Hi,
For ksa1381 ksc3503 in E-group (if my memory doesn't fail me) you might want to look up Sonnya here on diyaudio. He might be able to help you.
Mogens
Here are the Sanyos, E group, certainly the PNP has no problem with early effect. The last 3 lines show the cluttering i taked about. Just ignore the non straight lines.
I hope this bug will be eroded in future software updates.
The next two are 2SJ74BL and 2SK170BL. the 2SJ has 10mA Idss, the 2SK has 8mA Idss.
The last picture shows 2SJ 10mA, 2SK 8ma and last 2SJ 8mA.
Sorry JJ506, i do not know you and my source has only a few for prototypes anyway.
So far i did not test the KSC, KSA, only original Sanyo.
You are in the far east and get them more easy than i do.
I hope this bug will be eroded in future software updates.
The next two are 2SJ74BL and 2SK170BL. the 2SJ has 10mA Idss, the 2SK has 8mA Idss.
The last picture shows 2SJ 10mA, 2SK 8ma and last 2SJ 8mA.
Sorry JJ506, i do not know you and my source has only a few for prototypes anyway.
So far i did not test the KSC, KSA, only original Sanyo.
You are in the far east and get them more easy than i do.
Attachments
Last edited:
Here is the last graph, i forgot to post it.
The 2SJ74BL and 2SK170BL are certainly not complimentary. That is known from research done by EUVL. The Fets can be made complimentary though by using a slightly higher Idss 2SJ74 and then degenerating it until the curves fit.
The 2SJ74BL and 2SK170BL are certainly not complimentary. That is known from research done by EUVL. The Fets can be made complimentary though by using a slightly higher Idss 2SJ74 and then degenerating it until the curves fit.
Attachments
The yellow graph is a 2SJ74 with a little less Idss then the green graph 2SK170 so part of the missmatch comes from the Idss difference. On the other hand you can see that the yellow curve has a steeper bend. 2SJ74 have higher Gm with the same amount of idle then the 2SK170.
From graph 3 post 8266 you see that the 2SK170 needs less Vds to go into saturation.
That will allow the 2SK170 to work on a voltage as low as 3V to be linear.
From graph 3 post 8266 you see that the 2SK170 needs less Vds to go into saturation.
That will allow the 2SK170 to work on a voltage as low as 3V to be linear.
Today i measured 2SA1015GR, 2SC1815GR and the 2SA970GR, 2SC2240GR pair.
The 2SA1015, 2SC1815 are good to about 6mA so somewhat better then 2N4401, 4403.
The 2SA970, 2SC2240 is really awesome. Very nice up to 12V, the maximum i can measure.
See the curves of the PNP 2SA970. Very high early voltage. That is the major problem in PNP but here it is fine.
I would say that the good reputation of Japanese audio transistors compared to the European counterparts is deserved, at least at high current.
The picture is taken with the snipping tool. Thanks Frans for that idea.
I had something called Snaggit some years ago that did the same but now that feature comes for free with newer versions of windows.
The 2SA1015, 2SC1815 are good to about 6mA so somewhat better then 2N4401, 4403.
The 2SA970, 2SC2240 is really awesome. Very nice up to 12V, the maximum i can measure.
See the curves of the PNP 2SA970. Very high early voltage. That is the major problem in PNP but here it is fine.
I would say that the good reputation of Japanese audio transistors compared to the European counterparts is deserved, at least at high current.
The picture is taken with the snipping tool. Thanks Frans for that idea.
I had something called Snaggit some years ago that did the same but now that feature comes for free with newer versions of windows.
Attachments
GR, as i remember Hfe was around 200.
I will buy more of those also in BL.
I think they are among the best. Rbb`is around 30 Ohm i think so there are some that get lower but paralleling will do the trick.
I am still searching for very low Rbb´ones that are also complimentary.
So far the Sanyos come closest-
I will buy more of those also in BL.
I think they are among the best. Rbb`is around 30 Ohm i think so there are some that get lower but paralleling will do the trick.
I am still searching for very low Rbb´ones that are also complimentary.
So far the Sanyos come closest-
Maybe this thread needs some attention http://www.diyaudio.com/forums/parts/222485-lowest-noise-bjt-transistor.html#post3220201
I measured some more candidates. The Zetex medium power SMDs do well up to 7mA. The picture shows the PNP, always the achilles heel.
I have also a surprising find : our old BC140, BC160.
Very high early voltage on the PNP and very symmetrical.
Actually here i did a little trick. The PNP had Hfe of 170 and the NPN over 300 so i adjusted the base current until they match. With this trick you can cancel distortion when the values are known. Matched for Hfe this gives a very good pair. I know from the 80th that they have decent low Rbb´too. Grundig used them in their MC heads, so did Stan Curtis and a friend of mine, a physicist. Some selection for noise is necessary here. I know that from the physicist so they do not come easy but when they play, they play just fine.
I have also a surprising find : our old BC140, BC160.
Very high early voltage on the PNP and very symmetrical.
Actually here i did a little trick. The PNP had Hfe of 170 and the NPN over 300 so i adjusted the base current until they match. With this trick you can cancel distortion when the values are known. Matched for Hfe this gives a very good pair. I know from the 80th that they have decent low Rbb´too. Grundig used them in their MC heads, so did Stan Curtis and a friend of mine, a physicist. Some selection for noise is necessary here. I know that from the physicist so they do not come easy but when they play, they play just fine.
