Hello there, I hope you all had a great Xmas 🙂
I am out of touch with the newer 'Flat-pack' replacements for High Power TO-3 output transistors, like MJ15003 & MJ15004.
I understand that there is small selection of the newer devices that equal or exceed the 15003's and 15004's.
Could someone please post some examples with basic spec's. Or, post a useful 'LINK'.
Thank you.
I am out of touch with the newer 'Flat-pack' replacements for High Power TO-3 output transistors, like MJ15003 & MJ15004.
I understand that there is small selection of the newer devices that equal or exceed the 15003's and 15004's.
Could someone please post some examples with basic spec's. Or, post a useful 'LINK'.
Thank you.
Have a look at the MJL21194 and NJW21194 and their compliments. I know Ic and Pd are a little lower, but if you compare the SOA curves they are very similar.
https://www.onsemi.com/pdf/datasheet/mj15003-d.pdf
https://www.onsemi.com/pdf/datasheet/mjl21193-d.pdf
https://www.onsemi.com/download/data-sheet/pdf/njw21193-d.pdf
https://www.onsemi.com/pdf/datasheet/mj15003-d.pdf
https://www.onsemi.com/pdf/datasheet/mjl21193-d.pdf
https://www.onsemi.com/download/data-sheet/pdf/njw21193-d.pdf
You will never find a flatpack rated to 250W, because Maximum Tj is lower for flatpack than TO-3 (or any metal case). It is the epoxy case itself that is the limitation. In pulsed/audio operation, the MJL21194 can handle just as much current as MJ15003 at 50V, and more above that. If there is adequate cooling they will be fine. The 20A vs. 16A current rating is rarely needed, and the real limit is how low the beta can be allowed to fall at high current. If high current (over 10A per device) is required, also consider MJL3281. Gain may be higher at 15 or 20A. Any of those devices will allow audio-frequency excursions to 20A without damage (at low vce where it occurs) but lower gain requires more of the driver and increase distortion when those currents are required. Short-circuit current limits (where it can dwell there long enough to be an issue) are generally set to lower values.
It is a complete myth that faster output transistors cause oscillation. Normally, quite the opposite - loop stability improves if the non-dominant poles are all shoved up and out of the way. In the few cases it does happen the oscillation is local and the fix is as simple as adding a base stopper, bypass cap that should have been there in the first place but wasn’t, or both. There were a couple of NAD models where MJ15003 and 2SC2565 were alternates on the BOM. Obviously, you had to use the matching PNP. The mountings are totally different and device characteristics about as different as you can get. They made a heat sink that could take either.
It is a complete myth that faster output transistors cause oscillation. Normally, quite the opposite - loop stability improves if the non-dominant poles are all shoved up and out of the way. In the few cases it does happen the oscillation is local and the fix is as simple as adding a base stopper, bypass cap that should have been there in the first place but wasn’t, or both. There were a couple of NAD models where MJ15003 and 2SC2565 were alternates on the BOM. Obviously, you had to use the matching PNP. The mountings are totally different and device characteristics about as different as you can get. They made a heat sink that could take either.
Gain is much better with Newer devices and Ft is also higher that is the good thing. Total power dissipation more limited to the flat plastic things
You wont see Max ratings of 200c like T0-3. Most will be 150c and Most no more than 150 to 180 watts
Except for a few exceptions and likely more with a search. Common Big Boy packages will be T0-264 to support 200 to 230 watt transistor
So almost to the usual 250 watt or more metal cans of yesterday. But again Gain is way way better, much less current to drive than old T0-3
Many shrunk to slightly smaller T0-247 package still good for power and little wimpy small plastic stuff, still common T0-3P-3
and million other annoying numbers I cant keep up with. Basically small power transistor obvious when you see it. And obviously not a TO-264
For a low watt driver from 8 to 15 watts metal can T0-39 moved to plastic classic BD139/40 it is still T0-126 For Higher voltage ratings
And a heavy duty 30 to 40 watt Driver you would see with a Metal Can T0-66 will now be T0-220
Anyways dont care for much more than these for power Transistors
BJT 200 Watt
Matching NPN / PNP pair from On Semi
MJL3281A
MJL1302A
BJT 230 Watt
Matching NPN / PNP pair On Semi
MJL4281A
MJL4302A
With very good gain still and High Ft
The Toshiba cheapo transistors are still Good
are very widely seen in audio amps.
BJT 150 Watt ( 130 watt cough cough)
Matching NPN /PNP Toshiba
2SC5200
2SA1943
Toshiba use to make them in Big package T0-264
No longer- Only smaller then even smaller power package.
So not really " 150 Watt" more like 130 watt
Which is what American manufacture would rate them at.
Again if you want 2SC5200 / 2SA1943 in Larger Size with more Pad area.
Or Real 150 Watt version
American Manufacture On-Semi still makes them.
Basically high power legacy Fairchild Transistors.
( or any legacy Fairchild is made by them now)
BJT 150 Watt
Matching NPN / PNP pair Fairchild / On Semi
FJL4315
FJL4215
You wont see Max ratings of 200c like T0-3. Most will be 150c and Most no more than 150 to 180 watts
Except for a few exceptions and likely more with a search. Common Big Boy packages will be T0-264 to support 200 to 230 watt transistor
So almost to the usual 250 watt or more metal cans of yesterday. But again Gain is way way better, much less current to drive than old T0-3
Many shrunk to slightly smaller T0-247 package still good for power and little wimpy small plastic stuff, still common T0-3P-3
and million other annoying numbers I cant keep up with. Basically small power transistor obvious when you see it. And obviously not a TO-264
For a low watt driver from 8 to 15 watts metal can T0-39 moved to plastic classic BD139/40 it is still T0-126 For Higher voltage ratings
And a heavy duty 30 to 40 watt Driver you would see with a Metal Can T0-66 will now be T0-220
Anyways dont care for much more than these for power Transistors
BJT 200 Watt
Matching NPN / PNP pair from On Semi
MJL3281A
MJL1302A
BJT 230 Watt
Matching NPN / PNP pair On Semi
MJL4281A
MJL4302A
With very good gain still and High Ft
The Toshiba cheapo transistors are still Good
are very widely seen in audio amps.
BJT 150 Watt ( 130 watt cough cough)
Matching NPN /PNP Toshiba
2SC5200
2SA1943
Toshiba use to make them in Big package T0-264
No longer- Only smaller then even smaller power package.
So not really " 150 Watt" more like 130 watt
Which is what American manufacture would rate them at.
Again if you want 2SC5200 / 2SA1943 in Larger Size with more Pad area.
Or Real 150 Watt version
American Manufacture On-Semi still makes them.
Basically high power legacy Fairchild Transistors.
( or any legacy Fairchild is made by them now)
BJT 150 Watt
Matching NPN / PNP pair Fairchild / On Semi
FJL4315
FJL4215
Last edited:
Toshiba is still making the C5200 in both the N and L cases. SOA-wise, the MJL3281 is better. But physically larger with proportionally higher capacitance. That’s why the 0281 was created - to get back to something closer to original Toshiba. Fairchild/ON versions of the C5200 seem to be fine. I’ve never pushed them to limits myself, but you see them some pro amps interchangeably with Toshiba. They cost more. Toshiba C5200 are by far the cheapest solution, aside from fakes and second sources that often blow up in your face.
There isn’t any reason to choose MJL4281 unless high voltage operation demands it. Nothing that the MJ15003 is suitable for requires it. Below 100V VCE, the SOA is identical to the 3281. If supply voltage is under +/-80V the 3281 is better. At higher voltage, having margin to primary breakdown is of concern. 350V capability comes at the cost of reduced beta below Vce=20. On +/-35 or 40 volt supplies typical of DIY efforts this cannot be afforded.
There isn’t any reason to choose MJL4281 unless high voltage operation demands it. Nothing that the MJ15003 is suitable for requires it. Below 100V VCE, the SOA is identical to the 3281. If supply voltage is under +/-80V the 3281 is better. At higher voltage, having margin to primary breakdown is of concern. 350V capability comes at the cost of reduced beta below Vce=20. On +/-35 or 40 volt supplies typical of DIY efforts this cannot be afforded.
