Hi,
are these parts equivalent to the MJL4281A (NPN)
MJL4302A (PNP):
2SA1943
2SC5200
2SA1216
2SC2922
MJL3281A
MJL1302A
2SC3281
2SA1302
MJL0281AG
MJL0302AG
2SA1943
2SC5200
2N3055
MJ2955
Which are the best and which are still available please ? I need those for the elliot P3A amplifier to drive 8 ohm speakers with about 90w rms in each channel.
thanks
ronald
are these parts equivalent to the MJL4281A (NPN)
MJL4302A (PNP):
2SA1943
2SC5200
2SA1216
2SC2922
MJL3281A
MJL1302A
2SC3281
2SA1302
MJL0281AG
MJL0302AG
2SA1943
2SC5200
2N3055
MJ2955
Which are the best and which are still available please ? I need those for the elliot P3A amplifier to drive 8 ohm speakers with about 90w rms in each channel.
thanks
ronald
Sorry but none of those are MOSFETs. P3a is not designed to use MOSFETs.
The P3a article tells you exactly what transistors to use and what not to use, also check the articles on fake parts.
Any of what you list is OK with the exception of 3055/2905 parts which are too slow, and 2SC3281/2SA1302 which are likely to be fake.
The P3a article tells you exactly what transistors to use and what not to use, also check the articles on fake parts.
Any of what you list is OK with the exception of 3055/2905 parts which are too slow, and 2SC3281/2SA1302 which are likely to be fake.
Hi,
are you planning to use 1pair per channel?
It is usual to use 3times the output power for the active devices as a first guess, then to check the SOAR at the temperature and speaker load you plan to operate at.
ESP will have done the sums for the recommended device and recommended load.
Change at your own risk!!!
None of the alternative devices can get close to the SOA of the 4281/4302 when operating at 50Vce.
are you planning to use 1pair per channel?
It is usual to use 3times the output power for the active devices as a first guess, then to check the SOAR at the temperature and speaker load you plan to operate at.
ESP will have done the sums for the recommended device and recommended load.
Change at your own risk!!!
None of the alternative devices can get close to the SOA of the 4281/4302 when operating at 50Vce.
Considering the time and money invested it seems a shame to me to use only a single pair of output devices on 40+ rails, and not a duo in parallel.
Not even with MJL4***s, at 42Vce they only manage something like 5 Amps continuous current, and that's just at 25C. I do recall that Mr Elliot added a reference to not going into 4 Ohms loads, personally i think any power amplifier should be able for that.
A single pair of 2SA1943/2SC5200 is fine at +/-35Vdc rails.
At rail voltages above 40Vdc i'd favor 2 pairs in parallel, regardless if MJL4281/4302 or 2SA1943/2SC5200 are used.
The NPN high voltage MJL has a crummy hFe behavior at low Vce values, not one i'd favor for amps with relatively low rail voltage levels.
And unless someone is using the cheap OnSemi route, MJE15*** drivers are silly devices for such an output stage.
Toshiba drivers such as the 2SA1837/2SC4793 shown below would be a much better choice, certainly when the choice goes to Jappon output devices.
The fake Toshiba situation also seems to hinder professionals.
A guy like Rolf Gemein from Duisberg/Germany has just about had the longest track record for using Toshiba BJT output devices, starting with the RG1 power amp in 1985/86.
He apparantly was forced to switch Japanese brand a few years ago.
His RG10 integrated amp has 50vdc rails on a 4-device output stage, 2 pairs of 2SA1943/2SC5200 up till around 2004. Not sure what kind of drivers are used in the current MK4 model, the MK3 had the posted 2SA1837/2SC4793.
And at $6000 in Germany that's a serious expensive amp, even more expensive here because of the higher tax structure.
The previous model had a 430VA toroid and 60.000uF BH's caps for the output stage, plus an outboard 300VA for the preamp section. The current one has 450VA/300VA's and Sankens, $130 higher price tag.
Not even with MJL4***s, at 42Vce they only manage something like 5 Amps continuous current, and that's just at 25C. I do recall that Mr Elliot added a reference to not going into 4 Ohms loads, personally i think any power amplifier should be able for that.
A single pair of 2SA1943/2SC5200 is fine at +/-35Vdc rails.
At rail voltages above 40Vdc i'd favor 2 pairs in parallel, regardless if MJL4281/4302 or 2SA1943/2SC5200 are used.
The NPN high voltage MJL has a crummy hFe behavior at low Vce values, not one i'd favor for amps with relatively low rail voltage levels.
And unless someone is using the cheap OnSemi route, MJE15*** drivers are silly devices for such an output stage.
Toshiba drivers such as the 2SA1837/2SC4793 shown below would be a much better choice, certainly when the choice goes to Jappon output devices.
The fake Toshiba situation also seems to hinder professionals.
A guy like Rolf Gemein from Duisberg/Germany has just about had the longest track record for using Toshiba BJT output devices, starting with the RG1 power amp in 1985/86.
He apparantly was forced to switch Japanese brand a few years ago.
His RG10 integrated amp has 50vdc rails on a 4-device output stage, 2 pairs of 2SA1943/2SC5200 up till around 2004. Not sure what kind of drivers are used in the current MK4 model, the MK3 had the posted 2SA1837/2SC4793.
And at $6000 in Germany that's a serious expensive amp, even more expensive here because of the higher tax structure.
The previous model had a 430VA toroid and 60.000uF BH's caps for the output stage, plus an outboard 300VA for the preamp section. The current one has 450VA/300VA's and Sankens, $130 higher price tag.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.