Juntuim, are your output devices running cold with little or no bias? Insert an emitter resistir on one of the OP devices and see if its dissipating current or not. It might be an issue with the kit you have. I have a similar kit. Sounds thin and grainy at low volumes. Gets much better once warm and at higher volumes. My drivers run very warm... My speakers are 84w/db, so you and I are probably operating the amp in different regions of linearity. 
One of another reason I got confused about the negative report on the amplifier because not to a long time ago I read this awesome comment. Of course, that is the original (not clone) Dartzeel. Please read all feedback on the next Youtube post.
YouTube
Patrick J Newton
1 year ago
Yes, I just uploaded as proof for the manufacturer (who is resolving the problem). I liked it so much I bought another one. I have had Krell, Spectral, MBL, Linn, Naim. This beats everything including the MBL 9008A monoblocks ($52k retail) which I just sold. I run direct from CD/Streaming source. Will never look back. Despite the hum on this particular instance ( caused by voltage regulator) I would say it's a fantastic amp. The build quality is great. You can find them online through the usual sites
What I want to say maybe your clone need some more tweak, a bit more attention etc.(Please do not give up easily) Danny made some interesting simulation on this amp I don't know if he is willing to share it openly or even that would help at all.
I think I will go on and order pair empty PC boards, I do have most of the semiconductors from ON semi (at least I know those will be originals, nowadays you can trust on those you get at the eBay. ) if I do like the result with one pair transistor and still I feel the need to have the higher power version what is very hard to get (empty PC boards) at the moment. Momentarily I am occupied with something more important than DIY. After when I have the time and my health allowed me to return to warm up the soldering iron I will put a clone together.
YouTube
Patrick J Newton
1 year ago
Yes, I just uploaded as proof for the manufacturer (who is resolving the problem). I liked it so much I bought another one. I have had Krell, Spectral, MBL, Linn, Naim. This beats everything including the MBL 9008A monoblocks ($52k retail) which I just sold. I run direct from CD/Streaming source. Will never look back. Despite the hum on this particular instance ( caused by voltage regulator) I would say it's a fantastic amp. The build quality is great. You can find them online through the usual sites
What I want to say maybe your clone need some more tweak, a bit more attention etc.(Please do not give up easily) Danny made some interesting simulation on this amp I don't know if he is willing to share it openly or even that would help at all.
I think I will go on and order pair empty PC boards, I do have most of the semiconductors from ON semi (at least I know those will be originals, nowadays you can trust on those you get at the eBay. ) if I do like the result with one pair transistor and still I feel the need to have the higher power version what is very hard to get (empty PC boards) at the moment. Momentarily I am occupied with something more important than DIY. After when I have the time and my health allowed me to return to warm up the soldering iron I will put a clone together.
Juntuim, are your output devices running cold with little or no bias? Insert an emitter resistir on one of the OP devices and see if its dissipating current or not. It might be an issue with the kit you have. I have a similar kit. Sounds thin and grainy at low volumes. Gets much better once warm and at higher volumes. My drivers run very warm... My speakers are 84w/db, so you and I are probably operating the amp in different regions of linearity.
Yes, they are running very cold. Actually I was wondering that last night. Im not native english speaker and my english isn't so good, so I kinda need very specific instructions what to do
One of another reason I got confused about the negative report on the amplifier because not to a long time ago I read this awesome comment. Of course, that is the original (not clone) Dartzeel. Please read all feedback on the next Youtube post.
YouTube
Patrick J Newton
1 year ago
Yes, I just uploaded as proof for the manufacturer (who is resolving the problem). I liked it so much I bought another one. I have had Krell, Spectral, MBL, Linn, Naim. This beats everything including the MBL 9008A monoblocks ($52k retail) which I just sold. I run direct from CD/Streaming source. Will never look back. Despite the hum on this particular instance ( caused by voltage regulator) I would say it's a fantastic amp. The build quality is great. You can find them online through the usual sites
What I want to say maybe your clone need some more tweak, a bit more attention etc.(Please do not give up easily) Danny made some interesting simulation on this amp I don't know if he is willing to share it openly or even that would help at all.
All about this hobby is subjective, everyone likes different sound
I wrote earlier that I made permanent groundings and 100hz buzz went away, it's really silent now actually. I know also that these exactly same boards that i'm using now got a good review also from previous owner. He used these with 86db/w speakers, and I'm with 96db/w speakers. Why I was so disappointed is that I was expecting good and solid bass, at least close what Aleph 5 offered, but didn't get even close.I'm not certainly giving up on this one, it's still on my main system! I think that this amp runs too cold with my sensitive speakers, can't get enough power out of it with reasonable volume levels? Would higher bias work?
Higher bias should definitely work. The originals are biased to 20 watts pure class A, thats a lot of heat, that our kits are missing. I have not been able to figure out the bias situation yet, not sure if its the DC correction circuit or what, but the drivers seems to be biased ok (running very hot) so beats me why there is no bias on the OP devices... the kit is a lower priority for me right now... if you find out why, please do share.
Measuring them personally (signal generator, oscilloscope, 8 Ohm dummy load with 1 KHz sine wave), I can confirm that on my amplifier made with clone cards and photographed in my previous posts, at least the first 10/15 Watt per channel worked in full "Class A", thanks to the high static polarization values of the final power transistors.
The same fact is also confirmed for the original darTZeel NHB-108 reviewed by Wes Phillips by the magazine Stereophile of April 24 2005.
And I can assure you that the cards heat up a lot, despite the generous heat sink mounted.
The same fact is also confirmed for the original darTZeel NHB-108 reviewed by Wes Phillips by the magazine Stereophile of April 24 2005.
And I can assure you that the cards heat up a lot, despite the generous heat sink mounted.
Some facts do not add up
The amp cannot be biased up to 20W in pure Class A at 55V rail voltage with one pair power transistors! One pair plastic case power transistor cannot take that much power and heat! It is biassed up to 20W in Class A the rail voltage must be much lower. Otherwise, it will blow up at the first warm up.
I can also accept that my word is being questioned, but also denying evidence of what Sterophile says in the following paragraph seems to me a bit excessive:........
"Attention must be paid
The darTZeel NHB-108 Model One amplifier (NHB stands for "Never Heard Before") is based, Delétraz says, on the triple principles of simplicity, purity, and reliability. That means the 108 employs open-loop input and output stages with no global negative feedback. Nor does it use relay switches or fuses in the signal path. It employs a scant six transistors per polarity side from input to output, and only a single pair of bipolar output transistors per channel. The NHB-108's output is biased to produce about 20W in class-A, and it runs rather warm—as in Careful around that heatsink, buddy!"
Bha ......... are not what is the concept of "much" power and heat, but the MJL3281A/MJL1302A complementary final transistors, are devices capable of 15 Amperes of continuous current collector and 200 Watts of power.............. each!!
And also considering the yield in "Class A" equal to 25% theoretical and 20% real, realized at the Safy Operating Area of the MJL3281A/1302A, I don't see big limits to manage 6/7 watts in "Class A" for each of the complementary transistors.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
My objections to this pleasant theory are only based on the available circuit values.
As for the theoretical possibility...20W class A into 8ohms is about an amp of bias current, or 60W dissipation per device - manageable with these output devices provided they are adequately heatsinked.
Which of course raises the question: do the Chinese resistor values correspond to the original circuit, or are they safely tweaked so that less customers are likely to cause fire and mayhem?
As for the theoretical possibility...20W class A into 8ohms is about an amp of bias current, or 60W dissipation per device - manageable with these output devices provided they are adequately heatsinked.
Which of course raises the question: do the Chinese resistor values correspond to the original circuit, or are they safely tweaked so that less customers are likely to cause fire and mayhem?
There is a further issue. The output, being a diamond buffer does not allow simply changing the bias of the outputs. The CCS and the other transistors in the buffer increase their own dissipation accordingly. Which is not a problem on its own, but all this heat has to be dissipated somewhere and the best that can be said about the Dartz heatsink is that it looks very elegant and red, but not big at all. At 1A bias, unless i am mistaken, the dissipation crawls upwards of 250W/channel.
I can also accept that my word is being questioned, but also denying evidence of what Sterophile says in the following paragraph seems to me a bit excessive:........
"Attention must be paid
The darTZeel NHB-108 Model One amplifier (NHB stands for "Never Heard Before") is based, Delétraz says, on the triple principles of simplicity, purity, and reliability. That means the 108 employs open-loop input and output stages with no global negative feedback. Nor does it use relay switches or fuses in the signal path. It employs a scant six transistors per polarity side from input to output, and only a single pair of bipolar output transistors per channel. The NHB-108's output is biased to produce about 20W in class-A, and it runs rather warm—as in Careful around that heatsink, buddy!"
Yes, I read that, that is why I wrote something does not add up! I am not questioning your word, again I read that to. If the original Dartzel runs 55VDC rail voltage that is a lot of power to handle 1 pair power transistor! To biased to give 20W in pure Class A cannot believe that!
For example, I built several Hiraga Class A amplifier the 20W and 30W and that runs hot, very hot. Yes, the Dartzel has a large heatsink based on the size estimation it would be enough for 20W Class A operation at the right rail voltage, 55VDC it just too much. Usually, for that rail voltage, we use 2 or more pair power transistors 1 to 5W Class A and after Class A/B operation. How long that 1 pair transistor last? All those data from ON semi is true at room temperature only! My last word if it runs up to 20W Class A the rail voltage cannot be 55VDC. I know there is some amplifier design we can switch to Class A for a couple watts or run it all in Class A/B, these case we do not have that option.
Just purchased pair empty Dartzeel PC boards
300W-500W After Stage Amplifier Circuit PCB Board Amp base on darTZeel NHB-108B | eBay
I would rather use my own semiconductors and purchase the rest of the parts even if this way will cost more.
I will use hard wire to use 3 pair power transistors
Of course at first, I try it with one pair
300W-500W After Stage Amplifier Circuit PCB Board Amp base on darTZeel NHB-108B | eBay
I would rather use my own semiconductors and purchase the rest of the parts even if this way will cost more.
I will use hard wire to use 3 pair power transistors
Of course at first, I try it with one pair
I'm very happy to read it and it's a good choice to buy and select components.
As you rightly said, do not deny yourself the experience with a single pair of complementary transistors, which in terms of balance, speed, timbre and musicality will give you great satisfaction.
And maybe as an alternative to the BJT MJL1302A/3281A, you can decide to mount the pair MJL4281A/4302A which is a bit more performance in terms of power dissipation.
As you rightly said, do not deny yourself the experience with a single pair of complementary transistors, which in terms of balance, speed, timbre and musicality will give you great satisfaction.
And maybe as an alternative to the BJT MJL1302A/3281A, you can decide to mount the pair MJL4281A/4302A which is a bit more performance in terms of power dissipation.
Hi
Post-133 seems to have workable numbers but the assumptions in post-134 seem wildly inaccurate. To expect the bias and resulting dissipation to crawl to 250W/ch assumes that zero care has been taken to make the bias temperature stable.To me, the Dartzeel seems to be more hype than fact, but it is generally easy to avoid thermal runaway.
The topology is very basic and proven at this point, based on diamonds that can be set up to run class-B quite easily. The 20W class-A claim can be true alongside whatever the rated output is to require 55V rails. has anyone measured the amp to see if the rails sag under load? or to see what the idle current actually is? This would be more useful than conjecture.
Post-133 seems to have workable numbers but the assumptions in post-134 seem wildly inaccurate. To expect the bias and resulting dissipation to crawl to 250W/ch assumes that zero care has been taken to make the bias temperature stable.To me, the Dartzeel seems to be more hype than fact, but it is generally easy to avoid thermal runaway.
The topology is very basic and proven at this point, based on diamonds that can be set up to run class-B quite easily. The 20W class-A claim can be true alongside whatever the rated output is to require 55V rails. has anyone measured the amp to see if the rails sag under load? or to see what the idle current actually is? This would be more useful than conjecture.
Well, couple of weeks listening behind. It seems that my heatsink is way too big for this build and you have to really crank it up to get even a little heat. But it seems to make a difference, sounds much better after a while.
I have reserved my biggest sinks for Class-A anyways, so this will move into smaller case at some point (extra 4U laying around, without sinks), and thinking of smaller caps with cap multiplier. Can't do it just yet though.
I have reserved my biggest sinks for Class-A anyways, so this will move into smaller case at some point (extra 4U laying around, without sinks), and thinking of smaller caps with cap multiplier. Can't do it just yet though.