High speed, high Hfe complementary transistors?

[Tim Williams]

I'm designing an IGBT gate drive. I need a complementary pair of high gain,
high current transistors capable of switching in under 200ns (Zetex would be
perfect, but all the datasheets I've looked at seem to be in the 600ns
range).

Or (since someone's probably going to post about it whether I like it or
not) you could give a suggestion for a high/low side (half bridge) IGBT
driver rated 600V.

Through hole suggestions please.

Tim

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

 

[Fred Bartoli]

"Tim Williams" <tmoranwms@charter.net> a écrit dans le message de
news:d_35f.8024$lN2.5617@fe04.lga...

I'm designing an IGBT gate drive. I need a complementary pair of high
gain,
high current transistors capable of switching in under 200ns (Zetex would
be
perfect, but all the datasheets I've looked at seem to be in the 600ns
range).

Or (since someone's probably going to post about it whether I like it or
not) you could give a suggestion for a high/low side (half bridge) IGBT
driver rated 600V.

Through hole suggestions please.

Use the ZTX618-619/718-720 from Zetex. They are absolutly perfect for this.
The timing you're mentionning are not switching time: they include the
storage time as well. More, these times are for saturated switching, which
you won't have for a CC output stage and a capacitive load. You'll be in the
few 10s of ns range.


--
Thanks,
Fred.

 

[Tim Williams]

"Fred Bartoli"
<fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote in
message news:4354fd92$0$22055$626a14ce@news.free.fr...

Use the ZTX618-619/718-720 from Zetex. They are absolutly perfect for
this. The timing you're mentionning are not switching time: they include
the storage time as well.

Ok, but if the follower had to respond in the opposite direction in under
0.5ns, wouldn't one transistor turn on significantly faster than the other
turns off, causing shoot-through?

(Yeah, it won't, but if there were say 40nH in series with a non-resistive
gate capacitance, hmm that could never happen anyway with all the resistance
present...)

More, these times are for saturated switching,
which you won't have for a CC output stage and a capacitive load.
You'll be in the few 10s of ns range.

True, I'll be hard pressed to saturate a follower, especially for much of
the entire cycle given a capacitive load.

Thanks!

Tim

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

 

[Fred Bartoli]

"Tim Williams" <tmoranwms@charter.net> a écrit dans le message de
news:tEi5f.8682$lN2.2480@fe04.lga...

"Fred Bartoli"
fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote in
message news:4354fd92$0$22055$626a14ce@news.free.fr...
Use the ZTX618-619/718-720 from Zetex. They are absolutly perfect for
this. The timing you're mentionning are not switching time: they include
the storage time as well.

Ok, but if the follower had to respond in the opposite direction in under
0.5ns, wouldn't one transistor turn on significantly faster than the other
turns off, causing shoot-through?

(Yeah, it won't, but if there were say 40nH in series with a non-resistive
gate capacitance, hmm that could never happen anyway with all the
resistance
present...)

More, these times are for saturated switching,
which you won't have for a CC output stage and a capacitive load.
You'll be in the few 10s of ns range.

True, I'll be hard pressed to saturate a follower, especially for much of
the entire cycle given a capacitive load.

0.5ns? And for IGBT drivers? Why not asking for ps?

Do you understand that 0.5ns is 15cm light travel?

Sorry to say that, but I guess you have to get down on earth and better
understand all the details of how your bridge works, where the parasitics
are, how to wire your system to deal whith them,... Then you'll get more
realistic figures, much more than pulling them at random from your hat.


--
Thanks,
Fred.

 

[Tim Williams]

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
"Fred Bartoli"
<fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote in
message news:43566a59$0$12749$626a14ce@news.free.fr...

0.5ns? And for IGBT drivers? Why not asking for ps?

Do you understand that 0.5ns is 15cm light travel?

Oops, 0.5ľs.

...

(Unfortunately the rest of your post has no content to continue this
conversation.)

Tim

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

 

[Fred Bartoli]

"Tim Williams" <tmoranwms@charter.net> a écrit dans le message de
news:z6u5f.6$fm7.4@fe06.lga...

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
"Fred Bartoli"
fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote in
message news:43566a59$0$12749$626a14ce@news.free.fr...
0.5ns? And for IGBT drivers? Why not asking for ps?

Do you understand that 0.5ns is 15cm light travel?

Oops, 0.5ľs.

500ns... piece of cake.

Go for the Zetex.


--
Thanks,
Fred.

 

[Jim Adney]

On Wed, 19 Oct 2005 11:01:40 -0500 "Tim Williams"
<tmoranwms@charter.net> wrote:

"Fred Bartoli"
fred._canxxxel_this_bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote in
message news:43566a59$0$12749$626a14ce@news.free.fr...
0.5ns? And for IGBT drivers? Why not asking for ps?

Do you understand that 0.5ns is 15cm light travel?

Oops, 0.5ľs.

No, I think it's really much closer to 0.5 nSec.

c is approx 1 ft/nSec. 15 cm is approx 1/2 ft.

That's in vacuum, though, so it will be slower in wires, but only by
25-35%.

-
-----------------------------------------------
Jim Adney jadney@vwtype3.org
Madison, WI 53711 USA
-----------------------------------------------