Maker Pro
Maker Pro

Transistor selection (ii)

T

Tim Williams

Jan 1, 1970
0
So for example, say I wanted a transistor:
Vcbo >=200V
Ic 0.1-2
Pc >= 10W
fT >= 100MHz
How would I go about that?

*Wading* through Digikey's stock, I found a 2SC5993 that's almost there
(180V). Onsemi has a nice parametric search, but nothing in the range. I
tried going to Fairchild's website, but their search sucks (and the whole
website isp barely in English anyway). I've had better luck skimming over
Japanese transistor tables! Too bad anything in there is probably
unobtainium.

Tim
 
J

James Arthur

Jan 1, 1970
0
So for example, say I wanted a transistor:
Vcbo >=200V
Ic 0.1-2
Pc >= 10W
fT >= 100MHz
How would I go about that?

*Wading* through Digikey's stock, I found a 2SC5993 that's almost there
(180V). Onsemi has a nice parametric search, but nothing in the range. I
tried going to Fairchild's website, but their search sucks (and the whole
website isp barely in English anyway). I've had better luck skimming over
Japanese transistor tables! Too bad anything in there is probably
unobtainium.

Tim

You might try looking for CRT driver transistors.
http://www.semiconductor-sanyo.com/search/list.asp?CLCD=94

Sanyo's 2sc4188seems to meet your requirements:
Vcbo=200V
Ic = 100mA
Pc = 10W
Ft=150MHz

http://www.semiconductor-sanyo.com/ds_e/ENN2557B.pdf

HTH,
James Arthur
 
J

Joerg

Jan 1, 1970
0
James said:
You might try looking for CRT driver transistors.
http://www.semiconductor-sanyo.com/search/list.asp?CLCD=94

Sanyo's 2sc4188seems to meet your requirements:
Vcbo=200V
Ic = 100mA
Pc = 10W
Ft=150MHz

http://www.semiconductor-sanyo.com/ds_e/ENN2557B.pdf

Yep, I was just about to post the same. Often it is cheaper and more
practical to take a few video transistors in parallel instead of one
expensive boutique version. They are really cheap. The BF series
contains some as well.
 
B

Benj

Jan 1, 1970
0
Tim said:
So for example, say I wanted a transistor:
Vcbo >=200V
Ic 0.1-2
Pc >= 10W
fT >= 100MHz
How would I go about that?

Don't know how YOU would go about that but I'd look in my "secret"
references and come up with

2SC1819
300v
15W
100Mhz
..1 Ic.

The "answer" is an old set of data books. Back in the heyday of
discrete transistors it was a subscription service. They no longer
seem to exist and even a Google search turned up nothing. They were
called D.A.T.A. Inc. And had volumes for Transisor, diodes, SCRs and
many other things. The one I have is from 1978 so they aren't too up
to date put often you can cross-reference to more modern parts.

Someone REALLY needs to put the last version of those old books on the
internet!
 
W

Winfield Hill

Jan 1, 1970
0
So for example, say I wanted a transistor:
Vcbo >=200V
Ic 0.1-2
Pc >= 10W
fT >= 100MHz
How would I go about that?

*Wading* through Digikey's stock, I found a 2SC5993 that's almost there
(180V). Onsemi has a nice parametric search, but nothing in the range. I
tried going to Fairchild's website, but their search sucks (and the whole
website isp barely in English anyway). I've had better luck skimming over
Japanese transistor tables! Too bad anything in there is probably
unobtainium.

MCM is taking orders (for later delivery) for
"subs" for Sanyo's 2SC4188 (200V 10W 150MHz) and
for Toshiba's 2sc2238b (200V 25W 100MHz), and
2sc1569 (300V 12.5W 100MHz), and for Matsushita -
Panasonic's 2sc1819 (300V 15W 100MHz).

Toshiba's 2sc2238 (160V 25W 100MHz) is in stock.
 
D

Don Lancaster

Jan 1, 1970
0
Winfield said:
MCM is taking orders (for later delivery) for
"subs" for Sanyo's 2SC4188 (200V 10W 150MHz) and
for Toshiba's 2sc2238b (200V 25W 100MHz), and
2sc1569 (300V 12.5W 100MHz), and for Matsushita -
Panasonic's 2sc1819 (300V 15W 100MHz).

Toshiba's 2sc2238 (160V 25W 100MHz) is in stock.

We have quite a few extremely hard to find 2SC and similar ic's in stock
at http://search.ebay.com/_W0QQfgtpZ1QQfrppZ100QQsassZabeja

Prices are far below anywhere else.

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com
 
T

Tim Williams

Jan 1, 1970
0
Digging through my parts box, I found exactly four 2SC1569's, which probably
came from old TV video outputs. Why it isn't three or six, I don't know...
http://harmsy.freeuk.com/nic.percival/2sc/2sc150.html
300V, 0.15A, 12.5W (awfully low for a TO-220), 40 hFE, 100 fT.

These sort of transistors seem to be hard to come by. Just what do they use
for oscilloscope deflection, anyway? Er well, used to, anyway. I suppose
they're almost all digital and sampled stuff *these* days... but back in the
old days they needed *something* good for hundreds of MHz bandwidth and
hundreds of volts deflection. I can imagine pentodes doing that, or even a
distributed amp, but there's only one tube in my Tek 475.

Speaking of HF transistors, would you happen to have SRF397 in your records
anywhere? Also, PT4576 or PT4578, CA2206, SS4556, etc. I've pulled a bunch
of RF types that don't seem to show up anywhere.

Tim

P.S. I'm thinking of modding my Heathkit IO-103. I'm thinking either
improve what's there (not much to say for it, as is..), rip it out for a
bare X-Y display (raw CRT leads ran to the front panel, or with deflection
amps of some bandwidth), or maybe try making a whole sampling 'scope. I
must say it would be impressive to hike bandwidth from 10MHz to 1GHz.
 
W

Winfield

Jan 1, 1970
0
Digging through my parts box, I found exactly four
2SC1569's, which probably came from old TV video outputs.

I can send you the datasheet for the Toshiba 2sc1569,
if you can't get it anywhere. According to the plots,
this transistor has Cob of 7pF at 10 volts, dropping to
5pF at 100V and 4pF at 300V. They don't mention Ccb,
but you'd likely drive it with a low Z signal so that
shouldn't matter.

With a Tj 150C spec and Pc = 12.5W at Tc = 25C, you can
calculate RthJC = 10C/W, to worry about your heatsink
design. Keep the heatsink-to-ground capacitance down.

p.s. The high thermal resistance is a necessary result
of this part's very small die, which is necessary to get
the low capacitance you're counting on for a video amp.
 
T

Tim Williams

Jan 1, 1970
0
Winfield said:
I can send you the datasheet for the Toshiba 2sc1569,
if you can't get it anywhere.

I've found it at DatasheetArchive, but thanks anyway.
According to the plots,
this transistor has Cob of 7pF at 10 volts, dropping to
5pF at 100V and 4pF at 300V. They don't mention Ccb,
but you'd likely drive it with a low Z signal so that
shouldn't matter.

Yeah, I was thinking cascode, or an emitter follower driver (that is to say,
almost a darlington, with excessive E1-GND quiescent current available).

I may've puttered around with this transistor before, or at least one like
it. No, the one I was messing with was TO-202 and housemarked, that's
right. In any case, it showed a nasty slowdown under, say, 30V, as if
saturation voltage was just horrendous. I see on this datasheet that hFE is
good (>50 typ. @ 25C, 1-100mA Ic), but I also see that it drops madly with
Vce on the Ic-Vce plot. It looks like a fricking pentode! Is this a result
of the small die's resistance?
With a Tj 150C spec and Pc = 12.5W at Tc = 25C, you can
calculate RthJC = 10C/W, to worry about your heatsink
design. Keep the heatsink-to-ground capacitance down.

Original in the HK scope are D40N1 transistors, which are merely 50MHz, but
3pF @ 20Vcb, half that of the C1569. I guess the moral is I get to kick up
quiescent current a good bit. Oh, we were talking about heatsinks -- they
mounted them to live heatsinks in open air, which is more or less okay
because the chassis is pretty open, but will nonetheless add a few pF.
p.s. The high thermal resistance is a necessary result
of this part's very small die, which is necessary to get
the low capacitance you're counting on for a video amp.

Ah, true. I feel weird actually running transistors at their ratings...it's
almost like I'm using tubes here...

Tim
 
P

Phil Hobbs

Jan 1, 1970
0
Tim said:
P.S. I'm thinking of modding my Heathkit IO-103. I'm thinking either
improve what's there (not much to say for it, as is..), rip it out for a
bare X-Y display (raw CRT leads ran to the front panel, or with deflection
amps of some bandwidth), or maybe try making a whole sampling 'scope. I
must say it would be impressive to hike bandwidth from 10MHz to 1GHz.

Good luck getting 1 GHz without a travelling-wave deflection structure
in the tube.

Cheers,

Phil Hobbs
 
J

JosephKK

Jan 1, 1970
0
Tim Williams [email protected] posted to sci.electronics.design:
Digging through my parts box, I found exactly four 2SC1569's, which
probably
came from old TV video outputs. Why it isn't three or six, I don't
know... http://harmsy.freeuk.com/nic.percival/2sc/2sc150.html
300V, 0.15A, 12.5W (awfully low for a TO-220), 40 hFE, 100 fT.

These sort of transistors seem to be hard to come by. Just what do
they use
for oscilloscope deflection, anyway?

Along the way the changes in o'scope CRT deflection went from simple
parallel plates to distributed deflection. This added quite a bit to
the capacitance but improved the deflection sensitivity from many
(tens of) volts per inch to (many tens of) millivolts per inch. This
reduced the required magnitude of deflection voltage to fast
transistor range. Trying to speed up a HV deflection amplifier
without resorting to ultrafast HV FET's may be impractical.
 
T

Tim Williams

Jan 1, 1970
0
JosephKK said:
Along the way the changes in o'scope CRT deflection went from simple
parallel plates to distributed deflection. This added quite a bit to
the capacitance but improved the deflection sensitivity from many
(tens of) volts per inch to (many tens of) millivolts per inch. This
reduced the required magnitude of deflection voltage to fast
transistor range. Trying to speed up a HV deflection amplifier
without resorting to ultrafast HV FET's may be impractical.

Reminds me, I think I read in my Tek475's manual something about multiple
deflection. The CRT has a number of pins along it.

And that leads to another question ;-) -- would there be any value in
MOSFETs for this? Probably cascoded, so the biasing is still handled by a
nice and fast transistor. RLC of the gate is the ultimate roadblock, isn't
it? Could still be pushed harder with excessive drive current, which I
don't really have much of a problem with, as I have a few fast, ampy
transistors on hand.

If you meant JFETs, where do you find JFETs at more than 30V, 10mA and TO-18
/ TO-92 / etc., anyway? JFETs are so unpopular, it's not fair! (Checking
Digikey, I see no more than 55V or 50mA -- 500mA is mentioned but it's got
to be a Fairchild typo.)

Tim
 
S

Simon S Aysdie

Jan 1, 1970
0
If you meant JFETs, where do you find JFETs at more than 30V, 10mA and TO-18
/ TO-92 / etc., anyway? JFETs are so unpopular, it's not fair! (Checking
Digikey, I see no more than 55V or 50mA -- 500mA is mentioned but it's got
to be a Fairchild typo.)

I don't know in particular, but check Interfet. High I is hard to
come by in a JFET (among other things).

http://www.interfet.com/

greater than 100V, but low I:
http://www.interfet.com/pdf/DS_2N6449_50.pdf
http://www.interfet.com/pdf/DS_2N6449_50.pdf
http://www.interfet.com/pdf/DS_IFN6449_50.pdf
 
J

JosephKK

Jan 1, 1970
0
Tim Williams [email protected] posted to sci.electronics.design:
Reminds me, I think I read in my Tek475's manual something about
multiple
deflection. The CRT has a number of pins along it.

I suspect that would be a tradeoff, complexity versus cost.
And that leads to another question ;-) -- would there be any value
in
MOSFETs for this?

Then, no. Now yes. Some of the specs for wideband and SHF mosfets
today are spectacular compared Tek 475 vintage.
Probably cascoded, so the biasing is still
handled by a
nice and fast transistor. RLC of the gate is the ultimate
roadblock, isn't
it? Could still be pushed harder with excessive drive current,
which I don't really have much of a problem with, as I have a few
fast, ampy transistors on hand.

Nothing wrong with that. Just remember that as speed continues to
increase, volts as well as volts/naonsecond matter to handle L issues
in a small fractions of an inch.
If you meant JFETs, where do you find JFETs at more than 30V, 10mA
and TO-18
/ TO-92 / etc., anyway? JFETs are so unpopular, it's not fair!
(Checking Digikey, I see no more than 55V or 50mA -- 500mA is
mentioned but it's got to be a Fairchild typo.)

Unpopular? They are very useful and popular in tuners and UHF/SHF/EHF
front ends, just not so much elsewhere anymore.
 
W

Winfield

Jan 1, 1970
0
Simon said:
I don't know in particular, but check Interfet. High I
is hard to come by in a JFET (among other things).

No, not so hard. You find these by looking for low Ron
JFETs, which happen to also have very high Idss. An
example is the J105, with Rds < 3 ohms and Idss = 500mA.
It's available in TO-92, sot-223 (JFTJ105) packages.

But note, for linear use one wouldn't be likely to use
a possible-high operating current, like 500mA, unless
Vds was awkwardly low, because the die would overheat.
However we certainly do use large-die JFETs in linear
appliations, namely low-noise amplifiers, because they
have very low e_n. Then we set Id = 5mA or 10mA, etc.
 
Top