Simple Killer Amp - Listening impressions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello all,

I'm one of the few people here that decided to acquire a Simple Killer Amp.
When I first saw the topology, I was impressed by its simplicity and
specificities. Never saw such a thing, so I ordered a GB150D to test it by
myself. The PCB is nice, well done, compact. I didn't test the mosfets but
the small trannies : well matched. The instructions are clear for a trained
builder. The GB150D can output up to 150W in AB or 50W in class A. I decided
to go for a light class A and built a big dual mono power supply with what I
had on hand. Four 300VA earth screened transformers, two high power bridged
rectifiers, a CRC arrangement wich per rail is composed of 10mF + 1R +
2x22mF low ESR. It's actually oversized because I first made the amps run in
AB mode with 50mA per mosfet to test overall stability. Offset and bias are
stable enough though the trimming principle is incredibly simple. There's no
cap in the signal path and the bootstrap caps are those provided in the kit.
To be able to test it easily and as I don't have a dc coupled preamp on
hand, I replaced the input 1nF polystyrene cap with an 470pf one and added
an 10K ALPS pot. A few other resistor have been downscaled to fit the +-29V
provided by my actual psu. At this time I have no cabinet, everything is on
the floor. When time permits (next year), I will increase the bias (have to
mod some resistors that are actually a safety for not overbiasing) and test
class A performance. From what I can hear now, it will be putted in a
cabinet and not dismantled like many others.

My actual system, a CD14 by Marantz, an AKSA amp, a kt88 cathode coupled
push pull amp and Cabasse Iroise LS given for 92dB/W/m. An easy task and
load for all my amps.

The first thing that impressed me, was the very low residual output noise
with shorted inputs. My 2mV/div Tektronix in diff mode give me a nice 0.6mV
rms noise. I have to put the ear in close contact with my dual concentric
medium/trebble driver to detect something.

The sound is precise, more than my other amps. The bass are tight and very
well damped. No distorsion is audible, totally uncolored on the whole range.
The music flows easily without any agressivity and that's the second thing
that impressed me. It's not usual for me as I most the time listen to tube
amps or hybrids. Overall impression that it's as usual less life-like than a
tube amp but it's so nice and where is the truth ?
I like that amp and will try to get the best from it !

To follow are the listening impressions of Hubert, a friend of mine, who is
way better than me when it comes to put into words what we hear. It's an
experienced speaker builder and he has my AKSA at home too. It's somewhat
different from the usual techno babble we can read on this forum, but will
give you an idea of what you can expect from such a schematic. And of
course, I have no financial interest with the designer of this project.

Happy new year to all,
Francis


/*---------------------------------------------------*/
>
> Hello Francis,
>
> Here are my feelings about this new amp we listened yesterday afternoon.
>
> The stereo was my current system:
> -source: not a true performer but I perfectly know its listening
> characteristics: a Micromega Optic BS using a bitstream DAC, at least the
> "music flow" is preserved in term of fluidity and details remain present.
> -speakers: 1801's Ellis audio, a standed monitor 2-way using a 7" Seas
> magnesium and a 1" Hiquphon tweeter, with a perfect filter (for these
> drivers): not the ultimate performer in bass range but very revealing on
the
> global bandpass, particularly free of distortions; its relative lack
remains
> in macro-dynamics but only if compared to very good 3-way and...84db
> efficiency...
>
> IMHO, a real killer amp must be dynamical, revealing, with an extanded
> bandpass from sub-bass to top ends and neutral.
> So, lets go with this one:
>
> - dynamic: particularly equaly reproduced from sub-bass to top end. Often
> amps have more dynamic in midrange or trebles, this one not; good point.
In
> a first view I guessed a little lack in this area because of the poor
> efficiency of my 1801's but finaly no.... certainly more "near to the
truth"
> than others.
>
> - details: as with the other solid-state amp I currently use (no no, I
will
> not write its brand-name but as you know because you are its owner, it is
a
> well-known and very appreciated kit in the worldwide audioholic community)
> here is THE problem: inner details (micro details) are not on par with my
> tweaked valves-amp Rogers E40a: i.e. the lack of reverberations and
> prolongation of musical notes is obvious, not only in the upper range but
> also in the low-mid. Thus, timbres are affected, but not the musical flow;
> fluidity is preserved even if a few harmonics are "breaked down".
>
> - frequencies extand: a perfect straight line, allowing to create a right
> image in term of sizes of the instruments and place of recording, in
> relationship with true soundstages; good point.
>
> - neutrality: the last criterion, for a reason: NEVER HEARD SUCH A
> DISTORTION FREE ELECTRONIC: very, very uncommon: my bad cds aren't so bad,
> they don't become good cds but they are less bad, obviously. NO, my
speakers
> aren't harsh sometimes, no the magnesium driver doesn't distord in the
> upper-mid on some cds as I guessed before !!!
>
> To summarize:
> An almost perfect amp; if the lack of inner details could be avoided, this
> amp would be THE killer for anyone enjoying neutrality but only once
linked
> to a top source and to top speakers at "real" listening level. Others (me
> included) may prefer less accurate electronics, with different listening
> characteristics (notably to compensate less accurate sources and/or
speakers
> or to match bad room acoustics or to listen at moderate level), but that's
> another "commonly discutted" issue.
>
> Thanks again leting me listen to all your electronical buildings, allways
> very instructive.
>
>
> Cheers,
> Hubert
>
 
Well Hi Francis and Hubert,

Welcome to DIY Audio.

Thanks for your honest first impressions of my SKA amplifier and thanks Francis for being one of the very first to 'take a chance' on my first DIY kit offerring.

I hope to hear your further impressions, and, knowing your competence in electronics, any suggestions or tweaks you may come across.

I think it might sound better in a box with neat wiring layout.

Cheers,
Greg

:D
 
Thanks Dimitri, Mastertech, and Hugh,

I knew the amp was pretty good but our (wife and I) impressions don't count for much. All kudos go to Francis who has experimented and built himself a creditable amp. I am happy to have facilitated this.

However there a quite a few more who have purchased a pair as a, I suspect, end of year holiday project.

C'mon guys, I can build one in 2 hours to go! Bit longer to get it in a box.

Cheers,
Greg

:D
 
Hubert-
To summarize:
> An almost perfect amp; if the lack of inner details could be avoided, this
> amp would be THE killer for anyone enjoying neutrality but only once
linked
-------------------------------------------
I f the lack of inner details could be avoided, thats what my friend
jean told me a long time ago, without details and resolution
the amp wont be precise and music will suffer, it all comes down
to the exact implementation of wideband art, all my amp have one
feature in common-"high resolution" thats no neutrality thats
musicality!

Regards
 
I think inner details are overrated.
The outer details I find more important.
The details between these these two extremes, including lower and upper details,
should always be present, in my opinion.

I prefer a good overall performance, that does not disturb the music I listen to.
Lack of some inner details is no big problem for me.
Especially with more heavy pop music played with electronical instruments.

Think this Killer Amp would be almost perfect for me.
;)
 
Hi Francis,

Had I known you would be running the amp directly from a 10K pot, I would have suggested reducing that 1nF polystyrene input capacitor even as far as 100pF, to restore detail at mid settings where source impedance is as high as 2.5K + the 1k of the input filter. In reality it only needs to be large enough to ensure stage bandwidth for the bootstrapped diff'l stage. It's something Chicco and I discussed at length recently in the SKA thread.

there are those of us that believe there is always something lost, however small, when you use such passive sourcing. But it's sooo convenient with CD source.

I'll have to recommend this in the assembly instructions.

Cheers,
Greg

:D
 
rellum-
), I will increase the bias (have to
mod some resistors that are actually a safety for not overbiasing) and test
class A performance. From what I can hear now, it will be putted in a
cabinet and not dismantled like many others.
---------------------------------------------
this guy has been modifying the circuit without consulting the designer no wonder the performance is varied, take a look what
modification is about to be done next! it is not a simple as that
mate! if youre after top performance

Regards
 
Hi Mastertech,

The amp uses open cct sensitive pot biasing and limits the available range deliberately. So, for Class A biasing, lower values of CCS biasing resistors must be used. Francis knew this and was simply commenting that he will need to ascertain the best values so that Class A biasing was at the the top of the available range, with the supply voltages chosen.

He did, in fact, contact me privately over this and, with his extensive practical experience, found quite capable of setting it for his needs with little input from me.

He's simply exploring the design and what it can do. It's his prerogative.

Cheers,
Greg
 
mastertech
it is very few recordings that have inner details
so, this is no problem
as long as all normal details are there

as mastertech pointed out, this is not 100% standard setup
amplifierguru however says he think is no big problem
with rellum version


but here is my general reflection:
----------------------------------
My first reaction also, was that this listening test
was not done with the recommended setup.

An amplifier is a system within itself.
It is designed to work good as system.

Also this is a little system within a working environment.
This is power supply, input source and output load.
It can be more or less tolerant to different environments,
but designer had surely some specific requirements
in mind when making design.
Some important factors are for example
voltage and current capacity of power supply and impedance of load connected.


Listening impressions from this test is not the final word,
even if it can give some little hints and is very interesting.

I am sure will come other verdicts here
with a setup that follows instructions 'in the book'.
:)
 
amplifierguru said:
Hi Mastertech,

The amp uses open cct sensitive pot biasing and limits the available range deliberately. So, for Class A biasing, lower values of CCS biasing resistors must be used. Francis knew this and was simply commenting that he will need to ascertain the best values so that Class A biasing was at the the top of the available range, with the supply voltages chosen.

He did, in fact, contact me privately over this and, with his extensive practical experience, found quite capable of setting it for his needs with little input from me.

He's simply exploring the design and what it can do. It's his prerogative.

Cheers,
Greg

Greg,

Do you think the design will benefit from more class A?

It will be interesting to see if there is any audible
difference.

Cheers,

Terry
 
Hi Terry,

Yes indeed. Francis is quite capable of reconfiguring the amp from one mode to another so his resultant impressions will be interesting.

I, personally, would like to believe there is nothing in it from low bias AB to Class A. Simply because I have tamed all higher HD to well below -90 dB and 2HD/3HD around -86 dB each at 50mA per MOSFET. Reasonable thought would have us believe such levels are inaudible and well masked by low order HD from dynamic loudspeakers which is orders of magnitude higher.

Let's wait and see.

Cheers,
greg
 
Hello,

Yesterday I've increased Iq to run one channel at 1.4A and replaced input RC
filtering caps with some silver mica 100pf/500V caps. Good overall
stability. No overrun. Exemplary offset stability. To perform
some comparisons, I decided to run one channel AB and the other class A. I
made some mono listening tests yesterday evening with manual swapping. First
impressions : no changes. Tomorrow, I connected my switcher. It's a self
made one with wired remote controled relays. Many contacts are in parralel,
the off contacts feed a dummy load to keep each amp allways loaded and
ground is also switched to avoid ground loops. The only change that I
noticed when swapping is in the bass region. No more
impact but a fuller bass sound, with slightly more energy. I connected my
scope at the inputs of the amps and with the ADD+INVERT mode had a flat line
in the middle. Connecting the probes on the speaker outputs, the flat line
is somewhat modulated but at low frequency only.
The less efficiency of the bootstrap caps at LF gives less open loop gain
and thus, increase the output impedance. It wasn't noticeable on two way
speakers or with a low bass content on my own one.
In class A, higher transconductance of the output mosfets gives some more
gain that allows for that bass extension. Clearly, this amp has nothing to
gain with class A. Energy waste.
The bootstrap caps could be replaced with some biggers and best quality. The
P mosfets could be replaced with some IRFP9140. Sadly, I have only two of
them on hand and they are not easy to provide.
A more clever solution could be to add an error correction circuit. But not
an Hawksford one, output stage has gain here. Why not an opamp based one
wich action could be limited to the low frequency range ? A bass extender
could also do a nice job...

This amp is clearly a HIGH FEEDBACK design, but a clever one.
Won't suit everybody's taste.

To summarize, it combines : low cost, high power if needed, very low
distorsion, good definition and an incredible softness wich allows for long
listening sessions.

Cheers,
Francis
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.