Yes avondale have been selling NAIM clones for a long time, I have always been surprised that they have the right to do so, it must be believed that NAIM audio no longer has or has no property rights over its old designs. The use of MJ15003/04 output is not very convincing for me, these cards may work well but on the sound quzlity I do not think it can match the original NAIM with these transistor output. For my part my goal is to do as well as an original NAIM, I do not pretend to do better.
Original value of the capacitor is 39pf. I've also increased it to 47pf. For 39pf, amplificator is not stable.
Not yet, I had collected two 0.22ohm emitter resistors on my test PCB to complete the new amplifier. So I bought other transmitter resistors and some components to do some tests, except that yesterday I couldn't get it to work properly again, I haven't found the problem yet, the test amplifier seems to be working at rest but when I 'injects a signal the gain is much too high. I need to troubleshoot this test PCB to get started!
In the meantime I tested my old amplifier with the GBF (which I did not have before I was doing my tests on a PC)
The first time I did all the measurements, everything worked perfectly without oscillations or HF instabilities.
A week later I redid the tests and I had small HF instabilities, this amp however using 15003 with Mica as a miller capacitor.
I wonder if it's coming from my LED lamp on the desk.
In all cases these instabilities are random I have the impression that it depends on what is lying around on the sector and that it must go through the earth and the ground loops. I think it's HF parasites in the sector that are abnormally amplified in the amplification stage?
Or another hypothesis is that my measuring device has a problem with my GBF or the oscilloscope so I tested another amp and I had no problem but I had to do it again since the phenomenon is random.. .
I did the listening test between the two amps on my HIFI system there is no photo, the new amp is clearer less aggressive in the treble, the timbres are much better and it even goes a little more far into the details. And I was not the only one to hear the difference, my brother and my sister-in-law were present too
In the meantime I tested my old amplifier with the GBF (which I did not have before I was doing my tests on a PC)
The first time I did all the measurements, everything worked perfectly without oscillations or HF instabilities.
A week later I redid the tests and I had small HF instabilities, this amp however using 15003 with Mica as a miller capacitor.
I wonder if it's coming from my LED lamp on the desk.
In all cases these instabilities are random I have the impression that it depends on what is lying around on the sector and that it must go through the earth and the ground loops. I think it's HF parasites in the sector that are abnormally amplified in the amplification stage?
Or another hypothesis is that my measuring device has a problem with my GBF or the oscilloscope so I tested another amp and I had no problem but I had to do it again since the phenomenon is random.. .
I don't think it is critical MJ15003 works but with 2 Mhz they are a bit slow. I do test with Mj15024 the sound is better than Mj15003 the difference is bigger than I expected. With 2N6341 the sound is very good and they have much power.
I did the listening test between the two amps on my HIFI system there is no photo, the new amp is clearer less aggressive in the treble, the timbres are much better and it even goes a little more far into the details. And I was not the only one to hear the difference, my brother and my sister-in-law were present too
In the 1980s, there was a hotly competitive Japanese market for new, fast switching and better gain linearity transistors because even cloth-ears consumers could tell that hi-fi amps based on such advanced "ring-emitter", "perforated emitter" or multi-emitter power transistors sounded better than standard 4MHz Ft power transistors - simple as that. In the years preceding that revolution, Naim's Julian Vereker had already adopted fast switching types in his acclaimed products and was quick to adopt the best of these new LAPT types from Sanken, as they sounded superior to all the various standard technology types of power transistors used up to then. Check the later interviews of the man himself that are still available on the 'net.
Now, you can't buy anything in decent linear amplification that doesn't use these or similar technology audio transistor types, with Sanken's best LAPT versions extending to an Ft of 70Mhz or so and Toshiba's 5200/1943 types still providing the affordable versions. If you want confirmation of the linearity facts, read any of Douglas Self's tomes from the 3rd ed. up of his "Audio Power Amplifier Design Handbook". Maybe also take a look at the history of what even conservative manufacturers like Onsemi had to do to catch up to the Japanese manufactures with their MJL and MJE150XX driver series semis.
Now, you can't buy anything in decent linear amplification that doesn't use these or similar technology audio transistor types, with Sanken's best LAPT versions extending to an Ft of 70Mhz or so and Toshiba's 5200/1943 types still providing the affordable versions. If you want confirmation of the linearity facts, read any of Douglas Self's tomes from the 3rd ed. up of his "Audio Power Amplifier Design Handbook". Maybe also take a look at the history of what even conservative manufacturers like Onsemi had to do to catch up to the Japanese manufactures with their MJL and MJE150XX driver series semis.
3.3 MHz is way above power diode switching and audio frequencies so unless there are also lower frequency, sub-harmonics of 3.3Mhz present, it's likely from another source - perhaps it's source is RF that is radiated by the oscilloscope's PSU. Check all possible sources in the vicinity of your test equipment by turning each of them completely off at their power source or perhaps just use a different PSU for each, if they are powered externally. Of course, you would also need another, different type of oscilloscope to isolate and test yours.Hi, I bought an oscilloscopeAnd I saw an oscillation at 3.3 mhz on output. It looks like come from power supply (not all sure). Did you see that kind oscillation (some kind of dumped sin wave)?
Last edited:
it sounds really good, better than my old version, for now it's been working for more than two weeks in my HIFI system, I haven't detected any overheating but as long as I don't understand the origin of these high frequency instabilities I am not totally serene about this design. on the other hand this problem also seems to exist on my old version of the NAP 250 and this one worked for three years without problems!
It sounds great but I also used high quality components and the cost is really high...you have to take that into account too.
here an exemple of the problem, i connect my laboratory , oscilloscope and GBF on the same Wall PSU socket and i test simply test with GBF it appear easily in 1000 Hz sinusoïdal 100mv signal.
I quite agree with Ian Finch, I have to do tests like this because depending on when the problem is present or not when I tested my old amp the first time the problem was totally absent. I'll see if I find the configuration that causes the problem to appear then I'll test another amp in the same way to see if by chance it doesn't appear too, but you'll have to take an amp with a ground because the last time the one I I took did not have any and it can very circulate parasites in the ground of the electrical installation.
I would not have time to do tests before Tuesday, we receive the family. So I'll let you know, i have to be careful about cell phone too maybe, cell phones can radiating and disturb.
Last edited:
In case you think that the output stage is oscillating check out the quasi complementary output here https://www.diyaudio.com/community/threads/1diffqc-amplifier.219354/
Here, 1 and 3 HF oscillation for 1 channal of my amplifier. 2, other channal from the same power supply. I put two 100nf caps on the amp. board, all goes bad
There is a post https://electronics.stackexchange.com/questions/323507/bridge-rectifier-4-diodes-vs-single-chip. here it's written some 1mhz oscillation and solution
"Red trace is without caps and without snubber. It rings at 1MHz. Adding 10nF cap across the diode lowers ringing frequency to 100kHz (green) which is no longer a problem, it also smoothes the edges, so the EMI problem is gone. Blue is with snubber added (R3/C3)." from that post.
There is a post https://electronics.stackexchange.com/questions/323507/bridge-rectifier-4-diodes-vs-single-chip. here it's written some 1mhz oscillation and solution
"Red trace is without caps and without snubber. It rings at 1MHz. Adding 10nF cap across the diode lowers ringing frequency to 100kHz (green) which is no longer a problem, it also smoothes the edges, so the EMI problem is gone. Blue is with snubber added (R3/C3)." from that post.
