Hi,
5pair of MJ21193/4 should be capable of 450W into 4ohms. 250W into 8ohms is peanuts for them. I had to push the 250W Tc to 130degC to get the 1second SOAR to just touch the 8ohm 60degree phase angle load line. Talk about boiling water!!
8pair of 200W plastic packaged anything cannot get close to those 5pair power outputs, reliably or otherwise.
5pair of MJ21193/4 should be capable of 450W into 4ohms. 250W into 8ohms is peanuts for them. I had to push the 250W Tc to 130degC to get the 1second SOAR to just touch the 8ohm 60degree phase angle load line. Talk about boiling water!!
8pair of 200W plastic packaged anything cannot get close to those 5pair power outputs, reliably or otherwise.
Obsolete
According to Onsemi the MJW0281Aand MJW0302A are obsolete.
Hi Chris,anatech said:Hi John,
Yep. From the first time I saw your avatar.
They are very good parts for sure! If you want a real treat for lower power, look at the MJW0281A and MJW0302A. Those are the best parts I have seen in a long time. They would probably make excellent drivers for a bunch of the MJ2119x series.
Hi Andrew, Joe,
I was thinking along those lines. I will say that often a home environment is harder on amps than a pro environment. At least the owners manual isn't sitting over the vents.
-Chris
According to Onsemi the MJW0281Aand MJW0302A are obsolete.
"The 21194/21193 have an Ft of about 4MHz whereas the 1281/1302 are good for about 30Mhz"
Hmmm........
Driving 56V peak at 4 ohms with two pair (as is being asked about), is 7A per device.
At 7A the Ft of the 3281 is about 3.5Mhz, same as the 21194.
Now if we were talking at 1A it would be a different story.
Hmmm........
Driving 56V peak at 4 ohms with two pair (as is being asked about), is 7A per device.
At 7A the Ft of the 3281 is about 3.5Mhz, same as the 21194.
Now if we were talking at 1A it would be a different story.
Hi QSerraTico_Tico,
Well, I'll have to order some to see if they are as good. The MJW0281A and MJW0302A were really nice parts. A hard act to follow. I had a specific application for these.
You know, the largest package and power output device isn't always the best part. These could have been the basis of one heck of a nice amplifier. The 50 ~ 100 watt class of amplifier would cover most applications.
-Chris
They are still available from Digikey. I believe they are the same die as the MJL variants. If they are obsolete, that's too bad. I liked the smaller package. I see they consider the MJW1302A and MJW3281A as replacements.
Well, I'll have to order some to see if they are as good. The MJW0281A and MJW0302A were really nice parts. A hard act to follow. I had a specific application for these.
You know, the largest package and power output device isn't always the best part. These could have been the basis of one heck of a nice amplifier. The 50 ~ 100 watt class of amplifier would cover most applications.
-Chris
The bottom line, in my opinion, is that the MJL21193/94 are excellent for the kind of use that I put mine to: a high power subwoofer amp.
For more sonically demanding applications, there are much better devices available, for the same price.
The MJL4302a/4281a that I have would be a great example. Highest voltage (350), very well matched, good SOA, high fT. Plus the convenience of the flat package.
Now to put them to use...
Chris, you have good results with the symAsym? Would like to go there with +/- 56 volts (I have a 750VA 40-0-40 toroid here as a paper weight). Change out the mpsa18 and 2n5551 to mpsa42, change the 2N5401 to MPSA92. BD135 to BD139, BC546B to BC550.
Drivers: MJE15034/35 (excellent drivers).
Paralell 2 pairs of the MJL4281A/4302A.
Thoughts? Pretty much going with what I have on hand.
Substantial power increase, but that's not important.
For more sonically demanding applications, there are much better devices available, for the same price.
The MJL4302a/4281a that I have would be a great example. Highest voltage (350), very well matched, good SOA, high fT. Plus the convenience of the flat package.
Now to put them to use...
Chris, you have good results with the symAsym? Would like to go there with +/- 56 volts (I have a 750VA 40-0-40 toroid here as a paper weight). Change out the mpsa18 and 2n5551 to mpsa42, change the 2N5401 to MPSA92. BD135 to BD139, BC546B to BC550.
Drivers: MJE15034/35 (excellent drivers).
Paralell 2 pairs of the MJL4281A/4302A.
Thoughts? Pretty much going with what I have on hand.
Substantial power increase, but that's not important.
Hi John,
-Chris
No disagreements there.
I am going to check some of these out.
Yes. The original idea was a nice little amp to drive in-wall speakers and computer speakers. These outperformed my expectations and have not been put into their intended service. They are to be experimented with. If I could make my own boards they would have been in service by this point in time.
Nice!
No, go with a cascode and retain the small devices. I am thinking J-Fets for myself. MPSA(4/9)2 are low gain (for a BJT) and slow. Not my first choice for an input pair. Should be an interesting project. Let me know how it works out. You really should built the 5.3 as a reference point.
-Chris
anatech said:
Hi Chris,
See your point about the MPSA42/92. I have in my stock also MPSA05/55 which have improved gain and fT, also a high enough voltage rating to use at +/-56V. Another possibility is BC546/56.
My interests are not with methodical, step by step reproduction of someone Else's design, but to boldly go (within reason), hoping for the best.
My objective is to use the components I have on hand for this one.
I have another project(s) happening right now, but will start this shortly.
BTW, what do you have in mind for a J-FET input? Usually not my cup of tea, but can be swayed.
Hi John,
Some 2SC2240 would be nice, these are tried and true classics for differential pairs. 2SK170 might be a good choice for a J-Fet variant. You can interchange these parts if you keep your collectors (drains) around 10 VDC. This has other advantages in keeping your diff pair cooler, always a good thing in my book. I've seen the supply run as low as 5 VDC. I like some headroom so you operate in a more linear portion of the curve, depending on the device.
-Chris
Some 2SC2240 would be nice, these are tried and true classics for differential pairs. 2SK170 might be a good choice for a J-Fet variant. You can interchange these parts if you keep your collectors (drains) around 10 VDC. This has other advantages in keeping your diff pair cooler, always a good thing in my book. I've seen the supply run as low as 5 VDC. I like some headroom so you operate in a more linear portion of the curve, depending on the device.
Well, of the designs I've seen over the years that I thought sounded really good, the FET front end seemed to be the smoothest. If you look at many of the Marantz amps from 1978 ~ 1980, Sansui, Luxman, you will find many use the FET front end design. It works well and there is still room to take them further. These transistors are normally not expensive. If you buy a bunch, I'll be happy to help you match them. Matching is the one secret weapon you have that only costs time. Same goes for a bunch of BJTs.
Understandable. Leave room in your design to change things slightly. Consider emitter degeneration resistors also. See the SymAsym thread for more informaton on that. Also on the board layout.
-Chris
Andrew T,
5 pairs for 250W . . . yes, because the amp needs to handle a reactive load with minimal intrusion of the protection circuitry - i.e. very load tolerant.
djk - sorry I don't get your numbers. I'm looking at the On MJL3281 data sheet (apologies, I stated 1281/1302 in error - it should have been 3281/1302) and for 7A for both polarity devices I get >20Mhz (Vce = 10V). Are you looking at another data sheet?
5 pairs for 250W . . . yes, because the amp needs to handle a reactive load with minimal intrusion of the protection circuitry - i.e. very load tolerant.
djk - sorry I don't get your numbers. I'm looking at the On MJL3281 data sheet (apologies, I stated 1281/1302 in error - it should have been 3281/1302) and for 7A for both polarity devices I get >20Mhz (Vce = 10V). Are you looking at another data sheet?
Bonsai said:
I said
I actually used 2sa1943/c5200 adjusted up from 150W to 200W since these are supposedly the Toshiba version of the obsolete 2sa301/281 and Onsemi's version is a high power version of the same transistor.
If you study the SOAR vs loadline of the 5pair MJ21193/4 (Tc<=200degC) compared to 8pair of 200W plastic packaged (Tc<=150cegC) you will find that the 200degC always outperforms the plastic package at elevated operating temperatures.
Note that I refer to power not speed since the thread started that way.
It's not often I manage to stay on topic so long.
It isn't uncommon for the junctions in the outputs to be running at 100*C, while the heatsink is running less than 50*C.
With a 100*C junction temperature it looks like the metal device will have twice the power handling capacity of the plastic type, due to the difference in maximum junction temperature (150* vs 200*).
With a 100*C junction temperature it looks like the metal device will have twice the power handling capacity of the plastic type, due to the difference in maximum junction temperature (150* vs 200*).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.