, the noise floor, its soundstage presentation and so on. Say two amps have the same platform and components, one is using a 150VA transformer and the other is using a 300VA, the later is going to have better dynamic range, be able to dig deeper. It all matters, maybe more than if the amp uses JFet or BJT's in the gain stage.
Just to be clear, maybe you misunderstood what I meant by balanced. I want an X amplifier, as in in balanced out vs single ended as it is atm.
So two mirrored sides like Pass usually does it. Each side takes care of half and opposite polarity. Symmetrical operation of the loudspeaker is utterly important to access the last transient gain. Super Symmetry should be relatively easy to inject but is not a must. An X amp allows for dual mono floating power supply, (not referenced to common ground). One of the marketing aspects used by Yamaha in their circlotron amplifiers
I am also reading the AFEC document and looks like an elegant Feed Forward solution, So while one are rearranging the KX2 to be fully balanced through and through, AFEC can be added.
And I am thinking about the protection board found in the e-amp 180W, if you can find a chip that will use the original software, that should be okay I think. No need to write a new program unless absolutely necessary.
I am all for SMD, that is what I want for most of the components anyway since the noise compared to TH is usually lower. And I checked, wire wound resistors does exist, but TH can be used as surface mount in a pinch.
And you mention THAT, I am considering those
.... I stumbled upon it a few days ago investigating complimentary BJT's suitable for a SuSy gain stage and while doing so, a SuSy design by Kevin Gilmore popped up. the KG Dynali Mk2 or "The Multi Amp aka Dynalo Mk2" as it is also called. This amp is using the THAT340 which looks to have very nice specs compared to the 2SK170/SJ74 JFet everyone loves to throw at everything ... LOL.I can also see that the Dynali Mk2 is using some constant currant source and there does exist another interesting solution here which interest me: A programmable CCS which accept a nice little shunt precision voltage source. "Current regulation is better than 10ppm/V from 1.5V to 40V" - The LT3092 can do 200mA.
These are early days for me and the KX2 amplifier platform, but I must say I like what I read and see in terms of performance. I am currently going over the AFEC document and need some time with it as well as the CFA paper you wrote. I know PASS LABS use CFA in their gain stages, but since I have not read the document yet, we might talk about two different CFA. And as I mentioned, this amp can be keep open to anyone who which to build one. Its not my IP anyway, I am merely looking at this as a smörgåsbord and considering what options are best for me and moving over to BJT from JFet, especially for the gain stage, could be a very nice one. Who wants to pay $10-12 per JFet and needing a bunch just so you can find the correct Id at the intended operating voltage, what a tiny nightmare.
I don't think one can hear the difference between a JFet and BJT if they have similar noise and capacitance values. If all else is similar, no. Humans does not have sophisticated enough hearing. In the end, you aren't just listening to the FET or Transistor, you are listening to the whole circuit, you are listening to the amps harmonic profile, its SNR, the noise floor, its soundstage presentation and so on. Say two amps have the same platform and components, one is using a 150VA transformer and the other is using a 300VA, the later is going to have better dynamic range, be able to dig deeper. It all matters, maybe more than if the amp uses JFet or BJT's in the gain stage.
Are there any alternative recommendations to these devices (KSA1220 and KSC2690) ?
