I used square vawe at different frequencies. Load was 4X22Ohm 100W in parallel. (5.5 Ohm 400W, cooled with ventilator.To measure what you read on your scope, are you using a square wave input to the amplifier? At what frequency? What is your resistive load on the output your measuring across? Curious
Now, I am building a circuit tor THD measurements with Arta software and sound card. It is true that some data is published but I just want to verify. I think you understand that many kit data are more than obviously exaggerated. Also, recomendations for biasing are inconsistent both with practice and theory. We shall see.
I've got a audio gen that will do square waves but it only goes from 20 to 5kHz.
I'll try it, thanks
You may observe crossover distortions inthat region too. Just stretch the graph to see the rising edge as an arc.
Here is also a link below to interesting conversation. Note that "optimal bias" isn't that clear concept and that opinions may differ for practical reasons.
https://www.diyaudio.com/forums/solid-state/252333-cfp-output-require-bias-ef.html#post3843292
Here is also a link below to interesting conversation. Note that "optimal bias" isn't that clear concept and that opinions may differ for practical reasons.
https://www.diyaudio.com/forums/solid-state/252333-cfp-output-require-bias-ef.html#post3843292
Actually, these amplifiers. It has nothing to do with the golden voice circuit.
Golden voice e405 e305 e series. Basically, MOSFET output is used.
K1058.
And no to3 power transistor will be used.
actually. The performance of plastic transistor is better than that of metal power tube. TO3。
Although the to3 transistor looks better.
Power amplifier should pay attention to performance. Don't look at his appearance. Even materials.
If you can. Try not to choose the model of imitation manufacturer.
Performance test of analog amplifier. It can be used as a reference.
For example, under the same test conditions. The distortion of l20.5 is less than that of Naim. Less than Quad, Krell, passlab all power amplifier.
This is just a comparison of the measurement results here. So that's what it means.
You can compare the differences. But under different conditions. The same amplifier will have different data. This is normal.
Even winter measurement and summer measurement. Its data could be ten times different.
Different power sources are different. It doesn't make any sense.
It's like you can use five. Or ten amplifiers. Use the same equipment. Comparative measurement was carried out. You can see the difference between them. Which is better.
That's what measurement means.
Of course, if you can find a circuit with lower distortion. In the case of the same measurement. I'm very interested. Measurement can find a better amplifier.
Thanks for kind advice,Performance test of analog amplifier. It can be used as a reference.
For example, under the same test conditions. The distortion of l20.5 is less than that of Naim. Less than Quad, Krell, passlab all power amplifier.
This is just a comparison of the measurement results here. So that's what it means.
You can compare the differences. But under different conditions. The same amplifier will have different data. This is normal.
Even winter measurement and summer measurement. Its data could be ten times different.
Different power sources are different. It doesn't make any sense.
It's like you can use five. Or ten amplifiers. Use the same equipment. Comparative measurement was carried out. You can see the difference between them. Which is better.
That's what measurement means.
Of course, if you can find a circuit with lower distortion. In the case of the same measurement. I'm very interested. Measurement can find a better amplifier.
I have already tested L20.5, at different vrms and different loads. I am using very good capacitance multiplier power supply with almost non-existing ripple. I might also add voltage control to the power supply to elliminate its influence on measurement results. Adittionaly I want to achieve repeatability of results and to studdy influence of the quiescent current on THD.
By the way, the measurement results are quite encouraging. This indicates a solid implementation of Douglas Selfs' design.
With kind regards
Thanks for kind advice,
I have already tested L20.5, at different vrms and different loads. I am using very good capacitance multiplier power supply with almost non-existing ripple. I might also add voltage control to the power supply to elliminate its influence on measurement results. Adittionaly I want to achieve repeatability of results and to studdy influence of the quiescent current on THD.
By the way, the measurement results are quite encouraging. This indicates a solid implementation of Douglas Selfs' design.
With kind regards
Nice to discuss l20.5 amplifier with you.
Douglas self is great.
However, l20.5 is not the same as self amplifier. Although 90% of them are the same.
However, all amplifiers in the world are at least 50% similar. So it doesn't matter.
Amplifier on self book. It can't be used directly. Because it has some problems.
But his main structure. There is no problem. It is also the same structure of most manufacturers.
Single differential amplifier, basically no big difference.
But in fact, we can't use self amplifier completely. Including a lot of completely duplicated circuits.
believe me. I'm not boasting about myself.
In my practical application, I found that there are some problems in the self circuit.
Self doesn't mention these problems in the book. So, the actual measurement. There are also long-term customer feedback verification results. Sometimes it's more important than studying alone.
Hi LJM,
Today and yesterday I have made a series of measurements and I have established that in particular case of L20.5 THD indeed doesn't change signifficantly with bias. Even with extremely low bias value, the THD was almost the same. However, I was able to observe with the use of oscylloscope that at low bias there is persistent 53MHz oscillation. So I have returned bias to higher value.
Right now I have on test bench four LJM amplifiers: L12-2, MX50LT, L20 V9. 2 and L20.5. Indeed, they all sound much better than some notoriously famous amplifiers. Simple and good. Yet, to attain really good results they need good power supplies and speaker protections. I use good 42V capacitance multiplier and can consistently achieve complete absence of audible noise, below - 100dB.
Congratulations.
Today and yesterday I have made a series of measurements and I have established that in particular case of L20.5 THD indeed doesn't change signifficantly with bias. Even with extremely low bias value, the THD was almost the same. However, I was able to observe with the use of oscylloscope that at low bias there is persistent 53MHz oscillation. So I have returned bias to higher value.
Right now I have on test bench four LJM amplifiers: L12-2, MX50LT, L20 V9. 2 and L20.5. Indeed, they all sound much better than some notoriously famous amplifiers. Simple and good. Yet, to attain really good results they need good power supplies and speaker protections. I use good 42V capacitance multiplier and can consistently achieve complete absence of audible noise, below - 100dB.
Congratulations.
Hi LJM,
Today and yesterday I have made a series of measurements and I have established that in particular case of L20.5 THD indeed doesn't change signifficantly with bias. Even with extremely low bias value, the THD was almost the same. However, I was able to observe with the use of oscylloscope that at low bias there is persistent 53MHz oscillation. So I have returned bias to higher value.
Right now I have on test bench four LJM amplifiers: L12-2, MX50LT, L20 V9. 2 and L20.5. Indeed, they all sound much better than some notoriously famous amplifiers. Simple and good. Yet, to attain really good results they need good power supplies and speaker protections. I use good 42V capacitance multiplier and can consistently achieve complete absence of audible noise, below - 100dB.
Congratulations.
You're right.
It is possible that the test will be limited by the test conditions. Different people have different power supplies. Or test equipment.
Maybe the data is not exactly the same.
But under the same test conditions. The test results of different amplifiers can be used as reference.
We can see which is better.
May I ask which one is your favorite or how they differ from each other? L15 mosfet seems not popular like other ones, don't know why.
@ljm_ljm can you give a brief info regarding L15 mosfet amp please?
Last edited:
Hi Terra...
I'm glad to hear from you All are good, just L20.5 has lower THD, but is it really audible? I doubt one can sense the difference, the distortion is so low. Also, connectors are similar so it is easy to work with multiple amps on the bench at once. Just pull wires from one and push them into another.
However, I would like to understand how these midgets function, not just to have a cheap hi-fi amp. Purchasing one is much cheaper if you account your own time.
L20.5 has these two SOT26 devices with erased labels so it is impossible to study it. If you follow another V9 thread you may have noted that Marigno likes L20 V9.2 I tend to agree with him - he has good taste and understands the subject.
Cheers dear Terra...
Hi Terra...
All are good, just L20.5 has lower THD, but is it really audible? I doubt one can sense the difference, the distortion is so low. Also, connectors are similar so it is easy to work with multiple amps on the bench at once. Just pull wires from one and push them into another.
However, I would like to understand how these midgets function, not just to have a cheap hi-fi amp. Purchasing one is much cheaper if you account your own time.
L20.5 has these two SOT26 devices with erased labels so it is impossible to study it. If you follow another V9 thread you may have noted that Marigno likes L20 V9.2 I tend to agree with him - he has good taste and understands the subject.
Cheers dear Terra...
Hi Berlusconi, me too
I thought you will pick L20.5 as it seems latest and most strict design by LJM. It is based on CFP toplogy and noted as "no crossover distortion" which means Class-A sound with Class-AB bias. I have both mosfet and bjt versions of Elvee's Circlophone (it also has Class-A features at Class-AB bias) and I want to compare it with other discrete amp kits. I think that LJM's boards are good candidates in this regard.
Thank you.
I do not really have my favourite, they just have diferent flavours. It is fun to compare them side-by-side. V9.2 is interesting, and more important, there is a schematic whilst L10.5 has that secret part with SOT26 devices.Hi Berlusconi, me too
I thought you will pick L20.5 as it seems latest and most strict design...
Have a pleasant evening.
Hi Terra...
All are good, just L20.5 has lower THD, but is it really audible? I doubt one can sense the difference, the distortion is so low. Also, connectors are similar so it is easy to work with multiple amps on the bench at once. Just pull wires from one and push them into another.
However, I would like to understand how these midgets function, not just to have a cheap hi-fi amp. Purchasing one is much cheaper if you account your own time.
L20.5 has these two SOT26 devices with erased labels so it is impossible to study it. If you follow another V9 thread you may have noted that Marigno likes L20 V9.2 I tend to agree with him - he has good taste and understands the subject.
Cheers dear Terra...
In fact, I really like l20v9.2
It is a three-stage Darlington output circuit. This is a very rare circuit.
I work to make it stable. Tried many different models. Transistor combination.
It's very good.
I have assembled it myself but from all parts from the kit. Price is a bargain so it pays off to order the ."finished" board.
There is no volume control - this is just amplifier board. Plan using soft start and speaker protection boards, and of course - power supply.
I ordered the kit, but it has been missing in the DHL system for two weeks now, it got within 200 miles of me and then fell off the conveyor...
The reason I ordered the kit was to use higher tolerance components and match all the parts. I don't think you can order just the boards.
Yes, I've almost always did so - for me it is fun to measure every individual part and to match parts so you know everything is done right. A bit time consuming, but fun, so relaxing.
And, I gave up on DHL - they have just high price tag and nothing more. They are "high-end" so to speak.
AliExpress Standard shipping works just fine, at least in Europe. I have always got everything in 3-4 weeks, including customs procedures.Good luck.