General > General Technical Chat
LM386 on Amazon: what did I just buy?
(1/1)
Sredni:
I know, I know. They have to be fakes, especially when I can buy 20 LM386 and 20 sockets for 8 bucks. But I am a sucker for this kind of "bargains" because if the chips work reasonably well I can experiment without shedding tears on the money lost.
So, these chips were advertised as LM386 N, but came with a marking that says LM386 M-82. See picture.
I measured the DC resistance between pins 1 and 8 (should be 1.35 kohm) and I read 1.44 kohm. The default gain should be 20 and... lo and behold, it came out as exactly 20.
The ampli (at least on breadboard) is kind of fussy when it comes to decoupling and ground paths - it oscillates wildly if the connections are not short enough, but this happens with the original chips as well (at least the newest chipsd from National Semiconductors, from what I have read online).
When experimenting with a resistive 10 ohm load and a 9V batter I was able to put 6.04Vpp (2.12Vrms) without clipping across the load. That would amount to 450 mW and it stayed there unfazed for several minutes (now my poor battery is almost drained...)
Therefore, I am now curious: has anyone found what are the relevant differences between this cheap ones and the original parts? How far can they be pushed?
Are there datasheets for these Chinese clones?
I even searched noopy's site to see if he had some decapped but to no avail.
GigaJoe:
can be clone
but it definitely relabeled, who wish to buy XL386 , or XD386
just keep in mind , an official clone (frankly - if it NJM is it a clone ??? ) cost 5-7 cents. HTC and UTC about 10-15 cents JRC 25-40 C.
IMPO, TI charging a 1$ per LM386 opamp - that a 1000% margin
in term of difference, diggin to the rabbit hole they are different, can be more simplistic schematic , diff tech process, that affect input cascade, or noise , or distortions ,
it not reflecting by any CN datasheet , usually it copy- paste.
magic:
I have bought and opened over a dozen "too cheap" opamps of various types from China.
They all turned out to be clones of a few common jellybeans like LM358, with fake markings.
I have also seen clone MC34063, no-name generic LM78 regulators which worked, and Noopy has a big collection of clone NE555 of course.
And these were new chips straight from factory with fraudulent TI/ST/NS logos. Another big market are recycled chips, usually less common types.
The first problem with those parts is that there is absolutely no consistency and you never know what you get. It may be this version today, and that version tomorrow, from the same supplier. I once got a mix of two different fake NE5534 in one lot.
A second common feature of knockoff chips is that while they more or less reproduce the original circuit topology, they tend to be smaller somehow. So if there are any problems to be found, I would first expect things like lower current limits, higher internal heating, decreased long term reliability (days or years, not minutes).
Navigation
[0] Message Index
Go to full version