Since I am (apparently) still able to get a few of the old 2SA1349 and 2SC3381 duals, I'd like to ask the forum for more information about these parts and their use. Borbely used them in the input of one of his amplifiers, a fully DC coupled one iirc. Since I am leaning towards BJT single diff input in my amplifiers, I am considering using these.
What I don't know is how they compare to something like a MAT dual or LM394, or the SSM replacement. These use dozens (or maybe hundreds?) of BJTs in parallel with interleaved construction, like a chess board, to achieve matching. Do the Toshiba parts use this technique?
Are the duals an improvement over a pair of selected BJTs (lab matched beta and vbe)?
Thanks,
Alex
What I don't know is how they compare to something like a MAT dual or LM394, or the SSM replacement. These use dozens (or maybe hundreds?) of BJTs in parallel with interleaved construction, like a chess board, to achieve matching. Do the Toshiba parts use this technique?
Are the duals an improvement over a pair of selected BJTs (lab matched beta and vbe)?
Thanks,
Alex
Last edited:
I don't know. The offset voltage specification is far worse than that of the LM394/MATsomething/SSMsomething dual transistors: 10 mV maximum, and you don't need interleaving to keep the offset voltage below 10 mV in a bipolar process. Then again, maybe the transistors are interleaved, but they didn't want to use an expensive tester, so they set the maximum offset voltage specification high to get away with a cheaper model.
The datasheets state explicitly that the transistors are on a common chip. As a result, their thermal coupling is far better than with a pair of selected transistors in separate packages. Their noise figure contours are also pretty good, by the way.
Besides, please check the datasheets, but I think the collector-base capacitances are much better (that is, smaller) than those of the LM394/MATsomething/SSMsomething.
Last edited:
Remember that back when these dual-devices were actually made, they were used in series production for commercial products. I don't think you should underestimate how much behaviour in electronics is really driven by cost and ease-of-assembly and not by absolute audiophile quality (you will see this in many other places as well
)The matching is probably "good enough" for most applications and hand-matching on a production line will have been too complicated and expensive. If you are only building a one-off you may be able to do better, but it's not really an accurate comparison