Gigapod said:
If I couldn't find MJL21193 / MJL21194
I would probably use
these instead:
MJ15024 and MJ15025 .... (= MJ15022/MJ15023)
OnSemi datasheets:
http://www.onsemi.com/pub/Collateral/MJ15022-D.PDF
http://www.onsemi.com/pub/Collateral/MJ15023-D.PDF
I would prefer these MJ15024/25 before MJ15003/4
MJ15024/25 should be easier to find than MJL21193/94
#########################################
So if you get my info right:
1. Try find MJL21193 MJ21194, if not found
2. try find MJ15022 MJ15023 or MJ15024 MJ15025
3. use MJ15003 MJ15004 .. they are easy to find.
#########################################
All the above 3 pairs will work well in power amplifiers.
MJL21193 will work a little bit better, in demanding circuits.
Regards
lineup
I suggest www.digi-key.com for 2SC5200/2SA1943.
You also know that you are getting genuine parts from digi-key.
You also know that you are getting genuine parts from digi-key.
Dave said:
Thank you Dave for the info. I previously ordered from digi-key and they are very reliable, the only problem being the delivery delay (small parcels sometimes take almost a month).
Good to know they have genuine 2SC5200/2SA1943. Also not expensive at all at < $4/pair in quantities >10.
Gigapod,
I checked your link, the website of "Selectronic".
I found this list of power transistors:
http://www.selectronic.fr/upload/produit/pagecatalogue/3-08.pdf
I suggest the following output devices from the list:
TIP2955/TIP3055: cheap, okey for max. 50W
TIP35C/TIP36C: slightly better then TIP2955/3055, max. +/-40V rails.
2N3055/MJ2955: bad linearity, TO-3 case. Choose TIP2955/3055 instead.
MJ15001/MJ15002 and MJ15003/MJ15004: excellent device, doesn't costs many. Very popular. Good Safe Operating Area. Two pairs from +/-50V produces 200W-RMS on 4 ohm, and they won't overstep DC-SOA even at 60° phase angle.
MJ15024/MJ15025: almost the same, but with higher max. Vceo.
TIP142/TIP147: a cheap darlington, good for max. 50-100W. Relatively good SOA, but bad linearity.
2N3442, 2N3773: old devices for quasicomplementary designs. Low Ft, bad gain linearity. 2N3773 has very good SOA up to 100V.
Drivers:
TIP29C/TIP30C and TIP31C/32C: good for max. +/-45V rails. Low Ft. For average designs.
MJE15030/MJE15031: very popular. High Ft, good gain linearity, high SOA. Hard to match.
I checked your link, the website of "Selectronic".
I found this list of power transistors:
http://www.selectronic.fr/upload/produit/pagecatalogue/3-08.pdf
I suggest the following output devices from the list:
TIP2955/TIP3055: cheap, okey for max. 50W
TIP35C/TIP36C: slightly better then TIP2955/3055, max. +/-40V rails.
2N3055/MJ2955: bad linearity, TO-3 case. Choose TIP2955/3055 instead.
MJ15001/MJ15002 and MJ15003/MJ15004: excellent device, doesn't costs many. Very popular. Good Safe Operating Area. Two pairs from +/-50V produces 200W-RMS on 4 ohm, and they won't overstep DC-SOA even at 60° phase angle.
MJ15024/MJ15025: almost the same, but with higher max. Vceo.
TIP142/TIP147: a cheap darlington, good for max. 50-100W. Relatively good SOA, but bad linearity.
2N3442, 2N3773: old devices for quasicomplementary designs. Low Ft, bad gain linearity. 2N3773 has very good SOA up to 100V.
Drivers:
TIP29C/TIP30C and TIP31C/32C: good for max. +/-45V rails. Low Ft. For average designs.
MJE15030/MJE15031: very popular. High Ft, good gain linearity, high SOA. Hard to match.
MJE15030/15031 have good SOA, high gain linearity, high Ft and quite high Vcemax, and fair price - they are most often used as drivers, but can also be used for low power outputs! If you need a larger 15-16A transistor with similar characteristics, like 2SA1943/2SC5200, two pairs of MJE15030/31 can be used instead in many cases.
It seems to me that 2sa1943/2sc5200 from Toshiba are currently best bipolar power transistors for hi-fi amplifiers. Their Ft and SOA are slightly inferior to some of Sanken bjts, but they have superior linearity and very very low capacitances.
P.S. strongest competitors are probably sanken's 2SA1215/2SC3264
P.S. strongest competitors are probably sanken's 2SA1215/2SC3264
Can you explain "bad linearity"? My NAD 3020 (original version) uses the 2N/MJ pairs. Is the TIP2955/3055 pair a drop-in replacement? Is it worth the trouble?
I am trying to decide if I should re-cap this amp, or just use it as a pre-amp for a chip-amp project. It's working fine, but definitely lacking in punch, and sounds quite compressed in the high end compared with my newer NAD 3125.
TDWesty,
NAD applied strictly matched output devices, they are performing there very well. Don't worry about them, that's an excellent amplifier with these old MJ/2N outputs.
My statement was concerned to MJ2955/2N3055 vs. 2SC5200/2SA1943.
So now there are already far better devices than that NAD used. But the genuine transistors aren't need to change.
Re-capping is a good idea for a such old amplifier.
NAD applied strictly matched output devices, they are performing there very well. Don't worry about them, that's an excellent amplifier with these old MJ/2N outputs.
My statement was concerned to MJ2955/2N3055 vs. 2SC5200/2SA1943.
So now there are already far better devices than that NAD used. But the genuine transistors aren't need to change.
Re-capping is a good idea for a such old amplifier.
sixtek said:
TIP2955/3055 and 2N3055/MJ2955 are now just same chips in different enclosures. Old "hometaxial" technology of original 2N3055 don't exist anymore from a twenty years or more and actual 2N3055 is just an equivalent device. But this equivalent is now a *real* complementary to MJ2955. Before was not so (and is hard to find criteria other than commercial for defining "complementary" the old 2N3055 with the more new MJ2955 which employ a totally different process!).
Hi
Piercarlo
AndrewT said:
The first one to mention
the really high gain and pretty fast
MJL4281 / MJL4302 in this topic was AndrewT.
They are among the best we can get
as well as their similar 'brothers': MJL3281A / MJL1302A
I did not mention those in this topic, although they were on my mind.
I did this for several reasons, to stay within the topic, as stated in first post:
- They are not as easy to find, in all countries
- I don't think they are very cheap ( I am not sure )
- They are NOT 'beginners' power transistors
... too fast high gain devices can cause oscillation troubles
... if not knowing how to deal with them properly
But if you are after some good output transistors
they are close to 'The Top of The Top' of bipolars!
As a good example, in a very good amplifier
have a look at this Michael Bittner's MikeB SymaSym circuit:
Links:
Product Data at www.onsemi.com
MJL3281A, NPN
http://www.onsemi.com/PowerSolutions/product.do?id=MJL3281A
MJL1302A, PNP
http://www.onsemi.com/PowerSolutions/product.do?id=MJL1302A
Regards
Lineup Audio Lab
http://lineup.awardspace.com/
lineup
Lineup,
You quoted MJL and MJW 0302/0281 characteristics (10% hfe matching from factory), not MJL1302/3281's
The best of all are NJL 0302/0281 series, with internal thermal tracking diode for best setup of bias current and near perfect thermal stability
See my latest hfe measurements
http://www.diyaudio.com/forums/showthread.php?postid=1107975#post1107975
Mihai
You quoted MJL and MJW 0302/0281 characteristics (10% hfe matching from factory), not MJL1302/3281's
The best of all are NJL 0302/0281 series, with internal thermal tracking diode for best setup of bias current and near perfect thermal stability
See my latest hfe measurements
http://www.diyaudio.com/forums/showthread.php?postid=1107975#post1107975
Mihai
roender said:
Thanks for info, roender
MJLxx81A and MJLxx02A
The come in a number of variations.
I have not used or measured data of this group of transistors from OnSemi.
I just tell people they exists and they may try to find out details for them selves.
Because they are good power transistors!
Thanks for any additional info.
I am still absorbing new knowledge ..... for future use
lineup
Hi gigapod
You can build good amps with 2N3055/MJ2955's and they are pretty cheap.
I've built a <.01% amp @50W using them which sounds great.
Many people think these are past their sell-by date, but I think that they are still OK as a starter.
The old hometaxial RCA 2N3055 was slow (800kHz) and had a large gain variation (non-linearity) ranging between (say) 120 at 100 mA and 30 at 4A. This was a problem in many 1970's amps which did not have enough gain to linearise the outputs with feedback. For example if your open loop gain is 2000 and closed loop 30, as defined by the feedback ratio, the actual gain will be around 29.6. But if the open loop gain changed to 8000 the closed loop gain becomes 29.92. That represents a distortion level of 1%. No wonder many amps sounded different. You can't get these RCA devices anymore, but ST and ON Semi still sell 2N3055/MJ2955's.
So you need a high open loop gain, and you should get good performance. You can use Miller compensation if you are careful about the input stage (see tcpip's forum), but JLH's method seems to be better.
The latest 2N3055 from ON Semi is similar, I suspect, to their MJ15003/4 processing. The SOA of the 2N3055 is up to 40V at 115W, and 60V at about 1A which is higher than any previous epitaxial 2N3055. And, they'll usually be fine up to +/- 38V power supplies for 60W amps, better than the SOA I have seen for TIP3055/TIP2955.
They have an fT around 2.5 MHz which is similar to MJ15003/4.
If you want to go to 70 W use MJ15003/4; otherwise the MJ21193/4 will run 100W into 8 ohms for a pair, but these are not so cheap!
cheers
John
You can build good amps with 2N3055/MJ2955's and they are pretty cheap.
I've built a <.01% amp @50W using them which sounds great.
Many people think these are past their sell-by date, but I think that they are still OK as a starter.
The old hometaxial RCA 2N3055 was slow (800kHz) and had a large gain variation (non-linearity) ranging between (say) 120 at 100 mA and 30 at 4A. This was a problem in many 1970's amps which did not have enough gain to linearise the outputs with feedback. For example if your open loop gain is 2000 and closed loop 30, as defined by the feedback ratio, the actual gain will be around 29.6. But if the open loop gain changed to 8000 the closed loop gain becomes 29.92. That represents a distortion level of 1%. No wonder many amps sounded different. You can't get these RCA devices anymore, but ST and ON Semi still sell 2N3055/MJ2955's.
So you need a high open loop gain, and you should get good performance. You can use Miller compensation if you are careful about the input stage (see tcpip's forum), but JLH's method seems to be better.
The latest 2N3055 from ON Semi is similar, I suspect, to their MJ15003/4 processing. The SOA of the 2N3055 is up to 40V at 115W, and 60V at about 1A which is higher than any previous epitaxial 2N3055. And, they'll usually be fine up to +/- 38V power supplies for 60W amps, better than the SOA I have seen for TIP3055/TIP2955.
They have an fT around 2.5 MHz which is similar to MJ15003/4.
If you want to go to 70 W use MJ15003/4; otherwise the MJ21193/4 will run 100W into 8 ohms for a pair, but these are not so cheap!
cheers
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.