DarTZeel, Stereophile's 2005 Amplification Component...

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

just noticed for some day ago in MikeB's thread (Terry Demol's post) a link to Stereophiles awarding of 2005 Amplification Component, DarTZeel NHB-108 Model One.

Well, I'm not so interested in "ready made products" and what Stereophile thinks, but nevertheless I got a bit curious to what is so "special" with this amplifier so I dug up the patent (issued as late as last year..) and grabbed the principle schematic.

The designer tells the amplifiers input and output stage is outside the NFB loop which is only around the second stage, which to me looks like effectively being more or less a CFP with Q5 inbetween.

The principle schematic is attached below.

Cheers Michael

BTW, the Swiss designer Hervé Delétraz says NHB stands for "Never Heard Before"... :D
 

Attachments

  • dartzell_nhb-108.gif
    dartzell_nhb-108.gif
    12.2 KB · Views: 2,689
serengetiplains said:


Price is of course relative to what you can afford relative to how much you value the art coming through your speakers relative to ___________ [beer, a BMW .... ]


Oh those diyAudio "politition" aspirants...

As a DIY:er I would of course brew my own beer, drink it up, and then drive around in my DIY BMW! :clown:

Wellcome back with some more technical related inputs to this thread, or else please spare your energy. :rolleyes:

Cheers Michael
 
It has in fact two NFB loops. Because of the low loop gain of the last loop - THD is a little high. But it is constant over frequency ! Whis is seldom achieved by amps with overall NFB.

So it would probably be wise to listen to it before uttering a verdict.

Is it overpriced ? IMO yes !

Regards

Charles
 
Ultima Thule said:

20 W Class A...

I should be more clear mentioning that only the bias level of the otput stage admits 20W in class A, else the amplifier delivers 100W RMS into 8 Ohm.

Charles,

I'm not shure where those two FB loops you see are fitted, I thought Q3, Q5 & Q7 being one (and the only) loop where the feedback path is from Q7 through R11, divided down to earth by R9, to Q3.

The diode, D1 & D2, arrangement is a bit peculiar to me, it seem to let Q11 & Q12 move around a bit until it's clamped by those diodes at more demanding drive levels/loads.

Cheers Michael
 
Upupa Epops said:
They walk from " wall to wall " in this redaction.... One year win Halcro with outstanding parametres, second year this s... :D ...Yes, probably for " nice distortion "....


Upupa,

Yes it's funny how Stereophile change taste for amplifiers, maybe they don't like "ultra clean" sound reproduction anymore... BTW darTZeel resembles of your "favorite amplifier" Ayre! :clown:

Cheers Michael
 
Hi, Ultima Thule,

In the AudioManuEN.pdf, the last page, it is written that the DC offset can vary from -590mV to +590mV :bigeyes: It can reproduce signal to 1Mhz.

I first know this amp from a post by Jon Marsh here, quite sometime ago.

Hi, Phase_accurate,

The diodes D3 and D4, is it usefull with bipolars? I saw this in classD gate mosfets, usually in half bridge mode. What does D3 and D4 do with bipolars?
 
Lumanauw,

those diodes you have seen in class D amp are for sure in the opposite direction and it's for maintaining faster switch-off than switch-on giving lesser crossconduction witch is much more important of course with switching circuit's and high frequencies.

But in darTZeel it's not in the same direction, it's a mystery to me.
I can just speculate as I can see some minor things but...

Oh, it's that much offset, not good in my book! :whazzat:

Cheers Michael
 
Hi, Ultima Thule,

You are right :D It is the wrong direction. Maybe it is to speed up turn-on? I think I've seen such diode in the "base stopper" position, but I seen it in the driver transistor (after VAS), not in the base of final transistor. In that schematic that I see, I suspect it is to help stabilize (prevent oscilation), due to mismatching impedance between VAS and driver base.

Interesting review by Stereophile, this amp sounds "different".

I think (I'm not an expert :D) there are some interesting thing that obvious in the schematic :
- No resistor from ground to either Q1 and Q2 (where these 2 transistors are the interface with "outside world" to receive signal). In Q1-Q2, usually there is a resistor to base (input impedance resistor, about 22k-100k) attached to base, to define ground point. I've tought about this alot. As NP said, a transistor is "blind", it doesn't know what kind of signal is coming to it's base. It receives all signal, whether it is audio signal or ground noise. IF the ground point is "dirty", Q1-Q2 will amplify the ground noise, because their emitors are connected to the supply (via R1-R2). Any difference between B-E will be amplified, the transistor doesn't care wheter it is the supply that is dirty or the ground that is dirty, relative to each other. Connecting base Q1-Q2 in this amp that way is clever, so Q1-Q2 cannot receive the "ground noise", only the emitors sense the "supply noise" between +and-rail.
-The gain loop Q3-5-7 is "current feedback loop". This is very fast, enabling the power bandwith up to Mhz numbers. Good thing that Q5 and Q6 is that polarity, so that R13-14 have to be big valued, means Q7-Q8 receiving little of "ground noise".
-I'm not sure about this, but I feel that using biasing scheme like VBE multiplier is degrading the sonics, due to VBE multiplier is not constant at all, depending on current flowing to it and frequency. He uses diamond buffer, no need for biasing scheme at all here. I think this helps the sonics alot.

From the schematic, gnd point is only connected to the emitors of Q3 and Q4. He said to ALWAYS connect the speaker when the power is ON. Is it possible that the DC offset is determined by the speaker itself? 2ohm speaker will give better DC offset than 8ohm speaker attached to this amp? And it is dangerous to turn on this amp without any load attached, why is that?
 
This thing sounds a bit finicky for my tastes. I do wonder at it being a Component Of The Year when, according to the review, it radiated heat like a toaster oven, ate fuses like popcorn whenever they twitched a wire, and finally blew up (admittedly when a slow-blow fuse was installed instead of a fast-blow) when Atkinson tried running a 19/20 kHz IM test.

Perhaps it's me, but I suspect most people dropping the price of a decent car ($18 K for six transistors a side? Holy smokes!) on an amplifier would like a bit more reliability.


Cheers,
Francois.
 
I was temped to whip up this amp real quick. It resembles the Hiraga alot with a few bells and whistles. The hiraga I built sounds very good the DC drift is alright it starts from 200mV and settles to about 20. So it would be interesting to built.
Does anyone have the exact patent number and or the location where it's deposited?
 
Ok, I read the patent, it's only two pages of text, and... well I let somebody else comment it! ;)

However the diode phenomena, the patent states the following:

[0020] The diodes D1 and D2 provide a very fast pulse
response through their p-n junction capacitance and protect
the resistors R19 and R20 against heating.
well.. :rolleyes:

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