There were several posts in this thread that indicate problems with this amplifier board. Earlier, it was reported that quiescent current should not be increased over 15 mA. Rallyfinnen has indicated that:
Below, I have shown a couple of representative measurements. Before publishing this I have assured that the presented results are really objective and repeatable. I have measured four boards, one was pre-built with all components provided by the seller. In all other boards I have replaced "genuine" TTC/TTA output devices with original Toshiba 2SA/2SC transistors purchased from Mouser. In one board I have replaced practically everything with genuine components. I have even tested the use of TIP31C/TIP32C in place of the original drivers. Everything has ended with the same result.
At low power the amplifier behaves normally but the increased output power results in asymmetric oscillation. This board oscillates at the positive side of signal. Quantitatively, the oscillations are quite substantial.
Now, please refer to two snapshots shown below. Measurements were taken at 17.8 Vrms, RL 8 Ohm and rails +/-34,5 V. I have presented enlarged snapshot of oscillations (in blue color) to show the real nature of the problem. You can observe significant distortion of the signal.
For me, this amplifier is useless.
Indeed, this board suffers from serious deficiencies. I am sorry for bringing bad news, but it is what it is: this board was sold by presenting fake-news, almost perfect "measurements". My measurements indicate that his is far from the truth. Quite the contrary.
Below, I have shown a couple of representative measurements. Before publishing this I have assured that the presented results are really objective and repeatable. I have measured four boards, one was pre-built with all components provided by the seller. In all other boards I have replaced "genuine" TTC/TTA output devices with original Toshiba 2SA/2SC transistors purchased from Mouser. In one board I have replaced practically everything with genuine components. I have even tested the use of TIP31C/TIP32C in place of the original drivers. Everything has ended with the same result.
At low power the amplifier behaves normally but the increased output power results in asymmetric oscillation. This board oscillates at the positive side of signal. Quantitatively, the oscillations are quite substantial.
Now, please refer to two snapshots shown below. Measurements were taken at 17.8 Vrms, RL 8 Ohm and rails +/-34,5 V. I have presented enlarged snapshot of oscillations (in blue color) to show the real nature of the problem. You can observe significant distortion of the signal.
For me, this amplifier is useless.
I bought a Quad 405 ready built amp board.
It was a 100watt 100khz oscillator.
I binned it.
It was a 100watt 100khz oscillator.
I binned it.
OK, so what can one do to mitigate the oscillations on this board? Any solutions?
My knowledge in this field is too modest to do that. LJM has equipment to measure his boards and knowledge to solve the problem if there is a solution. ... and yet he sells them and I'm not surprised. On aliexpress/ebay there is a faulty Krell KSA-50 board for sale with error - it smokes as soon as you power it because zenner has wrong connection . Yet these people don't care at all, they keep selling them. They're from a world that we'll never understand. Absolute lack of personal integrity and no respect for the values we share.
I agree about Aliexpress sellers. I have a dispute with a seller of car Android units. I bought one because it claimed 4g internet functionality but it won't work with any UK networks, 3g is as fast as it goes and that is being switched off this year.
Also, the Bliueooth implementation won't work with diagnostic dongles, and the speedo display is only KPH. If you change the units to MPH, all that happen sis the letters KPH in the speedo dial change to MPH but the calibration remains in KPH. It's laughable. The seller has been promising a software update to address these issues for months. Aliexpress customer services are a waste of space.
Also, the Bliueooth implementation won't work with diagnostic dongles, and the speedo display is only KPH. If you change the units to MPH, all that happen sis the letters KPH in the speedo dial change to MPH but the calibration remains in KPH. It's laughable. The seller has been promising a software update to address these issues for months. Aliexpress customer services are a waste of space.
Just an idea how to avoid disappointments with these Chinese "Hi-End" amplifiers "at the price of breakfast". Chip Amps!
Reputable companies have invested considerable time and money in development of chip amps, first class professionals have been involved in their development. Reputable companies market chip amps, you may rely on them. Technical specifications are available and verified. There are many developers who have built great amplifiers based on chip amps. Any good chip amp based amplifier is better than any of these "Cheap Chinese High End Amplifiers"
And yet people want to believe that these cheap Chinese kits based on fake components are the right way towards Audio Heaven. Human mind is unfortunately so prone to manipulation, especially with low-price great products. There is no such thing - avoid trash.
Reputable companies have invested considerable time and money in development of chip amps, first class professionals have been involved in their development. Reputable companies market chip amps, you may rely on them. Technical specifications are available and verified. There are many developers who have built great amplifiers based on chip amps. Any good chip amp based amplifier is better than any of these "Cheap Chinese High End Amplifiers"
And yet people want to believe that these cheap Chinese kits based on fake components are the right way towards Audio Heaven. Human mind is unfortunately so prone to manipulation, especially with low-price great products. There is no such thing - avoid trash.
Last edited:
Well, allot of the electronics in your car, even my 2007 Lincoln airbag electronics for one, is made in china. I wish we could start manufacturing all this electronics at home like it was done back in the day...Silicone Valley...
Aliexpress has it, not individual sellers. I prefer to use Paypal rather than a credit card as it is easier to open a claim.
You're quite right.
When there are too many malicious competing products in the market. When competing for low-priced products. The quality will decrease.
Because sometimes when merchants have no profit, they will use lower quality materials.
I don't buy expensive products. But I won't buy the cheapest product either.
Many times. The quality of the product is determined by the buyer.
Be a high-quality product. The price will be higher. Many times, buyers choose to purchase the cheapest product.
At this moment. Authentic merchants have no choice but to lower prices and quality at the same time.
So I won't buy expensive products. But I won't buy the cheapest product either.
NO!
That's distorted logics - to the extreme.
The designer determines the quality of the product. In this particular case you are not only determining the quality of your product but, as the designer, you should bear the R.E.S.P.O.N.S.I.B.I.L.I.T.Y for the performance of your products.
Now, first, follow the link below:
Common decency
.... try to learn something and then come back and first address technical problem I have reported in post #122 about your design L20SE.
Now, let me continue this line of reasoning.
Response of the author of L20SE illustrates beautifully why you should avoid using these boards: There is no technical support. Seller will cease to respond as soon as you push the button "BUY". Or even worse: LJM participates conversations selectively. When he sees the opportunity to increase turn-over he will respond with yet another unrelated subject. But when you face the real world problem, he will refuse to respond. Have you seen how selective he is with his response to reported problems. Silence.
Even worse: a friend of mine, member of this community, has purchased 10 boards of L20.5, eight of them have blown and he asked LJM for advice. Instead of telling him the truth that all output devices are fake, he told him that failures were my friends fault. Excuse was that he has not fastened the output devices firm enough. I will refrain from using proper juicy word for this, I will say just that the answer wasn't truthful. And not helpful too.
Schematic Shows 50 volt Rail
So current sources are more likely set to this voltage.
Or could explain issues why bias issues and DC offset
if people just choose a random voltage
or if design fault.
You're right, but let me get back to that subject later in the penultimate paragraph.
Declared values are (from the sellers' site) :
Dual power supply range :+/-DC18V- +/-DC65V. (recommned +/-50V)
There is no official schematics. Someone has reverse engineered the board and published just the recommended voltage.
I have tested the board at different voltages. At higher rail voltages the onset of oscillations just appears at higher voltage, but doesn't vanish. It is also closely related to the value of the quiescent current and the real problem starts exactly here. As soon as you attempt to attain optimal quiescent current value, the onset of oscillation appears at even lower voltages. The upper limit on the Iq is lower than the optimal. So you simply get stuck without the chance to attain optimal values. In post #53 @90scaraudio has already indicated that the maximum tolerable value of the quiescent current is 15 mA,
Thanks for pointing into the direction of reasonable explanation.
Last edited:
Of course. Like L20SE, there will be no quality degradation.
So its price will not decrease either. Even if no one buys, it won't change.
L20SE L20.5 does not use fake transistors.
Actually, all you need to do is smash the transistor open. You can tell if it's fake.
I rarely appear on forums. Because visiting DIYAUDIO in China can be very slow. Opening a page takes 30 seconds. I may not have enough time.
And I don't understand English. Actually, in China. I have many clients. I use instant messaging software to communicate with them every day. Many products sold by international agents are not purchased by anyone in China.
And LJM is my product. Sales in China are still very high. This is enough to illustrate the problem.
L20.5 is indeed more prone to damage compared to L20V10.
L20V10 uses the same KEC and ONSEMI transistors. Almost no damage.
The reason is that I used a wrench to secure the L20V10. Use super strong screws. Nut.
L20.5 installation is more difficult. In addition, the stability of CFP structure is slightly worse than that of EF structure.
Do not replace any components for L20SE.
Because there will be problems. A malfunction has occurred.
L20SE without replacing any components. There won't be any problems.