Well that has been the case for a while, but I am more interested in having complementary pairs from the same hfe group. Since KSA1381 is only available in E grade right now and KSC3503 in D (small quantities) or E (rougly 1k minimum quantity), myself and maybe/hopefully some others are interested in the opportunity to get our hands on some KSC3503E.
@stinius directed me to this thread. I just want to set the record straight.
There is still plenty of authentic Sanyo stock floating around. Sanyo produced millions of surplus transistors. The handful or so I purchased in the past year were so old that they were starting to corrode.
The KSA1381E and KSC3503D are On Semi's Korean versions after they bought Sanyo. I'm pretty sure those two particular SKUs were produced in such vast quantities that they were intended to be the same rank but the 3503s couldn't cut the mustard and were de-rated.
Are the Sanyos any good? I'm sending some samples to @Berlusconi to test against their KSC counterparts so I guess we'll see.
On Semi left, Sanyo right:
There is still plenty of authentic Sanyo stock floating around. Sanyo produced millions of surplus transistors. The handful or so I purchased in the past year were so old that they were starting to corrode.
The KSA1381E and KSC3503D are On Semi's Korean versions after they bought Sanyo. I'm pretty sure those two particular SKUs were produced in such vast quantities that they were intended to be the same rank but the 3503s couldn't cut the mustard and were de-rated.
Are the Sanyos any good? I'm sending some samples to @Berlusconi to test against their KSC counterparts so I guess we'll see.
On Semi left, Sanyo right:
Last edited:
HFE is graded at some point in the production line, most likely BEFORE the wafer is even finished processing. It’s a lot of trouble and expense to sort at final test - that would just be a go/no-go. (At my work, if a designer or manager mentions BINNING requirements, the people who have to do the work come after you with torches and pitchforks!) So what you get would be all one rank when a batch comes out, and now they’re not making as many of these types as they used to. If the process is tight, you tend to get the same rank over and over so it’s probably all that is available. It may have even been years since they ran a batch. PNP tends to run higher at low current if they both have to pass some minimum value at full current rating (the slopes of the fall off are different, and that’s just physics). The rank is typically a 2:1 spread in hFE (can be both tighter or looser), but indicates where the center of the distribution was. You CAN still match in adjacent bins in you have to. Just need to start out with more.
I’m not sure if the KSC and 2SC are “exactly” the same - KSC at least used to it was made in one of the Korea facilities. Probably the same process, but the fab will have moved. Fairchild was making some KSC types before ON acquired them - all of those from places THEY acquired. (See Brock acquire. Acquire, Brock, acquire!). For a while, ON was still making some of those types with “2SC” part numbers. For some types (all SMD stuff now) they still do.
THE solution to the hFE rank issue is to design the circuit such that it is insensitive to hFE. If you are struggling with availability NOW, think what it’s going to be like in 30 years when your amp might need a new transistor. Might be a power surge or lightning strike, or some stupid 10 uF electrolytic goes bad and takes out something else.
I’m not sure if the KSC and 2SC are “exactly” the same - KSC at least used to it was made in one of the Korea facilities. Probably the same process, but the fab will have moved. Fairchild was making some KSC types before ON acquired them - all of those from places THEY acquired. (See Brock acquire. Acquire, Brock, acquire!). For a while, ON was still making some of those types with “2SC” part numbers. For some types (all SMD stuff now) they still do.
THE solution to the hFE rank issue is to design the circuit such that it is insensitive to hFE. If you are struggling with availability NOW, think what it’s going to be like in 30 years when your amp might need a new transistor. Might be a power surge or lightning strike, or some stupid 10 uF electrolytic goes bad and takes out something else.
There are also shady manufacturers, so you still have to beware. Maybe, just maybe, somebody will buy a few thousand apiece from Rochester, and decide to resell them in onsey twoseys (or dozens) on E-bay. Unfortunately you will have no way of determining if a seller is doing that or whether they are garbage - other than buying and trying.
Hence why I dropped $300 on my beloved Sankens before On Semi moved the operation to Korea. It will require a great deal of discipline to keep some of them in the "vault" for that long.
Most of the circuits will probably work fine with slightly higher hfe mismatch, but if better matching lowers distortion you might as well hunt for those matching pairs
Edit: I bought some similar types from Taobao a few months ago and measured them just recently. Apparently I managed to stay mostly clear of fakes and have a nice selection of different types now, mostly Sanyos. But only 10 pairs or so each, lets hope the seller keeps sending out the good stuff...
Last edited: