Just as the title states....good price, just bought 4
2SJ109 P CHANNEL JUNCTION IC - eBay (item 110503178874 end time Apr-04-10 20:38:40 PDT)
2SJ109 P CHANNEL JUNCTION IC - eBay (item 110503178874 end time Apr-04-10 20:38:40 PDT)
I am not claiming at all those 2SJ109s are not genuine.
There are these days MANY companies who claimed to be registered in Hong Kong. In actual fact, they are nothing more than that, a registration and an address, with no real activities in Hong Kong. The actual operation is 99.9% in China. The reason why they want a registration in Hong Kong is that HK still has law and order, and respect international laws and copy rights.
I always treat those companies as suspect. Lately we made an enquiry in HK for MAX9401. Someone in China replied by email saying he had them. But he did not even quote the right package. When we asked further, there was no reply. So we avoid any direct purchase from China.
It is a real pity, because a few bad sheep will ruin the reputation of the entire country.
Patrick
There are these days MANY companies who claimed to be registered in Hong Kong. In actual fact, they are nothing more than that, a registration and an address, with no real activities in Hong Kong. The actual operation is 99.9% in China. The reason why they want a registration in Hong Kong is that HK still has law and order, and respect international laws and copy rights.
I always treat those companies as suspect. Lately we made an enquiry in HK for MAX9401. Someone in China replied by email saying he had them. But he did not even quote the right package. When we asked further, there was no reply. So we avoid any direct purchase from China.
It is a real pity, because a few bad sheep will ruin the reputation of the entire country.
Patrick
I think it's more than a few....it seems, to me at least, to be systemic and SOP for most "audio" offerings from China that appear on ebay.
I am sorry to say that matched 2SJ74s will never be the same as a single 2SJ109 and will always be second best. It is easy to find better Idss matched 2SJ74s than in a single 2SJ109, if you have a large enough pool to match from. BUT in many applications, such as a diff pair, it is not only Idss that matters, but the entire Id vs Vgs curve. And if you have curve-traced enough of 2SJ74s and the like (and I did, more than 2000, both singles and duals), you will realise that Idss match does not always mean match over the entire working range.
On top of that, even with good thermal coupling between 2 single FETs (e.g. by using a special heatsink or silver epoxying them face to face), they are still not as well coupled as on a single silicon substrate next to each other. And He who has measured JFETs knows that they DO drift quite a bit thermally. Heard of thermally induced distortion ?
So I am afraid that while we shall have to live with single JFETs from now on, I for one still think that they are not the same as those unobtainium duals.
Just my 2 cents.
Patrick
On top of that, even with good thermal coupling between 2 single FETs (e.g. by using a special heatsink or silver epoxying them face to face), they are still not as well coupled as on a single silicon substrate next to each other. And He who has measured JFETs knows that they DO drift quite a bit thermally. Heard of thermally induced distortion ?
So I am afraid that while we shall have to live with single JFETs from now on, I for one still think that they are not the same as those unobtainium duals.
Just my 2 cents.
Patrick
Last edited:
I was going to suggest they pick up one of your cute little heatsinks to lessen some of the thermal effects...all I can say is for our diy applications we need not sweat the tiny stuff unless you want to enter the kilobuck business of high-end audio. It will be an embarassment selling it at that price-point if you didn't bother to match like you described above...(maybe for phonostages?)
Just my
0.0199.
Just my
0.0199.
I have listed two main reasons why the 2SJ109 is superior in certain applications. The first (match, especially in terns of transconductance, over working range) is by far the most important. You can argue that you don't want maximum distortion cancellation in a diff pair as a hobbyist. I tend to think many here would dispute that.
The fact that we have sold thousands of our "cute little heat sinks" will not change my mind about that. I still use my small reserve of 2SJ109s for the most critical applications.
Regards,
Patrick
The fact that we have sold thousands of our "cute little heat sinks" will not change my mind about that. I still use my small reserve of 2SJ109s for the most critical applications.
Regards,
Patrick
now apply a range of Vgs and check again like Euvl is telling you.Just to add that I do get about 0.03mA difference when testing with my garage door open...
I bet the match gets worse as the Vgs increases. Thermal coupling does not help reduce this error.
The advantages of the 109 over a matched 74s is a given and well known but it is also a given that the duals are almost extinct and it does not change the fact that the 74s will pretty soon be our only choice. This choice also dwindles as the supply dwindles as it is going the way of the duals. No argument -the diyers who can't sleep with the distortion difference can always look for 109s...and like you always say they can also have the next best thing -no argument either.
Of course...the many variables you can play with can also make you go mad.
I have no problem with anyone proposing 2SJ74s as a substitute for 2SJ109s which are next to unobtainable. There are also people who contact me by PM to buy matched JFETs. And I do explicitly state if they are Idss matched or curve tracer matched.
I do have a problem with a generalised suggestions that one gets "better match" with matched 2SJ74s than 2SJ109s. Matching means more than a single point (Idss) match, for me at least. Unless of course one states explicitly "curve tracer matched". Then, and only then, one can argue that they are, in matching terms, potentially better than a random sample of 2SJ109, which may have an Idss mismatch of up to 3%.
Patrick
I do have a problem with a generalised suggestions that one gets "better match" with matched 2SJ74s than 2SJ109s. Matching means more than a single point (Idss) match, for me at least. Unless of course one states explicitly "curve tracer matched". Then, and only then, one can argue that they are, in matching terms, potentially better than a random sample of 2SJ109, which may have an Idss mismatch of up to 3%.
Patrick
Last edited:
My Swap Meet thread specifies the jfets as Idss matched. The largest difference that I have listed is at 0.74% which of course is way better than3%...which justifies in a way my "better..." statement.
You'd probably need at least 50 sj109s to get a couple pairs that closely matched...the question is where to get that many and for how much?
Btw, how much are you selling Curve Tracer tested jfets?
You'd probably need at least 50 sj109s to get a couple pairs that closely matched...the question is where to get that many and for how much?
Btw, how much are you selling Curve Tracer tested jfets?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.