L20.5 AP SYS Test data My design

It is difficult to compare these two. I drove Fiat Cinquecento (midget) and Hilux (indestructible pick-up truck) and liked them both, though they belong to two quite different worlds. A60+ is quite impressive beast. Sound is different. Also A60+ has everything you need: power supply section and speaker protection. If you want the real beast, go for A60+. It sounds fantastic and can drive any speaker you want. On the other hand these LJM simple midgets let you play and learn.:)

Thank you for response. I'm actually a Circlophonist for some quite time. I like its sound qualities but I have a feeling that I can try something different and more powerful for a chance.

The last time I was checking out amplifier kits on Aliexpress/Ebay, they had a bad reputation due to the fake components and bad layout design. I'm checking out again but it seems that I'm seeing a quite different scene now. LJM kits tend to be most popular and well documented/measured ones. I think I have a little chance to regret on these ones.

There are many options to choose like those based on Accuphase, Krell, YBA designs. I really like the high quality looking of those ones with TO3 metal power transistors. A buyer confirms the authenticity of power transistors on E505 kit.
 

Attachments

  • O1CN01UwijOF2I0duJSsCPY_!!90439224.jpg
    O1CN01UwijOF2I0duJSsCPY_!!90439224.jpg
    485.9 KB · Views: 405
  • U2c75978147864aeda912a34349e3b5abQ.jpg
    U2c75978147864aeda912a34349e3b5abQ.jpg
    557.1 KB · Views: 410
...I think I have a little chance to regret on these ones.
There's nothing to regret. These midgets are just fine and I will take 3 LJMs boards and populate them with components of my own choice. I intend to use them in a tri-amping system initially conceived on NHB-108 clones. Though, these aren't as good as some "published" results indicate. You should see a square wave response below 100 Hz. Not so great.

..
There are many options to choose like those based on Accuphase, Krell, YBA designs. I really like the high quality looking of those ones with TO3 metal power transistors. A buyer confirms the authenticity of power transistors on E505 kit.
There are DIY-kits and you may replace almost everything with components you trust. I've built A60+ from a kit and have added my 5200/1930 and capacitors. I will rebuild it again, more stringently because they deserve more money.

I can understand that every seller on Aliexpress claims that their components are genuine but I doubt their understanding of the word "integrity".

If you plan to buy anything consider seriously purchasing just a board or a DIY kit at the most, otherwise your money is thrown away and you don't get the quality you've expected. That isn't prejudice, just experience.
 
Hi Berlusconi,

Just an update on the modification you suggested on the L20.5 boards well I finished rebuilding the boards today and fitted them into a box and have been running them in ten minutes on five minutes off blocks. All seems to be well but the boards definitely run hotter than they did previously. I am keeping my fingers they work reliably now.
 
Hi Berlusconi,

Just an update on the modification you suggested on the L20.5 boards well I finished rebuilding the boards today and fitted them into a box and have been running them in ten minutes on five minutes off blocks. All seems to be well but the boards definitely run hotter than they did previously. I am keeping my fingers they work reliably now.
Well done Billy,:up:

Increased temperature is normal because you have now increased the bias current to about 220 mA, hence dissipation has increased but not so significantly. For comparison, you should see Krell KSA-50, which heats like a stove.

Now, first make sure that there is no signal at the input (important) and measure voltage between the "OUT" connection and the middle pin of any of the output devices. It should be between 20 and 25 mV.

There's more to follow, but first let's see how your amplifier behaves after the re-vamp.
 
Last edited:
Hi Berlusconi,

Many thanks for the information - I am not sure I have done this correctly but it looks like one board measures 30ma and the other board 32ma. I am not sure I am using the correct setting on the multi-meter.
The measured values are close enough to roughly establish repeatability, but are you sure about the units? You should measure voltage and that is in mV. Then you may calculate current in mA by using the Ohms law. Resistance should be 0.15 Ohm if we have the same resistors.

We may also simplify this: If the heat sink heats too much then use a bit more conservative value (higher value) of the resistor that you've replaced. This would decrease the bias.

The problem is accuracy of the resistors built into the board and characteristics of other components. Therefore direct comparison of values that me and you measure is impossible. But, fortunately, we may compare the values of the measured current and observe the most obvious: temperature. If the heat sink heats too much use the resistor with higher value. But do not be afraid of 40-50 Celsius.

PS: Where do you measure?
 
Last edited:
Hi Berlusconi,

I measured the 910 ohms resistors and picked two that were exactly that figure. The temperature is so hot I can only keep my fingers on the heatsink for a maximum of three seconds. I have much bigger heatsinks available if needed but I am still waiting for delivery of the 953 and 976 resistors. Hopefully they will bring the temperature down a little.
 
Hi Berlusconi,

I measured the 910 ohms resistors and picked two that were exactly that figure. The temperature is so hot I can only keep my fingers on the heatsink for a maximum of three seconds. I have much bigger heatsinks available if needed but I am still waiting for delivery of the 953 and 976 resistors. Hopefully they will bring the temperature down a little.
You're right Billy,
Lets wait for the resistors with higher resistance. By the way, I have my board on a massive heat sink and do not have to worry about the higher dissipation.

Next, how is the sound? Mine has improved significantly at higher bias values.

I slowly understand why Chinese want us to run the board on so conservative bias: to make these small heat sinks "adequate" and to avoid destruction of "original" components. Just guessing.
 
Last edited:
Hi Berlusconi

The spare heatsinks I have are 330m x 152mm I have a feeling I will be using two of these heatsinks. One for each board. I have measured the boards correctly and as before they measure 30mv on one board and almost 33mv on the other board sorry I don't know why I said ma earlier. Sound quality was always pretty good on these boards with the only problem being they popped after a couple of weeks of use. More than one of those boards actually popped as they were turned on but not before being used for a minimum of two weeks. There is no doubt the improvement in sound quality is excellent even better than both of my Dartzeel Clones. They were a pig to get set up properly until I managed to bulk buy a lot of the flat type super caps needed to manage the bias. I also have a Quad one clone and a Quad two clone and four mono amplifiers using the L12/2 boards.
 
Billy,
30-35 mV is by all means too high so lets wait for resistors with larger resistance. This is obvious example where we can not apply the same values on two different boards because our components may differ significantly. I expect to receive all new package of components on Monday, including 0.1% resistors.

Meanwhile, I've found another possible culprit for destruction of these boards: improper grounding may drive my L12-2 into wild oscillation that can't be stopped without un-plugging the amplifier. This may happen even when you plug or un-plug the signal input connector. But, this is quite another subject to be solved.
 
Hi Berlusconi,

Yes, I knew that no matter where we bought our parts from they would probably add up to slightly different measurements but that could happen with any boards that's why the adjustable resistor is the best way to go if you are able to measure things properly. Incidentally the measured voltage on my inputs is 3mv. I had the amplifier turned on today and removed the both input plugs to measure them and plugged them back in while the amplifier was turned on - is that a no no?
 
Don't worry too much Billy,
I think we are heading into the right direction. Just look at the post cited below.
It appears that what I have empirically deduced is very well known fact "that the voltage drop across the emitter resistor has to be approximately 26mV for each of the output resistors." (Bob Cordell, "Designing Audio Power Amplifiers").
Now, I just have to (i) re-build this amplifier with the use of quality components, (ii) use two separate power supplies and (iii)perform good grounding. Otherwise this amplifier will not be so "blameless".

Good luck with selection of adequate resistor.


Hi Filthyone,

Bob Cordell explains the biasing principles in his "Designing Audio Power Amplifiers" book. Skipping all the theory behind it, the key principle for setting the optimal bias for emitter-follower type of output stage is that the voltage drop across the emitter resistor has to be approximately 26mV for each of the output resistors.

With 0.22R emitter resistors, corresponding quiescent current comes to 118mA per output pair. My practical experience shows that depending on particular design, 60-90 mA per pair is enough for suppressing crossover distortion to acceptable level.

10-15mA per pair would be enough for CFP output stage, but this is not the case in the amplifier we're talking about.

Cheers,
Valery
 
Hi Berlusconi,

Very interesting reading and to be honest I had no doubt about the direction you were going in and after using these amplifiers for the last couple of days I can only say they sound amazing. I have used a variety of different speakers and all seem to be driven to high levels without problem. Once I get the readings down to similar levels to yours I will rebuild a couple more of the boards as spares. My son and I did a brief comparison of my L12.2 amplifiers with the L20.5 and both of us agreed the L20.5 was the best.
 
Hi Berlusconi,

I hope today finds you well, here is an update on the boards after installing 953 ohm resistors in place of the 910 ohm resistors. With the 910's in place the readings were 31mv on one board and the other measured 33 mv. With the 953 ohm resistors in place the reading are board one 19mv board two measures 20mv so quite a reduction. The board with the 20mv runs a little hotter than the other board but I can now keep my hand on the heatsinks without trouble. Top work Berlusconi well done and thank you for sharing. Bill
 
Hi Billy,

The pleassure was all mine, it was inddeed a pleasant and fruitful conversation for me too. Without your initiative I couldn't have solved this dilema.

Today, I have bulit the second channel by using new components. Everything went just fine and I have made some changes.I have increased voltage to slightly above 30mV. I think it is better to belt this amplifier a bit harder to ensure complete absence of distortion. I have realised that the shape and the size of distortion is constant regardless of the amplitude of the output. This implies that distortion will become more dominant while listening at thel level of few watts and this is crucial for me because I listen at a couple of watts, usually chamber music or vocals. I simply need clear sound at low power levels.

Investment into larger heat sinks pays of whilst heating of amplifier is really rather modest.

I like neutral sound of this amplifier. It adds nothing and removes nothing, just amplifies and that's all I want from the amplifier.

Thanks for pleasant company.:)
 
Last edited:
Hi Berlusconi,

With reference to the voltage increase is that the same voltage I measured from the output on the board and the middle pin on the power transistor. I only ask because I am in the process of fitting new heatsinks which are each 3kg in weight for each channel and like you most of my listening is done at lower volume levels. With volume only going up when my son visits because he likes loud music. If that is the case I may put the 910 ohm resistors back in since the heatsinks I was using when it was getting very hot was no heavier than half a kilo. Bill
 
Hi Berlusconi,

Just an update now that the L20.5 rebuilt boards are now housed in their own box with two 3kg heatsinks. First off I think the heatsinks may be a little too big but they certainly keep the heat down. In fact I can't feel any heat at all, I turn the amplifier off every now and then to check the internal temperature of the power transistors and the resistors that previously burned with a bright blue light but all remains cool. I have completely separate power supplies and in true mono style there are no leads connecting anywhere except where the mains comes into the box on it's way to the individual power supplies. I have made very small ground stars for each board and there is absolutely no hum at all so I am very pleased with that. Measuring the boards as you instructed now shows 19mv for one board and the other 20mv so I am assuming I am in a good range to avoid any oscillation. As for sound quality lets just say I am very happy with what I am hearing.
 
Hi Billy,
I'm gad you're happy with the achieved results, me too. If we just observe the meaning of "Hi-Fi", we have achieved almost everything, there's (almost) nothing more. (according to Oxford English Language Dictionary: fidelity: "The degree of exactness with which something is copied or reproduced"). I'm not a native speaker, but I distinguish perfectly between HiFi and Hi-End. The former is just a marketing trick, a linguistic creation that means nothing and serves just to drag people into buying snake oil stuff. Regarding sound quality there's nothing more beyond the Hi-Fi. We are fortunate that our goals are reasonable and care much more for real things. Just the fact that you have a privilege to discuss your hobby with your son is more than anything else. That's an achievement we should be aware of. I also have great relationships with my daughter who is also a great person. Building these midgets is just fine addition to the quality of life we already have.

Right now I have ordered L20 V9.2 DIY board, not because it can be better. It's a midget I want to build and study, to have some more fun.

Have pleasant ongoing weekend and the May 1st holiday.