I have simulate simple CFA amplifier. One using single transistor on VAS and another using Hawksford cascode.
Hawksford cascode VAS, is better THD but negatif slew rate is suffer.
How can I get high slew rate like single transistor on VAS but using Hawksford cascode?
Hawksford cascode VAS, is better THD but negatif slew rate is suffer.
How can I get high slew rate like single transistor on VAS but using Hawksford cascode?
I think that KSA992/KSC1845 are quite bad, I changed them with models 2sa970/2sc2240.
I you want to simulate square wave you need to remove capacitors from input and feedback path.
Hi Kovax !
Thank you ! , and sory for my late answer ! , since somehow I didn`t notice your post !
BTW , check interesting Tandberg TPA series of non GNFB Amps !
Best Regards !
Nice find, thank you for sharing. Still need to finish building Marsh HP amp.
Best regards, Kovax
I did not realize that for such a long time, the audio industry has shifted over to compl/push-pull and CFA circuits... with or without GNFB. Since, i dont really like building things from raw/scratch - unless I have to -- just updating/mods to used amps is perfect for me.... time-wise.
What are the best of these type topologies/designs out there and which brands/models that are 'used' and not rare ($) to buy for such rework? Such as, the TPA series. --- would be a great help to the DIY community. Suggestions?
[Interesting servo circuit shown]
THx-RNMarsh
What are the best of these type topologies/designs out there and which brands/models that are 'used' and not rare ($) to buy for such rework? Such as, the TPA series. --- would be a great help to the DIY community. Suggestions?
[Interesting servo circuit shown]
THx-RNMarsh
Last edited:
I have had the Tandberg 3026, a great disappointment compared to my other amp, A 30 W version of the Hiraga. the Hiraga had less SS grain and was despite or because of it's low(ish) power way more intimate and close to the music. At the same time I also got the matching preamp. It was switched on for about a week before i sold it again. they simply did not fit my system or my taste.
Last edited:
Oh Well....
to each his/her own. Topology is just the beginning... execution and better transistors and optimising circuit values, bias, etc is always needed.
Wonder how it would sound with some optimizing done? Put the circuit into a SIMulator and see what it doea and then what can be done to improve it.
-RNM
to each his/her own. Topology is just the beginning... execution and better transistors and optimising circuit values, bias, etc is always needed.
Wonder how it would sound with some optimizing done? Put the circuit into a SIMulator and see what it doea and then what can be done to improve it.
-RNM
Last edited:
"Topology is just the beginning... execution and better transistors and optimising circuit values, bias, etc is always needed.
Wonder how it would sound with some optimizing done? Put the circuit into a SIMulator and see what it doea and then what can be done to improve it." RNM
There are two designs from the past that I would like to volunteer for this process: The Spectra Sonics 701 by Dilley. These were made specifically for recording studio monitors. They sounded very nice. They featured an all NPN push pull output stage and a novel cascode second stage. The other design that I would like somone to really put the simulation to is the Bart Lochanti T circuit as featured in the Marshall Leach LTIM and the ML Super Amp by professor Leach of GA Tech. Both of these were featured in AUDIO magazine as a construction article. These amps used old RCA transistors (2n5415 and 2n3439). I would think there are better transistors nowdays. Ray
Wonder how it would sound with some optimizing done? Put the circuit into a SIMulator and see what it doea and then what can be done to improve it." RNM
There are two designs from the past that I would like to volunteer for this process: The Spectra Sonics 701 by Dilley. These were made specifically for recording studio monitors. They sounded very nice. They featured an all NPN push pull output stage and a novel cascode second stage. The other design that I would like somone to really put the simulation to is the Bart Lochanti T circuit as featured in the Marshall Leach LTIM and the ML Super Amp by professor Leach of GA Tech. Both of these were featured in AUDIO magazine as a construction article. These amps used old RCA transistors (2n5415 and 2n3439). I would think there are better transistors nowdays. Ray
Thank you, Dadod. Unfortunately, my VFA is ready first. I surprise when I listening the sound. This is my first amp design using LTspice. May be next week, I will try my CFA.
grhughes,
A friend of mine was friends with Dilley and still speaks highly of the Spectra Sonics designs. I think that there were some problems with the power supplies if I remember correctly and possibly undersized heat sinks. I will have to ask my friend about that. I know they are still sought after in studio applications.
A friend of mine was friends with Dilley and still speaks highly of the Spectra Sonics designs. I think that there were some problems with the power supplies if I remember correctly and possibly undersized heat sinks. I will have to ask my friend about that. I know they are still sought after in studio applications.
Thanks Kindhornman! The thing that intrigues me is that they could be configured for balanced in and balanced out without transformers, Also an all NPN push pull output stage. The much more recent BURR BROWN OPA2604 also used an all NPN output stage. Prof Marshall Leach used an all NPN output stage in his Wide Bandwidth preamp very similar to Dilley's 110 mic preamp. GRH
This uses the "4-transistor tringlotron" concept that dates back to Hawksford's I/V converter, but I don't know if it originated there.
I use this forgotten technology in my Kuartlotron:
The Kuartlotron
This webpage might not be complete yet, please share your thoughts on what I did wrong, left out, or could do better.
I built a copy It wasn't difficult with 4 transistors. That thing is fantastic! It is much better than the chip buffers made of voltage amplifiers that aren't amplifying. Instead, the Kuartlotron, is actually meant to be a buffer and that works so much better.
Thanks Kean!!
Thanks Kean!!
Last edited:
The concept doesnt date back to hawksford, it predates it by more than just a couple of years. Pioneer patents for the circuit were granted in japan in 1977 and US patents filed in 1981 and granted in 1983. Pioneer launched the first range of amps using the circuit in 1978 ableit in japan only for the first couple of years.
BTW thanks for the models page, I expect you understood my post and that there was no disrespect intended but actually the opposite before it was deleted. Ive tried all model parameters even advanced ones by intergrated circuits manufacturers and so far there isnt a parameter that succesfully corrects the early voltage problem which is why I asked.
Im curious though to which part in Olivers paper you found that maybe the 26mv rule was across both emitter resistors.
😎😀
Distortion cancellation. Yeh !!
Any advantage to making a mirrored version of itself -- compl/push-pull?
Thx-RNMarsh
Thanks Bonsai.
You're right Manso, we were never on bad terms as far as I can remember.
Marsh - without the solid 1k SE bias resistor, the Kuartlotron has very low error-corrected output impedance, so a complimentary version could have a bias current impossible to control. This just indicates it's working as it should. I simulated a complimentary version here, but I think it's inferior to the SE circuit, and uses twice the parts. I had to do a thermal simulation because it CAN do thermal runaway; I would suggest starting with this thermally stable (in simulation) circuit if you want a complimentary version.
http://www.diyaudio.com/forums/analogue-source/154210-mpp.html#post3555281
There's some useful info in that thread, I started discussion on the Kuartlotron here:
http://www.diyaudio.com/forums/analogue-source/154210-mpp.html#post3554680
EDIT: Corrected links
You're right Manso, we were never on bad terms as far as I can remember.
Marsh - without the solid 1k SE bias resistor, the Kuartlotron has very low error-corrected output impedance, so a complimentary version could have a bias current impossible to control. This just indicates it's working as it should. I simulated a complimentary version here, but I think it's inferior to the SE circuit, and uses twice the parts. I had to do a thermal simulation because it CAN do thermal runaway; I would suggest starting with this thermally stable (in simulation) circuit if you want a complimentary version.
http://www.diyaudio.com/forums/analogue-source/154210-mpp.html#post3555281
There's some useful info in that thread, I started discussion on the Kuartlotron here:
http://www.diyaudio.com/forums/analogue-source/154210-mpp.html#post3554680
EDIT: Corrected links
I made a thread for the Kuartlotron, so we can discuss it there if you want.
http://www.diyaudio.com/forums/anal...kens-simple-error-correction-superbuffer.html
http://www.diyaudio.com/forums/anal...kens-simple-error-correction-superbuffer.html
That was just a theory I threw out there. Honestly, I don't remember the last time I read that paper or if I ever read it at all. Maybe I should have tried to find out before opening my mouth? 😀