Of course: your amp is overcompensated. I prefer to cure the problems where they occurs. I used 50pF instead of 220pF.
That is my way: max bandwidth with no peak with global compensation, then cure the overshoots on little signals.
The moment i took the screenshot was in NGFB, not obvious ?.
(I don't understand neither your "precise", who care about the filter precise symmetry at so high frequencies ?).
Ok, so you don't even want to look at my suggestion, even five seconds in sim ? Too bad...
Just judge, looking at those two graphs witch one seems the more healthy. ( everything equal elsewhere) and conclude yourself.
EOT for me, you don't seem to be opened to any dialog. I'm disappointed but it is not so important.
Attachments
Last edited:
By the way: "the CFA gain is defined as R12/(R19//R9)+1" was close to an insult. I took my first job as an electronic engineer in a R&D department of the bigest French HIFI company in 1970 !
Last edited:
Sorry Esperado, I did not want to insult you, that formula was there to shows what resistors play a role in CFA for all who are interested.
OK, you are experience designer, I am just hobbyist, but you are making beginner's faults and when I point to it you are offended.
Again you are wrong with NGNFB as it was not obvious to you, you missed that connection as it should be there for both nodes, the jumpers are on different position. I attached the schematic again with the jumpers, J1 for CFA and J2 for NGNFB.
You seems to not understand the supper pair(Baxandall pair) as you have left first transistors collectors not connected anywhere.
This design is finished, and I don't intend to change it, so maybe I will try your LP filter in some other design.
I am not disappointed in you as I know your temperament from other threads and posts.
Attachments
I showed-you the dumping, Just this part and just for you, nothing else. How can-you have an idea of my work in progress and what i try? It will appear strange, but sometimes, I explore circuits for some reasons or try alternatives. Don't you ? I was in the middle of something at the moment of this screenshot. Nothing else. And nothing you can lead to conclusions about my competencies as you did..
This, anyway, and despite your arrogance that, me too had noticed several time on several threads, don't change the fact that your pre-amp is a nice design, quite original, with amazing performance.
Have a Nice day.
Last edited:
Calm down Christophe, nobody knows everything. My knowledge is very limited regarding audio and I just try to learn some more now when I'm retired and could spend more time with it. I studied electronic but never used it in my professional life. I worked in telecommunication industry and mostly on software side.
And thank you very much for your opinion about this preamp, I appreciate it. I am not the one who can judge the sound competently, I know only from the feedback of the builders.
And most important, I don't want that you are angry with me.
Damir
Don't worry, Damir. Believe-me or not, you always had my sympathy (witch has no more value than this ;-)
My problem is the general lack of respect between contributors in this forum, witch makes-me a little paranoiac. Post like this one:
http://www.diyaudio.com/forums/solid-state/235695-no-nfb-line-amp-gainwire-mk2.html#post3484841 that shocked-me a lot. You share a clever idea, and what you got in return is this ?
I was disappointed because i tried to *share* my personal experience (fragile, because based on subjective point of view) with Diamonds and Super pairs. Be assured it was the result of a lot of listening tests during my life (half as a sound engineer, half as an audio designer, not a comfortable position here, and pretty uncommon. Audiophiles takes-you for a "technician", and objectivists for an "audiophile"). It was not a critic, it was the contrary.
Oh, believe-me or not, i am usually able to get an idea of the sonic "character" of a design at first sight on the schematic. This feeling was positive enough for i wanted to adopt your for a project. I'm confident of its "airy" and "fast" (detailed) character. The previous project that exited-me was the VSSA from Andrej. And the amazing engineering work of OS (slewmaster).
Please, if you read my posts on the forum, you will find that i had NEVER assaulted somebody witch had not done-it previously, and, I hope, not tell too much B.S. (read my lips ;-)
Friends ?
My problem is the general lack of respect between contributors in this forum, witch makes-me a little paranoiac. Post like this one:
http://www.diyaudio.com/forums/solid-state/235695-no-nfb-line-amp-gainwire-mk2.html#post3484841 that shocked-me a lot. You share a clever idea, and what you got in return is this ?
I was disappointed because i tried to *share* my personal experience (fragile, because based on subjective point of view) with Diamonds and Super pairs. Be assured it was the result of a lot of listening tests during my life (half as a sound engineer, half as an audio designer, not a comfortable position here, and pretty uncommon. Audiophiles takes-you for a "technician", and objectivists for an "audiophile"). It was not a critic, it was the contrary.
Oh, believe-me or not, i am usually able to get an idea of the sonic "character" of a design at first sight on the schematic. This feeling was positive enough for i wanted to adopt your for a project. I'm confident of its "airy" and "fast" (detailed) character. The previous project that exited-me was the VSSA from Andrej. And the amazing engineering work of OS (slewmaster).
Please, if you read my posts on the forum, you will find that i had NEVER assaulted somebody witch had not done-it previously, and, I hope, not tell too much B.S. (read my lips ;-)
Friends ?
Last edited:
I am comparing this preamp to my DCB1.
The DCB1 buffer does have no gain but it produces a really large soundstage with "almost" perfect imaging.
The Gainwire is more muscular and with NGFB it gets close to the DCB1 image size.
Choosing GFB, the image shrinks and the highs loose their natural "air".
The DCB1 buffer does have no gain but it produces a really large soundstage with "almost" perfect imaging.
The Gainwire is more muscular and with NGFB it gets close to the DCB1 image size.
Choosing GFB, the image shrinks and the highs loose their natural "air".
I suppose I was not clear enough. I usually proceed that way to compensate my amps.
First suppress any low pass filter. Set the global compensation in order to get the max available FLAT bandwidth (No peak at HF).
(here, C2=50pF).
Then, add a low pass filter just enough to kill any overshoot on transients of square waves (40KHz or such).
Hi RCruz,
Which PSU did you use for DCB1? I tried the BCB1 on a simple Z-Reg PSU (like the one with RJM Sapphire Headphone Amp) but I'm not satify as with CFA GainWire. May be I need to try again with PSU like yours ...
Which PSU did you use for DCB1? I tried the BCB1 on a simple Z-Reg PSU (like the one with RJM Sapphire Headphone Amp) but I'm not satify as with CFA GainWire. May be I need to try again with PSU like yours ...
RC, I wonder your listening impression. From what i had seen in simulations, the CFA mode is supposed to be both faster and creating less distortion (At this level, I mean No distortion ;-).
Too, the "expansive" behavior of the CFA mode is supposed to give subtle detail enhancements.
Last, with the speed inside the loop, no IM (TIM) to be suspected.
You report exactly the contrary. I should suspect one or several of the followings:
- You are influenced by your belief that GNFB creates negative effects.(A blind comparison should eliminate this point)
- Your enclosures are too generous in trebles. So, the (subtle) added details of the CFA mode, yet in excess, is masking a little the "soundstage" .
- Your power amp is not fast enough and (may-be) you'll have to decrease the transition frequency of your input filter.
- There is some instability at very high frequencies when the preamp is looped.
- You prefer soft reproduction to micro-dynamic.
Please, don't take this as any kind of a criticism, I have no Idea of your system, or your musical preferences. Those are just the questions i would ask myself in this situation. If the answer is negative at all those question, we will have to question all our believes about TIM.
Even with headphones I have the same impression.
I did listen to the preamp using different setups, the last one was with a Nagra pyramid feeding stenheim speakers and we all felt the same 🙂
I did listen to the preamp using different setups, the last one was with a Nagra pyramid feeding stenheim speakers and we all felt the same 🙂
Hi Damir,
May I ask why you used such a high value resistor for R18 ?
I would have thought that biasing Q9 and Q10 higher, maybe around 1mA, would have made Q9/Q10 faster, lower noise, and their base currents would compensate for the base currents of the helper and improve the current mirror balance (by some uA only, I admit)
May I ask why you used such a high value resistor for R18 ?
I would have thought that biasing Q9 and Q10 higher, maybe around 1mA, would have made Q9/Q10 faster, lower noise, and their base currents would compensate for the base currents of the helper and improve the current mirror balance (by some uA only, I admit)
I did extensive simulation and that value was chosen as the best.
I never tried to change that on the real pre amp to see if any sonic improvement could be gained.
- Status
- Not open for further replies.