I see the Infidel as a much more optimal CFA. Optimal in the respect that the backside diamond is a really good and very very close to ideal V to I converter that ensures close to perfect current summation in the feedback injection point. Normally you use a low value resistor network in the CFA feedback, but that also gives you issues, too low values and power/heat becomes a problem and too high the voltage modulation becomes a problem and you can push the transistor out of class A operation. by adding this back diamond you get a more ideal V-I conversion and a more dynamic linear feedback path.
I wonder if it could be build with a third Diamond in order to have a balanced input. 🙂
I wonder if it could be build with a third Diamond in order to have a balanced input. 🙂
It would seem with the CFA-X, Krypton-C and Infidel that we now have three different approaches to the CFA amplifier to compare. The CFA is the simplest followed by the Krypton in the middle and the Infidel as the most advanced of the three. Now how do they rank in sound on high quality speakers with a good source?
No matter how happy you are about the response you get, this is still hijacking a thread.
Mr. Kindhornman asked for opinions on CFA with good source. Is Audio Aero Capitole not a good source to your opinion? In this thread usually PC card or USB players are in use .. hijacking?!
You missed the REAL deal...
Elektor DIY Bat Detector Kit (140259-91) Elektor
We need to know when the Batman is in town too.
Just a quick note..
The use of such high resistor values in the feedback produces a high phase shift. Instead of 22K and 820R i would go for 2.2K and 82R.
But the easiest thing to try fix the oscillation. 22 - 100pF between in+ and in- will raise the HF closed loop gain.
Quiz time: Which British company has used this method?
The use of such high resistor values in the feedback produces a high phase shift. Instead of 22K and 820R i would go for 2.2K and 82R.
But the easiest thing to try fix the oscillation. 22 - 100pF between in+ and in- will raise the HF closed loop gain.
Quiz time: Which British company has used this method?
Ignore this if it has been tried. The polyester 0.1uF caps are probably not the most effective de-coupling caps at HF, perhaps a few 0.1uF 1206 SMD caps could be placed on the bottom side of the PCB to aid localizing the HF loops. Two could be placed right under C9 and C10, one more for V- could be placed near the jumper next to R21, another one more for V+ at where Cbp1 used to be.
Hi, Thimios,
Usually, Diamonds are not stable at very HF.
Have a look at:
www.esperado.fr - VSSA with Diamond input and DC servo
http://www.esperado.fr/vssa-diamond/dvssa.php
To see how i solved this problem, if it can help you.
Best regards.
How is Kypton C amplifier build compared to the two principles in the above article?
Anytging to improve to Kypton C compared to the optimized amplifier in the article?
Those caps are really good with something like 47 ohm basestoppers, for me this bandwith limiting is really good and can preven RF transmitter ingress into your circuit, maybe all that is needed is on the input side.
To the transistors at the diamond (Q5-Q8), basestopperrs and then at the base a small cap to GND, form a diamond rf filter. As previously said the 103 MHz is a radio station frequency, so maybe it's RF FM radio station reception. I once had an MC preamp that was a great tool for listening in at the nearby hospital internal radios too bad it was only the janitor telling where they needed cleaning an not where the royaly had babies or something interesting like that, well the ponit is that BJT circuits really need good input filtering.
I am not so sure that I would filter the backside diamond (q6-q8) as it also a pole in the FB
I am not so sure that I would filter the backside diamond (q6-q8) as it also a pole in the FB
Last edited: