Very High-Gain PNP Transistors...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
EchoWars

EchoWars

Member
Joined 2003
...for use in differential pairs. Let's say our criteria are as follows:

1. TO-92 case (or similar). No dual parts, such as a 2SA798.
2. Usable Vceo, 65V or better.
3. Minimum Ic of 50mA.
4. Reasonable bandwidth, 100MHz min.
5. Being in current production would be nice, but otherwise reasonably easy to find.

The best I have found:
Fairchild KSA992. Consistant gains of between 410 ~ 470 @ 1mA.

Zetex ZTX795A. High gains, typically between 510 ~ 575 @1mA, with the occasional device measuring a bit over 600.

Fairchild BC556C. High gains, usually better than 500, but all over the map, from 500 to 650 or so. I always find it a pain to match a pair of devices with these.

A while back, I pulled a few TO-92 PNP transistors from an old 70's Japanese amp, and out of curiosity measured gain. Of the eight devices, the lowest one measured about 750 @ 1mA. All were fairly close to that. I meant to write down the number, but I can't now find the devices, or remember the number. I'm sure they were long out of production, but I wish I could remember what they were. Maybe the best devices were in the past...

Any devices I'm missing? Love to hear your thoughts.
 
Terry Demol

Terry Demol

Member
Joined 2002
EchoWars said:
...for use in differential pairs. Let's say our criteria are as follows:

1. TO-92 case (or similar). No dual parts, such as a 2SA798.
2. Usable Vceo, 65V or better.
3. Minimum Ic of 50mA.
4. Reasonable bandwidth, 100MHz min.
5. Being in current production would be nice, but otherwise reasonably easy to find.

The best I have found:
Fairchild KSA992. Consistant gains of between 410 ~ 470 @ 1mA.

Zetex ZTX795A. High gains, typically between 510 ~ 575 @1mA, with the occasional device measuring a bit over 600.

Fairchild BC556C. High gains, usually better than 500, but all over the map, from 500 to 650 or so. I always find it a pain to match a pair of devices with these.

A while back, I pulled a few TO-92 PNP transistors from an old 70's Japanese amp, and out of curiosity measured gain. Of the eight devices, the lowest one measured about 750 @ 1mA. All were fairly close to that. I meant to write down the number, but I can't now find the devices, or remember the number. I'm sure they were long out of production, but I wish I could remember what they were. Maybe the best devices were in the past...

Any devices I'm missing? Love to hear your thoughts.
Click to expand...

2SA970 BL are pretty good and readily available.
120V / gain 350~700 / 100mA / 100MHz. All the ones I had on hand
measured gain >400.


cheers

Terry
 
EchoWars

EchoWars

Member
Joined 2003
Terry Demol said:


No worries.

If I may ask, why do you need such high beta?
Click to expand...
Nice, high input impedence, low base currents, low offsets. :up:

CBS240 said:


Hi

Fairchild makes similar devices except they name it with a K instead of a 2. They do this with all their typical transistors.:rolleyes: Anyway, they list them as complements in the datasheet
Click to expand...
The original '2S' numbers were properly assigned to the Japanese manufacturers by the JIS (Japanese Industrial Standard) council. Fairchild chose to label them as 'KS' to distinguish them as replacement parts with similar specs without stepping on the JIS system.
 
EchoWars

EchoWars

Member
Joined 2003
homemodder said:


Wrong, here you go, 2sa1240 been using them for more than 15 years. ;)
Click to expand...
Perhaps you typed that number in wrong? The 2SA1240 is s dual part in an oddball six-pin package.
jacco vermeulen said:


Glenn,

how about the Mitsu/Pana A979H and C2259H duals with a Beta of 600-1200.
As if you've never had Pioneer or Marantz items on your bench.
Obsolete, but still easy to find and surprisingly cheap.

(for 50V : Panasonic A798G )
Click to expand...
The A979 is one I've been looking to find in quantity for a reasonable price ($1 or so). I can get pieces here and there for prices between $3.50 and $10 each, but I'm looking for something like 100 of them, and not interested in shelling out stupid prices for them. Glad I don't need them too often.

I have a bunch of the A798 F & G, and they work very well when I can use them.
 
Ian Finch

Ian Finch

Member
Joined 2010
Paid Member
I don't think so. LM394 is a multiple transistor, with many parallel transistors formed on the one substrate to achieve otherwise impossible matching by virtue of averaging the individual junction properties.

You won't get anywhere near that match quality or low drift without extremely precise matching of duals, also on the same substrate. Then again, if you don't really need the "supermatch" properties, you may find these useful, if also obsolete: http://www.analog.com/static/imported-files/data_sheets/SSM2210.pdf
http://www.analog.com/static/imported-files/data_sheets/MAT02.pdf

You can still obtain LM394H if necessary - there are a few realistic opportunities on the net with a little help from Google. However, this recently revised doc. by TI suggests some renewed manufacturing interest: http://www.ti.com/lit/an/snoa626b/snoa626b.pdf
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom