QUASAR a QUAD405 reborn design

[tvicol]

Even I'm a member of this forum since some time now, I never introduced myself.
My name is Tiberiu Vicol, I'm 39 years old, married, father of two gorgeous girls - 3 and 5 years old - and, of course, audio is my free time hobby.
I build my first QUAD405 18 years ago. I was fascinated, as many others, by the principle behind this amplifier and (at that time) by his sound.
Because this amplifier is more than 30 years old and since then many good things happened in electronics, I would like to share with this community my vision about reincarnation of QUAD405 which make use of few modern components. I renamed this amplifier to QUASAR. :-)
This project is the result of 6 years of work. Many versions and configurations have been tested, simulated and voiced.
My schematic contains some modifications mentioned by Bernd Ludwig and Keith Snook and I would like to thank to both of them for QUAD405 contribution.
As some of you have noticed, some versions of QUAD405 reborn are available on the net.
The version, which I present now, is very stable and was already tested and compared with several audiophile amplifiers. I'm very pleased with the sound of this amplifier and I hope that you will enjoy as well. ;-)
For sure I'll continue to improve this amplifier and further versions will be published.
So, here is the schematic.


Finished product.


More pictures with finished product. :-D
http://forum.audiofil.ro/index.php?topic=803.0

Regards,
Tibi

[tvicol]

Seems that a finished project don't make too much buzz ...

OK. Here is the next version.

Regards,
Tibi

[tvicol]

Please change extension to .asc LTspice file.

Regards,
Tibi

[MadMutt]

That's just, well, superb.
Absolutley stunning metal work.

I'm sure others would love to know what you tried it against, and the results.

[tvicol]

Originally posted by MadMutt
That's just, well, superb.
Absolutley stunning metal work.

I'm sure others would love to know what you tried it against, and the results.

Enclosure is from here.

Regarding comparisons, I'll not give amplifier names because I don't want to start off topic discussions.
IMHO, making an amplifier and stating that this sound very good based on simulations, without any direct comparisons with some high-end gear, is not the right way.
I'm not a spice man. I use spice only to see if some ideas are working or not.
I design, measure, listen and compare, than again design, measure ...

Regards,
Tibi

[juma]

Great job Tibi !
Quad 405 was my high-school project too, and being a same age as you I understand and share your emotions for this amp.

The lack of response in the thread is probably due to output devices you chose - lateral MOSFETs are a bit expensive and not easy to find, +- 60 V PS dictates the use of higher voltage elkos (63 V type is not safe to use here) which are more expensive....

Anyway, you made a great amp which is a bit on the "monster" side where audience is not so big.

Keep up the good work !

[ekaerin]

Tibi,

Great work, all the way.

I really like the transformer you are using. Looks like something I saw in a Genelec active studio monitor once.

Also, Thanks for the enclosure link. Good looking and fair pricing.
The enclosure is usually what put my projects on hold.

/ Mattias

[tvicol]

Here is a picture with one prototipe I used for tests.
This is with vertical mosfets IRFP240/9240.

In this version R3=10Kohm; R10=100ohm; R17=22ohm; R19=220ohm
and only two pairs of IRFP output mosfets.
The sound with IRFP's is very good as well.
Large heatsinks must be used in order to avoid vertical MOSFET run avay.

Regards,
Tibi

[Leolabs]

Oh vertical fets!Why not normal BJTs???

[tvicol]

Originally posted by Leolabs
Oh vertical fets!Why not normal BJTs???

The amplifier is based on a low power A class with T7 & T8 and associated ccs.
This stage need to assure enough gain, in order to have "magic formula" Z1Z3=Z2Z4 working, and also enough current for final stage.
The output impedance, of class A stage, is also quite high at only 50mA.
IMHO to drive BJT's from only 50mA is not enough. At high output levels, a power BJT with a beta of 20-30 will overload class A stage and amplifier will fail into high distortions. I think here more appropriate are Darlington's or Mosfet's.
I didn't managed to get the amplifier stable with Darlington's.
During my tests, I also observed that the amplifier is more stable with very fast output devices. MOSFET's are much faster and this is another reason, beside high input impedance and more easy to drive. I don't want to go into a BJT vs MOSFET debate as long a thread is already discussed here .

As long this is a feed-forward error correction amplifier, another choice are IGBT's, which should work fine as current dumpers and maybe a future version will be with them.

Regards,
Tibi