NJW3281 NJW1302 Sustained Beta from OnSemi

I just received an e-mail from the product line manager at On Semi. Here's what he had to say about these parts:

"Our roadmap for the coming year is centered on the TO-3P package we are getting ready to introduce. This portfolio of devices uses the same die as our TO-264 parts. The lower cost TO-3P package will hopefully allow us to capture more of the cost competitive mid-range consumer business without sacrificing performance"
 

taj

diyAudio Member
2005-02-23 8:49 pm
I'm a bit confused. What's the difference between the MJW (1302/3281) and NJW series? Just TO-247 vs. TO-3P? I don't see much difference between the two in terms of physical footprint. Is there an electrical difference?

Obviously the NJW is smaller than the MJL series TO-264 package, but doesn't the MJW series already provide the size advantage they're after?

Or is there more to it than I'm seeing?

..Todd
 

KSTR

Member
Paid Member
2007-07-17 2:35 am
Central Berlin, Germany
OnSemi case codes:

NJW: TO3P
MJ: TO3
MJD: DPAK4
MJE: TO220/225
MJF: TO220-ISO
MJL: TO264
MJW: TO247
NJL: TO264-5 (with integrated t-comp diodes)

An effective difference in parameters other than Vcbo/Vceo seems to be in the 0302 vs 1302 (0281 vs 3281, resp) varietes, the "0"-version are smaller die (less Pd) and have less capacitance. For same number devices vs. case style (NJW, MJL, MJW) I see no significant differences (besides Vcbo/Vceo)

- Klaus
 

megajocke

Member
2003-01-11 8:01 pm
Did anyone else notice that the Rth(j-c), power dissipation, thermal derating and maximum power dissipation don't add upp for many of the ONSemi plastic parts? (MJL21193, MJL3281...)

Rth(j-c) is stated as 0.7K/W and thermal derating as 1.43W/K. This would suggest a power rating of 180W. Or a typo as this is the same values as stated for the 250W metal cans.

Strange is that the mjw3281a datasheet says 0.625K/W (this works out to 200W) but the MJL3281 in the bigger TO264 says 0.7K/W. I have a hard time believing that the smaller package has better thermal properties... :D

I wonder too what is the advantage of TO3-P vs. TO247. Maybe it is cheaper in some way. Or maybe it's just the japanese transistor looks they are after :p
 

MJL21193

Disabled Account
2007-03-10 1:20 am

Attachments

  • 1a1.gif
    1a1.gif
    10.5 KB · Views: 1,340
megajocke said:
Did anyone else notice that the Rth(j-c), power dissipation, thermal derating and maximum power dissipation don't add upp for many of the ONSemi plastic parts? (MJL21193, MJL3281...)

Rth(j-c) is stated as 0.7K/W and thermal derating as 1.43W/K. This would suggest a power rating of 180W. Or a typo as this is the same values as stated for the 250W metal cans.

Strange is that the mjw3281a datasheet says 0.625K/W (this works out to 200W) but the MJL3281 in the bigger TO264 says 0.7K/W. I have a hard time believing that the smaller package has better thermal properties... :D

I wonder too what is the advantage of TO3-P vs. TO247. Maybe it is cheaper in some way. Or maybe it's just the japanese transistor looks they are after :p


Thank you for contacting ON Semiconductor

We have some inconsistencies in our Data sheets. We will be checking all our Audio products data sheets and we?ll correct as required.

Regards,

MartinB
ON Semiconductor
Technical Support Center
 

ozonek

Member
2006-06-12 5:10 am
Any update from ONSEMI regarding to the TO3-P vs TO247/TO264 package's thermal resistance?

Here in China a local hi-fi magazine reviewed in a recent issue the new NJW0302/0281, NJW1302/3281 and some other popular output BJT. To my surprise, NJW0203/0281 was ranked 1st based on some subjective listening tests and NJW1302/3281 was ranked 2nd. Has anyone compared the NJW series with the older MJLs, any difference in performance?
 
1) The 0302 has cob of 400pF vs the 1302's 600pF. So while the power handling gets a 25% (33%) boost, the cob is worse off by 50%.... is this a trade-off that everyone will be favorable to? My guess is it would go both ways.. i.e. ease of installing less number of devices etc.

2) When these 2 devices were tested, was the VA design essentially the same? Just looking at RAW cob numbers... it might be possible for the 1302's to load the VA more adversely than the 0302.... granted a properly designed VA would not suffer the same fate... but it is possible to have audible effects.


Anyway a buddy of mine repairs QSC's BGW's PEavey's etc. He has has that his customers consistently prefer the sound of the MJ15024/25 compared to the MJ21193/4/5/6.... this is contrary to what I would have thought.... but that's one first hand exprerience I know of... (Though a tad unrelated) ;-)
 

ozonek

Member
2006-06-12 5:10 am
Thanks K-amp for reply. From the info on the magazine, they only change the output BJTs and adjust idle current to same value. Seems no adjustment was done on the VA stage, so higher loading caused by Cob on 3281/1302 is possible for some audible effects then..

BTW, comparing high early voltage (less Ic change due to Vce change, as in 21193/21194 series) with sustained beta (less hfe down with current increase, as in 3281/1302 series) which factor is more important for linearly amplifying analog signals? Need some comment on this..thanks :rolleyes:
 
Ozonek -
that is interesting information. I am using the NJW0302/0281 in a class A amplifier running about 1.5A idle bias current, and like how it sounds.

Could you let us know what the other BJTs were?
and what the magazine is called and in what issue the test was made?



Sigurd

ozonek said:
Any update from ONSEMI regarding to the TO3-P vs TO247/TO264 package's thermal resistance?

Here in China a local hi-fi magazine reviewed in a recent issue the new NJW0302/0281, NJW1302/3281 and some other popular output BJT. To my surprise, NJW0203/0281 was ranked 1st based on some subjective listening tests and NJW1302/3281 was ranked 2nd. Has anyone compared the NJW series with the older MJLs, any difference in performance?
 

megajocke

Member
2003-01-11 8:01 pm
ozonek said:
Thanks K-amp for reply. From the info on the magazine, they only change the output BJTs and adjust idle current to same value. Seems no adjustment was done on the VA stage, so higher loading caused by Cob on 3281/1302 is possible for some audible effects then..

BTW, comparing high early voltage (less Ic change due to Vce change, as in 21193/21194 series) with sustained beta (less hfe down with current increase, as in 3281/1302 series) which factor is more important for linearly amplifying analog signals? Need some comment on this..thanks :rolleyes:

If not using triple darlington output then the higher and more linear (at low currents) beta of the sustained-beta types will be more useful for minimizing VAS loading.

When using a triple darlington this doesn't have as big importance.
 

ozonek

Member
2006-06-12 5:10 am
Sigurd,
The magazine is called "Radio and Television" (in Chinese: ÎÞÏßµçÓëµçÊÓ) published since 1958 but now devoted mostly on AV equipment review and DIY despite its original name. It is in the No.5 of 2008 issue, in which they reviewed 5 pairs of power BJTs. Below is the list of BJTs together with a score out of 10.

No.1 NJW0302/0281 ON 8.5-9.5
No.2 NJW3281/1302 ON 8-9
No.3 NJW21194/21193 ON 7-8.5
No.4 FJL4315/4215 Fairchild 7-8.5
No.5 2SC5200/2SA1943 Toshiba 7-8


Sigurd Ruschkow said:
Ozonek -
that is interesting information. I am using the NJW0302/0281 in a class A amplifier running about 1.5A idle bias current, and like how it sounds.

Could you let us know what the other BJTs were?
and what the magazine is called and in what issue the test was made?



Sigurd

 

ozonek

Member
2006-06-12 5:10 am
megajocke, thanks for reply. so the name sustained-beta implies the beta does't drop at low currents?

megajocke said:


If not using triple darlington output then the higher and more linear (at low currents) beta of the sustained-beta types will be more useful for minimizing VAS loading.

When using a triple darlington this doesn't have as big importance.