Dartzeel amp schematic - build this?

Hi Marigno

Thank you for your extraordinary contribution to this thread.

In you amp the original 108's distortion level has been reduced by perhaps 3 orders of magnitude, so the new distortion spectrum is entirely determined by the non-switching output. Which i still am to simulate and try to comprehend.

Not sure i share your enthusiasm about the dynamic ccs. It can save a modest amount of heat but i cannot see it as an improvement. ARC used it? ARC also produced a large number of very mediocre sounding gear, especially at the time of the D400.

The 108 and its variants still appear a mystery to me. Excessive and ugly distortion spectrum starting from 0W, yet doesn't seem to negatively affect the sound at least at lowish power/higher impedance.
 
WOTS-WHA216 sounds just fine :) And maybe you just found the Secret Sause in Dartzeel NHB458/468 :) An input balance servo circuit, no feedback and a ton of output transistors
I am blown away by your knowledge and skills. I just need some time to digest all this info.
And you got me wanting electrostatic speakers, also :) They seem to be Very Rare in Europe, though.
 
When the distortion presents a monotonically decreasing level on harmonics then you obtain a good sound even if the total distortion is high. Perhaps could this be the spectrum of NHB-108 distortion?. I designed a preamplifier (my first WOTS) that is simply made out of a 6H30 followed by a mosfet buffer, both working with high current, a lot higher than needed; no CCS at all. I think the distortion is high but the sound is really rich and detailed, i like to say that the preamplifier gives back to the sound what was lost during the process of recording. What a lie! High current in single-ended stages, usually means "single-ended" sound, whose distortion is not low but is monotonically decreasing on harmonics. So you obtain a very charmer sound.
But NHB-108 is not single-ended or, better, it is made by two opposite, symmetrical single-ended ciruitries meeting themselves at the collector join of the last transistor of the two VAS and this means push-pull. So we do not expect to see monotonically decreasing level on harmonics because the even harmonics are lower in level (cancelation occurs as a consequence of push-pull) than the odd ones. But again, without any doubt, the sound is amazing.
All my equipment is in Rome and I'm spending a lot of time in the US now, I met a guy who was the Acoustat dealer in SLC and, for $850, I bought a pair of 1+1 knowing I had to modify them. What can you get for $850? A couple of B&W 801 is worth about $3000. Second hand stuff, i cannot afford any new couple of speakers that you can call speakers. For Hattori: be aware that electrostatic speakers means high disturbance given by the electromagnetic field, if your main amplifier is sensitive to this disturbance, it will heat a lot (usually it is fixable).
 
When the distortion presents a monotonically decreasing level on harmonics then you obtain a good sound even if the total distortion is high.

You be the judge of how monotonic is the distortion. 1W and 25W into 8ohms simmed. The 4ohms are a lot uglier.
 

Attachments

  • 108 1W 8ohms.JPG
    108 1W 8ohms.JPG
    82.2 KB · Views: 331
  • 108 25W 8ohms.JPG
    108 25W 8ohms.JPG
    92.9 KB · Views: 334
Ugly indeed. But not that rare. Ugly only at the sight, our ears love it! I still can't believe that I received chinese fakes, tons of them! Usually so bad things happen to other people! In order to discover the fake transistor I understood I have to measure the E-C capacitance: for MJE15032/3 it measures around 70pF while in the fakes one, having a smaller chip, it is around 30pF. E-C capacitance not always is reported on datasheets so you must have a good one to compare. Is there a better method that avoids to crack the transistor?
Here nobody complains about assembled board, perhaps these have good transistors, the fake ones blow around or before 100V suply voltage.
 
Two transformer are the best solution for the sound quality, but not the solution for the hum. It should not hum, even with only one transformer. Something is wrong. I use the two legs decoupler you see in enclosed picture; try it, one only leg in your case.
 

Attachments

  • Decoupler.JPG
    Decoupler.JPG
    45.6 KB · Views: 357
Last edited:
I use a extern Servo. Only "mods" is resistors at input from 510 kohm to 110 kohm. Some Zeners are changed to 6.1 v. Base resistors is 20 ohm. Now bias is as I wish. Input transistors are within 1 mv. I don't use those terrible Wima capacitors at input. I hate Panasonic M etc. Think transformers will arrive next week.
 
Last edited:
Help me decide

How many output pairs are really needed? " " I had a bunch of MT200 so I used 11 pairs a channel. " " Using the NHB-108 is very simple even because you can buy the pcb.
I do not have any transistors lying around, I need to purchase them for the project. So which output transistors should I get? The standard pairs from Onsemi or the Sankens you are using?
Do You see anything wrong with this board?
1pcs Luxury Power Amplifier Board AUDIO Imitation DarTZeel NHB 108 Power Amplifier Circuit Board|Amplifier| - AliExpress
It has the DC servo, Standard NHB 108 design - but three output pairs.
I was thinking of buying just the board and get components separately. On the other hand, this seller claims they use matched Original Onsemi transistors and that the board is designed for AC voltages up to +-45V.
So maybe I can get this readybuilt board, after all, and just modify/change/add a few components according to your schematic? Will three pairs of output transistors be enough?
FYI: I have (3 in a package deal) 500W 42-0-42 VAC toroidal transformers, 2X10.000uF 80Vx6 rectifier/power boards, fairly large heatsinks and a "power delay soft start temperature protection board", already.
What would you do if you were starting from scratch with the board and had no components (or pcb) lying around?
You obviously do not think the DC servo makes the amplifier sound bad?
 

Attachments

  • stuff.jpg
    stuff.jpg
    206.2 KB · Views: 313
I use a extern Servo. Only "mods" is resistors at input from 510 kohm to 110 kohm. Some Zeners are changed to 6.1 v. Base resistors is 20 ohm. Now bias is as I wish. Input transistors are within 1 mv. I don't use those terrible Wima capacitors at input. I hate Panasonic M etc. Think transformers will arrive next week.

Why are Wima caps, terrible?