Comparing LME49810, 49830 and 49811

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
We used Sprint-Layout.
This software is cheapest, very easy to use and you can export your layout to the GERBER and EXCELLON format which are supported by almost every PCB-manufacturer...

Sprint-Layout is a great product. I use it as well. My only grumble is that Sprint-Layout does not include a schematic edit/ capture/ import mechanism and that makes it much too easy to mess up a PCB layout. I did that as recent as last week!
 
I've read all 61 pages of this. I plan to build 3 channel amps on 5"x7"x2" heatsinks on the back in stainless motor drive enclosures I've salvaged. I bought 24 LME49810 before the end of production. Reading this I'm afraid building with point to point wiring on nema-CE board will lead to oscillation, as this seems to be quite a problem with these IC's. Panson_hk boards he is selling require a heatsink not available from my distributors located in the middle of the pwb, which has two channels. The other boards shown here show separate heat heat sinks for the IC also not available from my distributors. The power transistors are out at the edge fo the board. With the huge 5 HP motor drive heatsinks I have , I hadn't planned to use a fan.
Is anybody selling a PWB with IC and power transistors all located on one edge?
Is anybody else using TIP142-147 output transisotrs instead of these mythical thermal trak transistors discontinued years ago? Am I nuts not to use separate LME15032/33 drivers and a lot of separate resistors?
Tbe tube amp I intend to replace has no feedback and probably 10% distortion at full power. 0.X% distortion would be a huge improvement.
I don't have up to date windows for economic reasons, having partially to do with them obsoleting the office software every time they update. I detest the one finger interface the market is driving to. Linix schematic software so far, GEDA ****s. I imagine the layout software is as bad. Nor can I keep a printer cartridge that doesn't dry up in 10 weeks due to lack of use, nor am I going to etch anything myself using ferric chloride that legal disposal of which costs $500 a 55 gallon barrel, plus $12 a month barrel rental. So my options are purchased PWB's as is or point to point wiring executed with a soldering iron
 
sgrossklass said this about darlingtons on a thread on solid state:

Quote:
Originally Posted by OdieM
If I struggle too much I will just revert to the original design using TIP142/TIP147's for the complete project (Stereo 3-way active crossover).

If so, rather go with a Darlington made up of two discrete transistors (EF2). General purpose Darlington transistors are not well-suited to audio power amps, they have like 2 kOhm emitter resistors for the drivers when you'd rather want around 220 ohms (or even 100-120 ohms in some cases), giving very low driver current and slow base charge depletion in the outputs. This negatively affects gm linearity (which you should be concerned about with 4 ohm loads) and shoot-through at high frequencies.
close quote sgrosskloss
Which explains in a fact based way why the sample in the LM49810 datasheet may not be for hifi use. The TIP142-147 and the TO3 2N6382-87 both families have 8000 ohm resistors driver base to driver emitter, then 50 ohms driver emitter to output emitter. Many amp designs have a 220 ohm resistor between driver emitter and speaker or centerline power to bleed off base charge somewhere besides the speaker.
 
Last edited:
Hi Panson,
I hope you're still around here :)

It's about LM49811 + Thermal Trak initial serie boards, I've got the one with 2 pairs of TT.

Here is my question : I'd like to switch to non Thermal Trak power TR, can you confirm that everythings remains the same on driver board while going to non-TT on power board ?

Regards,
Nicolas
 
Hi,
I want to build stereo amplifier for home usage on lme49810, lme49811 or lme49830. After reading datasheets I concluded that 830 optimized for fets, 810 for bjt transitors and 811 for Darligtons. Also 811 has the least powerfull output, but also have smallest distortions.
From observatons 811 is the best choice for my requirements, right? However, I do not know what transistors to use with it. I found std03 from sanken and it seems they do not need Qmult, but people overally do not like Darligtons.
I do not have power supply, so I need to design it also, what voltage to choose?
 
Hello panson, I have re-read the content of pages 1-61, thank you for sharing the experience!

Although the LME49830 is designed to drive FETs, it seems that the totem pole drive circuit can also drive BJTs in principle. Considering that the second-hand LME49830 is easier to buy, so recently I tried to use the LME49830 to drive the BJT output stage.

The current output stage adopts a two-stage EF structure. The power tube uses 5 pairs of NJW0281/0302 in parallel, and the driver stage also uses NJW0281/0302. In order to improve the performance of the driver stage when using high-power tubes, I tried to use the driver stage quiescent current of 50-60mA to avoid Hfe drop and FT drop under small current. From the result, it can work normally!

8 ohm resistor load + 50W output power + 80Khz bandwidth, 1Khz THD+N: 0.0012%, 20Khz THD+N: 0.0048%. The result is not bad~

I tried to use a 12pf CC compensation capacitor. The circuit is very stable and the square wave has no overshoot. I also tried to use TMC compensation, 12pf+60pf+5.2KΩ, the circuit is still very stable, and the high frequency distortion is significantly reduced. 8 ohm resistance load + 50W output power + 80Khz bandwidth, 20Khz THD+N: 0.0025%, high frequency distortion is reduced by nearly 50%!

However, I also encountered a weird problem during the test. I used a 3uH double-layer air-core inductor as the output inductor, and connected a 3.3Ω resistor in parallel to the inductor. When I enabled it, something unexpected happened: the distortion deterioration was very obvious, especially the high frequency distortion.

I tried to change the direction of the inductor, and the result did not change; I tried to check the material of the inductor, and I confirmed that it was made of copper, there was no problem; when I short-circuited the output inductor, the distortion returned to normal. This is a strange problem, I have never encountered it before~

It is still unclear what causes the problem, is it because of the distributed capacitance of the double-layer inductor? If I can't find the reason, I'm considering canceling the output inductor. I measured the speaker cable in my hand. The 3 meter cable has an inductance of 2uH, which seems to replace the output inductance.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.