Searching \ for '[EE]: Protecting FETs' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=protecting+fets
Search entire site for: 'Protecting FETs'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Protecting FETs'
2002\10\10@194100 by Jinx

face picon face
In an H-bridge I'm using IRL3402 FETs. These have an internal
diode with a forward voltage of 1.3V. Should an external diode
have a forward voltage less than this, for example a power
Schottky ? If so, there's a choice of diodes with Vf ranging from
0.38V to 0.98V - how important is the magnitude of lesserness ?

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\10@202741 by Bob Blick

face picon face
On Fri, 11 Oct 2002, Jinx wrote:

> In an H-bridge I'm using IRL3402 FETs. These have an internal
> diode with a forward voltage of 1.3V. Should an external diode
> have a forward voltage less than this, for example a power
> Schottky ? If so, there's a choice of diodes with Vf ranging from
> 0.38V to 0.98V - how important is the magnitude of lesserness ?

Important to efficiency in a low voltage h-bridge. Unless you are beating
the thing real hard, the diode loss can be larger than the forward
loss. That is, if you are doing PWM.  Full on does not use the diodes.

High reverse voltage schottkys generally have higher forward drop, so if
you can choose lower voltage ones, do it. But they'll always have a lower
drop than the diode in the FET.

Cheers,

Bob

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\10@205442 by Jinx

face picon face
> > 0.38V to 0.98V - how important is the magnitude of lesserness ?
>
> Important to efficiency in a low voltage h-bridge. Unless you are beating
> the thing real hard, the diode loss can be larger than the forward
> loss. That is, if you are doing PWM.  Full on does not use the diodes.

This is for motor driving - PWM and reversibility, and protection
against spikes and stalling is critical. A few $$ extra at assembly
time is no big deal if it means no time on the service bench

> High reverse voltage schottkys generally have higher forward drop,
> so if you can choose lower voltage ones, do it. But they'll always
> have a lower drop than the diode in the FET.

The lowest Schottky in my list is the IR 42CTQ030, 30V 40A 0.38V
which I think should complement the IRL3402. The bridge will run at
either 12V or 18V (the IRL3402 is a 20V 85A 180W 10mohm device)

As for stalling protection / detection. I know some people measure
voltage drop across a resistor in the supply line, but wonder if a
current sensor, eg Hall Effect, is a viable alternative. Not knowing
the absolute current through the motor isn't as important as knowing
when it's passed a safe limit so the drive can be shut down

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\11@121655 by Roman Black

flavicon
face
Jinx wrote:
>
> This is for motor driving - PWM and reversibility, and protection
> against spikes and stalling is critical. A few $$ extra at assembly
> time is no big deal if it means no time on the service bench

> The lowest Schottky in my list is the IR 42CTQ030, 30V 40A 0.38V
> which I think should complement the IRL3402. The bridge will run at
> either 12V or 18V (the IRL3402 is a 20V 85A 180W 10mohm device)

Hi Jinx, so you're building a 40 amp PWM'ed
h-bridge for the first time? I hope you are going
to do testing on resistive load at reduced current
to start with, then add your inductive load in
series with your resistive load for further testing
and I suggest a real heavy duty load snubber to start
with, gradually reducing the snubber size as you
are confident that you are handling all recirculating
current paths etc and allowing for load accel and decel.
:o)

Re the schottky's, WES in Aust have;
S16C60A, TO220, $1.95 AUD, dual diode 2x 16A, 60v
S30D60C, TOP3, $2.55 AUD, dual diode 2x 30A, 60v

I have used these for some time, they are cheap and
quite rugged, Vf is good depending on what currents
you have. The dual diodes are available in both configs
which may help reduce package count.


> As for stalling protection / detection. I know some people measure
> voltage drop across a resistor in the supply line, but wonder if a
> current sensor, eg Hall Effect, is a viable alternative. Not knowing
> the absolute current through the motor isn't as important as knowing
> when it's passed a safe limit so the drive can be shut down

Since you don't need instant (ie PWM) current
sensing you can use a coil of heavy wire wrapped
around a reed switch. :o)
-Roman

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@134446 by Peter L. Peres

picon face
On Thu, 10 Oct 2002, Bob Blick wrote:

*>High reverse voltage schottkys generally have higher forward drop, so if
*>you can choose lower voltage ones, do it. But they'll always have a lower
*>drop than the diode in the FET.

Why not sense reverse voltage drop across FET diode and turn the FET on
with an opamp that overrides the bridge driver ?

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@175645 by Jinx

face picon face
> Hi Jinx, so you're building a 40 amp PWM'ed
> h-bridge for the first time?

Yes - I'm about to be deflowered. I've made a few H-bridges before
but not ones that could catch fire or make loud noises. I'm expecting
a bang or two. Hope it won't happen of course, but if you expect the
worst, things can only get better. That's why I'll spend a little extra
time at the design stage rather than rushing into it with crossed fingers

> I hope you are going to do testing on resistive load at reduced current
> to start with, then add your inductive load in series with your resistive
> load for further testing

Absolutely. It's going to take a little while to figure out what PWM curve
works best and I wouldn't be trying that with the full load

> and I suggest a real heavy duty load snubber to start
> with, gradually reducing the snubber size as you
> are confident that you are handling all recirculating
> current paths etc and allowing for load accel and decel.
> :o)

There's a snubber in there now, but I've yet to find out how spiky
the motor is and stuff. Part of the exercise is stresslessening by
using accel and decel routines

> Since you don't need instant (ie PWM) current
> sensing you can use a coil of heavy wire wrapped
> around a reed switch. :o)
> -Roman

I can test that without the H-bridge, ie power the motor straight
from the battery and stall it. Then figure out why it didn't work with
the bridge and it all went kablooey ;-)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@175655 by Jinx

face picon face
> *>High reverse voltage schottkys generally have higher forward drop,
> so if *you can choose lower voltage ones, do it. But they'll always have
> a lower *>drop than the diode in the FET.
>
> Why not sense reverse voltage drop across FET diode and turn the
> FET on with an opamp that overrides the bridge driver ?
>
> Peter

Are you saying use the FET to soak up the back EMF that the diodes
would have ?

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@181306 by Bob Blick

face picon face
Hi Jinx,

If you only care about protecting the FETs, not measuring the actual
current, why not measure the voltage drop when on?

For inspiration, check out my "fighting robot" brain:
http://www.bobblick.com/techref/projects/sv2hb/sv2hb.html

It uses a PIC, does PWM, drives two h-bridges, senses the voltage drop on
the output devices, limits the current. Almost no parts. Very robust.

Cheers,

Bob

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@185959 by Peter L. Peres

picon face
On Sat, 12 Oct 2002, Jinx wrote:

*>> *>High reverse voltage schottkys generally have higher forward drop,
*>> so if *you can choose lower voltage ones, do it. But they'll always have
*>> a lower *>drop than the diode in the FET.
*>>
*>> Why not sense reverse voltage drop across FET diode and turn the
*>> FET on with an opamp that overrides the bridge driver ?
*>>
*>> Peter
*>
*>Are you saying use the FET to soak up the back EMF that the diodes
*>would have ?

Yes. The FETs conduct backwards just fine when driven on. This is what is
used for active rectification. Since this is a first project you should
use conventional techniques imho ...

The conducting FET trick I've seen used in non-PWM bridge drivers where a
large inertial load is attached (small electric vehicle). Maybe it would
work with PWM but it would have to be fast. Usually higher power
electronics tend to go down in frequency.

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@231317 by Jinx

face picon face
> *>Are you saying use the FET to soak up the back EMF that the diodes
> *>would have ?
>
> Yes. The FETs conduct backwards just fine when driven on. This is
> what is used for active rectification. Since this is a first project you
> should use conventional techniques imho ...

imvho so do I ;-)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\11@233849 by Jinx

face picon face
> For inspiration, check out my "fighting robot" brain:
> http://www.bobblick.com/techref/projects/sv2hb/sv2hb.html

This is the circuit I'm thinking of using

http://home.clear.net.nz/pages/joecolquitt/FETbridge.html

Not shown is a charge pump (probably MAX232) to get 2 x V+

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\12@040930 by Peter L. Peres

picon face
On Sat, 12 Oct 2002, Jinx wrote:

*>http://home.clear.net.nz/pages/joecolquitt/FETbridge.html

Problems I can see:

If you chop the high side you will need very fast optocouplers. Chop only
the low side (leave the high side transistor on).

The high side drivers supplies are a little strange.

You sure hate decoupling capacitors.

May I ask what you used to draw the circuit ? (Looks like what I use ... )

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spam_OUTlistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@050201 by Jinx

face picon face
> *>http://home.clear.net.nz/pages/joecolquitt/FETbridge.html
>
> Problems I can see:
>
> If you chop the high side you will need very fast optocouplers. Chop
> only the low side (leave the high side transistor on)

I've been looking at Hewlett Packard's high speed TTL range. The
standard 4N25 opto-transistor is around 2us and 20% CTR, whereas
the HP (and others) data optos can be 40ns and 500% CTR. There's
a limit to the response time of the FET's gate but rather the opto be
too fast than too slow to keep the switching as hard as possible

> The high side drivers supplies are a little strange.

It's an emitter follower driver for current gain

> You sure hate decoupling capacitors

They're on the PCB, honest. The 2 x V+ part of the diagram isn't shown
and that's where they're drawn

> May I ask what you used to draw the circuit ? (Looks like what I use ... )

http://home.clear.net.nz/pages/joecolquitt/0symbols.html

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam@spam@mitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@063616 by Peter L. Peres

picon face
On Sat, 12 Oct 2002, Jinx wrote:

*>> May I ask what you used to draw the circuit ? (Looks like what I use ... )
*>
*>http://home.clear.net.nz/pages/joecolquitt/0symbols.html

Oh, it's not what I use. ;-)

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamKILLspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@072140 by Jinx

face picon face
> Oh, it's not what I use. ;-)
>
> Peter

Oh, and to gif it I use Paint Shop Pro. I think giffing is still a
proprietary algorithm, had no success finding a stand-alone
utility. If anyone knows of a freeware/public domain small
program that does giffing I've a few friends that could use it.
And also it would help the list, as jpg attachments of diagrams
get broken up pretty badly and are larger than they need to be

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam.....mitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@083131 by Roman Black

flavicon
face
Jinx wrote:
>
> >http://home.clear.net.nz/pages/joecolquitt/FETbridge.html


Hi Jinx, do you really need the push pull drivers?
A large motor like that will probably suit lower freq
PWM and FET turn-on/off time won't be as critical.
Its not like you need to run the thing at 50kHz or
anything, for best performance you may end up at a
few kHz. What's the motor inductance?
Likewise you won't need to pwm the top fets, only the
bottom ones so you may be able to source P-FETs for the
top two and simplify the driving a lot.
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email EraseMElistservspam_OUTspamTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@084420 by Dale Botkin

flavicon
face
On Sun, 13 Oct 2002, Jinx wrote:

> If anyone knows of a freeware/public domain small
> program that does giffing I've a few friends that could use it.

I've switched completely to png.  So far I haven't found anyone who can't
view them with common tools (like a web browser, if nothing else).

Dale

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@085255 by Roman Black

flavicon
face
Dale Botkin wrote:
>
> On Sun, 13 Oct 2002, Jinx wrote:
>
> > If anyone knows of a freeware/public domain small
> > program that does giffing I've a few friends that could use it.
>
> I've switched completely to png.  So far I haven't found anyone who can't
> view them with common tools (like a web browser, if nothing else).


My web browser doesn't view .png files. I have
to save them to hard disk and view them with a
freeware utility later. The image editors I paid
for years ago and use constantly as they have great
features won't load or save .png files. :o(
-Roman

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email @spam@listservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@092924 by Jinx

face picon face
> Hi Jinx, do you really need the push pull drivers?
> A large motor like that will probably suit lower freq
> PWM and FET turn-on/off time won't be as critical.

This is as much an exercise for me as much as anything else. I
doubt whether I'l have the time to become an "expert", especially
with analogue circuits, but hopefully by trying things like this I can
at least be less incompetent than I was yesterday. As you say, for
this particular motor and larger the PWM freq could be fairly low,
but I'd still like to explore a "universal" H-bridge, one that's adaptable
to any motor, large or small

> Its not like you need to run the thing at 50kHz or
> anything, for best performance you may end up at a
> few kHz. What's the motor inductance?
> Likewise you won't need to pwm the top fets, only the
> bottom ones so you may be able to source P-FETs for the
> top two

My understanding is that P FETs are less common and grunty ($ for $)
than N FETs, hence the reason for trying N all round. If, as you and Peter
suggest, the top pair need only on/off, then there's a possibility of using
relays (or even bipolar, SCR ?), assuming that they are economic and
up to it. If some on/off device were to be used, the PWMing could start
after the top device has been turn on, which I guess would minimise
start-up current effects in it

> and simplify the driving a lot.
> Roman

For the circuit or the programmer ? Assuming that opposing FET drives
are connected by inverters to give two-wire control, would that necessarily
add to the complexity ? What advantages are there keeping 4-wire control ?

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email KILLspamlistservKILLspamspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@095706 by Alan B. Pearce

face picon face
>This is the circuit I'm thinking of using
>
>http://home.clear.net.nz/pages/joecolquitt/FETbridge.html
>
>Not shown is a charge pump (probably MAX232) to get 2 x V+
>

I suspect you are making a rod for your own back ... or one of the many
similar expressions.


Get down to RS and buy an Intersil HIP4080. It has a built in charge pump
for the upper FET's, and all sorts of other goodies. It also has a built in
PWM oscillator section so you need only feed in a DC voltage, but it can be
fed with an externally generated PWM if preferred.

There are all sorts of useful app notes and data sheets on the Intersil web
site on the HIP408x family.

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservTakeThisOuTspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@140607 by Peter L. Peres

picon face
On Sun, 13 Oct 2002, Jinx wrote:

*>> Oh, it's not what I use. ;-)
*>>
*>> Peter
*>
*>Oh, and to gif it I use Paint Shop Pro. I think giffing is still a
*>proprietary algorithm, had no success finding a stand-alone
*>utility. If anyone knows of a freeware/public domain small
*>program that does giffing I've a few friends that could use it.
*>And also it would help the list, as jpg attachments of diagrams
*>get broken up pretty badly and are larger than they need to be

Send tiff or pdf. tiff compresses very well for bw line art.

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email spamBeGonelistservspamBeGonespammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@162117 by Russell McMahon

face
flavicon
face
The fabulous & FREE Irfanview produces GIFs and many other formats.
The rot has set in however as he is charging for JPEG 2000 above 640 x 480
resolution.
GIFs and jpegs settable in size/quality from medium/excellent to very very
tiny/pathetic.

       http://www.irfanview.com

> If anyone knows of a freeware/public domain small
> program that does giffing I've a few friends that could use it.
> And also it would help the list, as jpg attachments of diagrams
> get broken up pretty badly and are larger than they need to be

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email TakeThisOuTlistservEraseMEspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@172629 by Jinx

face picon face
> >Not shown is a charge pump (probably MAX232) to get 2 x V+
>
> I suspect you are making a rod for your own back ... or one of the many
> similar expressions.
>
> Get down to RS and buy an Intersil HIP4080.

$69 ???? Doubt it. However, I will look around for a better price. I had
considered using the ICM7667 driver

> There are all sorts of useful app notes and data sheets on the Intersil
> web site on the HIP408x family

Which I shall cast my beady eye over, thanks for the reference (RS
notwithstanding)

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email RemoveMElistservspamTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@210905 by Mike Singer

picon face
I've downloaded it.
Works just _phabulously.

Thank you.
Mike.


Russell McMahon wrote:
> The fabulous & FREE Irfanview produces GIFs and many other formats.
> The rot has set in however as he is charging for JPEG 2000 above 640 x
480
> resolution.
> GIFs and jpegs settable in size/quality from medium/excellent to very
very
> tiny/pathetic.
>
>         http://www.irfanview.com
>
> > If anyone knows of a freeware/public domain small
> > program that does giffing I've a few friends that could use it.
> > And also it would help the list, as jpg attachments of diagrams
> > get broken up pretty badly and are larger than they need to be

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservEraseMEspam.....mitvma.mit.edu with SET PICList DIGEST in the body


2002\10\12@223719 by Mike Singer

picon face
Another interesting product:
Softuarium WebPic Deluxe ActiveX control.
http://www.softuarium.com
(Have no idea who are these guys)
Could be build in VB project for example.
Could open or capture, resize convert to many formats etc.
There is a VB example project.
Installation is about 700k, but when compiled, project may
be very small.
Free evaluation version of the control is fully functional.

Mike.


Russell McMahon wrote:
> The fabulous & FREE Irfanview produces GIFs and many other formats.
> The rot has set in however as he is charging for JPEG 2000 above 640 x
480
> resolution.
> GIFs and jpegs settable in size/quality from medium/excellent to very
very
> tiny/pathetic.
>
>         http://www.irfanview.com
>
> > If anyone knows of a freeware/public domain small
> > program that does giffing I've a few friends that could use it.
> > And also it would help the list, as jpg attachments of diagrams
> > get broken up pretty badly and are larger than they need to be

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email EraseMElistservspammitvma.mit.edu with SET PICList DIGEST in the body


2002\10\13@014340 by Peter L. Peres

picon face
On Sun, 13 Oct 2002, Jinx wrote:

*>For the circuit or the programmer ? Assuming that opposing FET drives
*>are connected by inverters to give two-wire control, would that necessarily
*>add to the complexity ? What advantages are there keeping 4-wire control ?

4-wire control is a must unless you use a dedicated FET driver chip. The
reason is that with 2 wires and no special precautions the FETs in a group
will still be on when the others switch in. The result is a destroyed
bridge. With 4 wire you can program a MCU to drive the phases in correct
sequence. For example I would be tempted to use two general purpose IOs to
drive one each of the top transistors and the PWM output through two and
gates to the lower two transistors. The and gates would gate the pwm to
the correct lower transistor using the drive for the upper transistors as
select. The MCU can control the sequence like this. This scheme allows all
the control to be done in software (including active braking f. ex.).

As a safety measure I'd add a shunt type current sensor and connect its
comparator output through an extra and gate to the PWM directly. This and
a 50A resettable breaker (or bimetal current protector) should avoid
surprises caused by a stuck MCU.

Jinx, beware of the grounds. Your lower switches share ground with the MCU
control. At the usual bridge currents (10-20A for this kind of design)
you only need 0.25 ohms in the grounds to lift the lower FETs enough that
they will partially close, even if nothing else goes wrong.

Peter

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestEraseMEspamEraseMEmitvma.mit.edu


2002\10\13@014349 by Peter L. Peres

picon face
On Sat, 12 Oct 2002, Dale Botkin wrote:

*>On Sun, 13 Oct 2002, Jinx wrote:
*>
*>> If anyone knows of a freeware/public domain small
*>> program that does giffing I've a few friends that could use it.
*>
*>I've switched completely to png.  So far I haven't found anyone who can't
*>view them with common tools (like a web browser, if nothing else).

Older browsers on Wintel do not do png and png has no internal
compression, you need to zip the files. Apart from that it's an open
standard. tiff is still better for line art etc.

Peter

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspam_OUTspamKILLspammitvma.mit.edu


2002\10\13@042357 by Jinx

face picon face
> 4-wire control is a must unless you use a dedicated FET driver chip

I've looked through the pdf for the HIP4080 and AN9404 for it. There's
a lot of info in there about sequencing, dead times and propagation
delays that are relevant to what you've said. Even if I get an awful price
off Arrow tomorrow and don't use this chip, I think there's a lot to be
learned from Intersil's examples

> As a safety measure I'd add a shunt type current sensor and connect its
> comparator output through an extra and gate to the PWM directly. This
> and a 50A resettable breaker (or bimetal current protector) should avoid
> surprises caused by a stuck MCU.

Noted

> Jinx, beware of the grounds. Your lower switches share ground with the
> MCU control. At the usual bridge currents (10-20A for this kind of design)
> you only need 0.25 ohms in the grounds to lift the lower FETs enough that
> they will partially close, even if nothing else goes wrong

The first motor I'll be trying this out on uses 0.9A when running, which
shouldn't tax the circuit too much, but I have some 1/4HP motors that
will - maybe too much and the circuit would need beefing up.  I'll be
keeping power tracks short and thick

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspamspammitvma.mit.edu


2002\10\13@085838 by Alan B. Pearce

face picon face
>$69 ???? Doubt it. However, I will look around for a better price. I had
>considered using the ICM7667 driver

What???? The HIP4080AIP is GBP7.87 and the HIP4080IP is GBP10.51 on the RS
CD I have. Now even when multiplying by 3.3 to get NZ$ that is still
considerably less than what you quote.

The HIP4080AIB surface mount version is also GBP7.43.

maybe you better print this off and wander round to them. I know RS are
rarely the cheapest on the block, but that is of the order of twice what it
should be :))

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamspamspamBeGonemitvma.mit.edu


2002\10\13@090746 by Alan B. Pearce

face picon face
>Even if I get an awful price off Arrow tomorrow and don't use this
>chip, I think there's a lot to be learned from Intersil's examples

I don't know if you have EBV there Jinx, but they are the folk we dealt with
in the UK to get this chip here at work. This was before RS started stocking
it. We even got some free samples off them, but then we were in the market
for some expensive rad hard versions, so I suppose they figured it was worth
it. :))

I'm hoping there are still some samples around to fall into my lap when the
project finishes. :)))

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2002\10\13@100203 by Dale Botkin

flavicon
face
On Sat, 12 Oct 2002, Roman Black wrote:

> My web browser doesn't view .png files. I have
> to save them to hard disk and view them with a
> freeware utility later.

Really??  I'm surprised.  What are you using for a web browser?

Dale

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestSTOPspamspamspam_OUTmitvma.mit.edu


2002\10\13@110702 by Paul Hutchinson

flavicon
face
PNG has the excellent compression built in no need to zip the files, see
http://www.libpng.org/pub/png/

Paul

> {Original Message removed}

2002\10\13@135811 by Peter L. Peres

picon face
On Sun, 13 Oct 2002, Paul Hutchinson wrote:

*>PNG has the excellent compression built in no need to zip the files, see
*>http://www.libpng.org/pub/png/

Uhh, my bad. I confused png with ppm. I retract everything I said about
png (excepting the fact that older Wintel browsers do not support it).

Peter

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestSTOPspamspamEraseMEmitvma.mit.edu


2002\10\13@173049 by Jinx

face picon face
> >$69 ???? Doubt it. However, I will look around for a better price.
> > I had considered using the ICM7667 driver
>
> What???? The HIP4080AIP is GBP7.87 and the HIP4080IP is
> GBP10.51 on the RS CD I have. Now even when multiplying by
> 3.3 to get NZ$ that is still considerably less than what you quote

I have to admit to a fleeting visit by Mr Cock-up. The price I gave
was for the SMT version. The DIP is $18 + tax, which is acceptable
for a prototype, and I'd be quite keen to try it at that price

Arrow gave me a price of $7+. Bad news - 18 weeks lead time
and 1500 MOQ. They have the HIP4082 (1.5A version of the 2.5A
4080) for $13+ in lots of 48

> The HIP4080AIB surface mount version is also GBP7.43
>
> maybe you better print this off and wander round to them.

Er, perhaps not. I'd feel a little like Neville Chamberlain waving
"that" piece of paper about after his meeting with Herr Hitler

> I know RS are rarely the cheapest on the block, but that is of the
> order of twice what it should be :))

For the SMT version that seems to be true. And also I'm noticing
errors in their discrete semis tables. Makes it tricky to pick something
without checking it out with the manufacturer. Just one example, some
ZTX transistors are shown as having 10V Vce, in fact it's more like
0.36V

--
http://www.piclist.com hint: To leave the PICList
KILLspampiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu


2002\10\14@022831 by Peter L. Peres

picon face
On Mon, 14 Oct 2002, Jinx wrote:

*>errors in their discrete semis tables. Makes it tricky to pick something
*>without checking it out with the manufacturer. Just one example, some
*>ZTX transistors are shown as having 10V Vce, in fact it's more like
*>0.36V

Vceo and Vcesat are different parameters. There is no parameter called Vce
I know of except if listed in context (e.g. operating point = Vce=10V
Ic=2mA Gain=18dB Nf=...). Maybe the numbers you are trying to compare
aren't.

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@042536 by Jinx

face picon face
> Vceo and Vcesat are different parameters. There is no parameter
> called Vce I know of except if listed in context (e.g. operating point =
> Vce=10V Ic=2mA Gain=18dB Nf=...). Maybe the numbers you are
> trying to compare aren't
>
> Peter

I appreciate what you're saying, which is why the RS tables are not
thoroughly helpful. Zetex's pdf for eg the ZTX450 show Vceo = 45V
and Vcesat = 0.36V. No mention of Vce, and AFAICT that term is
not explained by RS, so in fact it's meaningless (to me anyway). I
already knew that Zetex types have lower saturation voltage than
other types (which is why I wanted them), it would have been nice
to confirm it at a glance or make comparisons. I should e-mail RS
and find out exactly what they mean by Vce

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@070918 by Russell McMahon

face
flavicon
face
> I should e-mail RS and find out exactly what they mean by Vce

They won't know. When the new catalogue came out some while ago I rang the
local manager to point out that many of the tables had utterly meaningless
column headings or, in many cases where they had meaning, utterly useless
column headings, and that the new ways of sorting many devices made it great
if you knew what part number you wanted but utterly useless if you were
searching for a new part by its parameters. (I wasn't just being cranky - I
thought they might be interested in knowing that somebody who had no idea
whatsoever about electronics had rearranged the catalogue). He said that
many people had said how much better the new catalogue was and that I was
the only person who didn't like it. I really can't imagine what to think
about people's preferences. (I'm sure Jinx will have some ideas :-) ). I'm
sure that Farnell would have been delighted when they saw it.

FWIW Vce simple means the voltage between collector and emitter. As the
conditions are not specified there is no other meaning. It's similar to the
Quantum Mechanics Copenhagen interpretation of what happens when no-one is
looking or like the sound of one dog barking but not as loud.


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@073829 by Jinx

face picon face
> > I should e-mail RS and find out exactly what they mean by Vce
>
> They won't know.

I e-mailed the UK office anyway. When / if I get a reply I'll pass it on

> the new ways of sorting many devices made it great if you knew what
> part number you wanted

I played Monty Python's The Cheese Shop with the Memec sales rep
today. Looking for better prices I read IR FET numbers from the RS
catalogue to him one after another. His answers were all along the
lines of "Not today sir, today the van broke down" and "Not much call
for it round here sir". It was almost amusing except I was getting peed
off because there seemed to nothing on the shelves. The least he
could have done was put me on hold to some bazouki music. I like
a nice tune

> He said that many people had said how much better the
> new catalogue was

The tables show Vce in the 1994 catalogue too

> and that I was the only person who didn't like it

I'm sure you've come across this before. Something obviously wrong
but "you're the first to mention it". IOW, other people just don't bother

> I really can't imagine what to think about people's preferences. (I'm
> sure Jinx will have some ideas :-) ).

Well, I got caught out thinking they meant Vcesat, even though that
didn't make sense either, so not entirely happy about that. Although
whenever possible I get mftrs data for the part so it's something of
nothing really

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@074034 by Olin Lathrop

face picon face
> Vceo and Vcesat are different parameters. There is no parameter called Vce
> I know of except if listed in context (e.g. operating point = Vce=10V
> Ic=2mA Gain=18dB Nf=...). Maybe the numbers you are trying to compare
> aren't.

The Vce spec probably means the highest voltage the transistor can withstand
when off.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@075111 by Jinx

face picon face
> The Vce spec probably means the highest voltage the transistor can
> withstand when off

Some table samples

BC549      Vceo 30    Vce 5
2SC3329 Vceo 80     Vce 0.1
TIP50        Vceo 400   Vce 10

????????????

All academic, only whoever did the RS tables knows for sure

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@081640 by miked

flavicon
face
If you mean covert to/save as, check out Irfanview.
> Oh, and to gif it I use Paint Shop Pro. I think giffing is still a
> proprietary algorithm, had no success finding a stand-alone
> utility. If anyone knows of a freeware/public domain small
> program that does giffing I've a few friends that could use it.
> And also it would help the list, as jpg attachments of diagrams
> get broken up pretty badly and are larger than they need to be
>

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@084927 by Jinx

face picon face
> > > I should e-mail RS and find out exactly what they mean by Vce
> >
> > They won't know.
>
> I e-mailed the UK office anyway. When / if I get a reply I'll pass it on

=====================================================

                      Dear Sir,
                              In reply to your enquiry relating to the
transistor tables and quotes of Vce .The lines for hfe Ic/mA and Vce should
be read together . Such that the gain value hfe is quoted at a particular
value of collector current and collector to emitter voltage .
                                                   Regards ,
                                                   John Taylor.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@085343 by Alan B. Pearce

face picon face
> > I should e-mail RS and find out exactly what they mean by Vce
>
> They won't know.

From my reading of RS data pages, it appears to be the Vce at which the beta
is measured.

Looking at my current (no pun intended) RS CD, for a BC547 there is 3
columns labelled (separated by vertical stripes here) hfe@ | Ic/mA | Vce, so
my guess is that these three columns go together. There is seperate columns
for Vceo max, and IC max.

Perhaps the NZ distributors have missed most of the columns out :)))))))

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@130259 by Roman Black

flavicon
face
Jinx wrote:

> I appreciate what you're saying, which is why the RS tables are not
> thoroughly helpful. Zetex's pdf for eg the ZTX450 show Vceo = 45V
> and Vcesat = 0.36V.


I bought a heap of Zetex "high perf" low-sat
transistors a couple of weeks back, and was VERY
disappointed with the results. :o(

In the SMPS circuits I was often getting better
results with a 4 cent BC327 than with a $1 Zetex
part.

The Vce sat voltages of 0.3v and 0.4v as shown in
the Farnell and Zetex specs are a bit of a joke,
requiring betas as low as 10(!!) or 50 when I
actually got the Zetex datasheet after being royally
pissed off when testing the things.

The gain of 160 or 300 as shown in the catalogue
were at high Vce and very low currents. The low-sat
of 0.3v at 3A which looks so impressive was with
gain of 10 (300mA base current) which makes them
almost useless.

As a comparison I have BD203 etc here, 20 year old
TV transistors with better low sat and AT a better
beta and hgiher current. And cost 45 cents.

Now for a $1 "super transistor" I expect low sat
AND high gain, which are available, but not from
Zetex it seems. Grrrr. :o{
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@135321 by Jim

flavicon
face
  "requiring betas as low as 10(!!) or 50 "

Huh?

Do you mean to say that you were required to
provide additional "base drive current" in
order to achive the desired Vcesat (volts collector
to emitter at saturation) you were looking for?

I'm sorry, but I have to address this issue -
'beta' values for a transistor are measured at
*specified* collector currents and is a parameter
that varies *greatly* over the range collector
currents which a device (a transistor) is
designed to operate ... in the not-so-old-days
"characteristic curve" of Vce versus fixed
or stepped base drive currents were a part of
a devices data sheet - one could then *easily*
determine the 'beta' of a transistor at various
"operating points" for a particular type number
of transisitor.

My fear now is that the thinkiing here is that
ALL transistor parameters are constants for
all operating conditions! This is NOT true and
has NEVER been the case!

(Sorry ahead of time for the rant!)

RF Jim (who has swept his fair share of semiconductor
devices!)


{Original Message removed}

2002\10\14@152451 by Peter L. Peres

picon face
On Tue, 15 Oct 2002, Jinx wrote:

*>Some table samples
*>
*>BC549      Vceo 30    Vce 5
*>2SC3329 Vceo 80     Vce 0.1
*>TIP50        Vceo 400   Vce 10

Does it not go on like:

BC549   Vceo 30         Vce 5   Ic 5    Beta 120 ...

?

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@160611 by Jinx

face picon face
> Does it not go on like:
>
> BC549   Vceo 30         Vce 5   Ic 5    Beta 120 ...

A complete entry is

                Ic/A     Pd/W    Vceo      hfe@      Ic/mA  Vce    fT/MHz

ZTX450    1.00    1.00    45.00   100-300     150     10     150 min

This is RS (UK)'s definition of Vce

"The lines for hfe Ic/mA and Vce should be read together. Such that
the gain value hfe is quoted at a particular value of collector current
and collector to emitter voltage"

In the Zetex pdf for it, the Vcesat for the ZTX450 is shown as 0.25V
to 0.35V @ Ic 150mA, Ib 15mA

http://www.zetex.com/3.0/pdf/ZTX450.pdf   (50kB)

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@164414 by Russell McMahon

face
flavicon
face
> All academic, only whoever did the RS tables knows for sure

Sadly, I'm pretty sure that whoever did the RS table has no idea whatsoever.
Have a look randomly at other electronic device tables in the latest
catalogue and you will come up with all sorts of interesting information.
Something very very strange happened when they made that catalogue (even if
everyone else likes it).


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@164422 by Russell McMahon

face
flavicon
face
> > > > I should e-mail RS and find out exactly what they mean by Vce
> > > They won't know.
> > I e-mailed the UK office anyway. When / if I get a reply I'll pass it on

>                        Dear Sir,
>                                In reply to your enquiry relating to the
> transistor tables and quotes of Vce .The lines for hfe Ic/mA and Vce
should
> be read together . Such that the gain value hfe is quoted at a particular
> value of collector current and collector to emitter voltage .
>                                                     Regards ,
>                                                     John Taylor.

That would be almost convincing if your list of 3 examples had not included
the entry

       2SC3329 Vceo 80     Vce 0.1

Now it's POSSIBLE that they are intending to state the hfe with the
transistor saturated but I rather doubt it. I suspect that John Taylor has
made a valiant attempt to explain what he sees.

Another possible explanation, also without perfect correlation with reality,
is that this parameter is MEANT to be Vebo. I looked at the relevant table
in the RS catalogue and, with that theory in mind, chose a device with a
very high figure for Vce. The Zetex ZTX600B has a claimed Vce of 10 and a
look at http://www.zetex.com shows it also has a Vebo of 10 volts. Not conclusive
but suspicious. [[With a Beta of 1,000 minimum at 1A this E-line (TO92
sized) part should be kept in mind for impossible applications. Vsat at
0.85v at 1A is high as it's a Darlington]].

But aaaagh - it does have it's hfe specd at Vce=10v ! :-)

As one last try - the MJD127 has an RS Vce of 1 volt but at
onsemi.com.cn/pub/Collateral/DataSheet/mjd122rev1.pdf
they list no parameter at 1 volt. It does have a Vce_sat of 2v. (Warning -
this is a pdf AND if you read it they will probably have to kill you)*.


           Russell McMahon


* - read url

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@193745 by Jim

flavicon
face
Are you guys relying on a distributer's
'presentation of the specs' as opposed
to the manufacturer's data sheet for a
device?

RF Jim

{Original Message removed}

2002\10\14@200123 by Jinx

face picon face
> Are you guys relying on a distributer's
> 'presentation of the specs' as opposed
> to the manufacturer's data sheet for a
> device?

Not relying, no, and to be fair to RS they do show a lot
of convenient information. Not so much as they used to
because there just isn't room in the catalogue with the
vast number of products they stock these days

An example from yesterday though of further research needed

The HIP4080 is available in AIB or AFB SOIC20 packages, and
the AFB is $6 dearer than the AIB. What's the difference ? RS
don't say, but neither does the HIP4080 pdf. I've been too busy
to find out, except that the M20.3 specs for SOIC20 don't mention
AFB or AIB

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@210806 by Russell McMahon

face
flavicon
face
> Are you guys relying on a distributer's
> 'presentation of the specs' as opposed
> to the manufacturer's data sheet for a
> device?

Or, az opozd to vot dey reli behav lik (even ven dey meet der specz (ask
Roman)), or .....


       RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\14@212014 by Russell McMahon

face
flavicon
face
> > Are you guys relying on a distributer's
> > 'presentation of the specs' as opposed

> Not relying, no,

Not even Jinx lives that riskily

> and to be fair to RS they do show a lot
> of convenient information. Not so much as they used to
> because there just isn't room in the catalogue with the
> vast number of products they stock these days

A decent presentation of useful data in the room available will allow you to
shortlist devices which may well suit and then use the datasheets to confirm
your choices. RS still have enough room to allow you to do this but someone
went mad two issues ago and turned a number of their tables into custard and
they have only partially got it back under control this time round . They
either include strange columns of data and/or exclude useful/normal data
and/or quote figures to many more significant digits than needed or in
inconsistent ways or .... The same sort of data is represented in different
notation for adjacent devices making column skimming hard(er).

Slew rate V/uS:  0.01 for one device, next one is 1.0e-03.
Aol V/mV = 12000.00 (wow) (LTC1150CN8 = 1000000.00!)

Many devices do not have the columns fully populated (or even at all
populated in some cases).
eg NJM386L has no heatsink - no other data is included in the 8 available
columns.

24LC01 is a 1024.00 byte device.



               RM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\10\15@040257 by Alan B. Pearce

face picon face
>I'm sorry, but I have to address this issue -
>'beta' values for a transistor are measured at
>*specified* collector currents and is a parameter
>that varies *greatly* over the range collector
>currents which a device (a transistor) is

True, but often to achieve a transistor saturation a "forced beta" is used.
When you get to doing relay switching on a space craft then the forced beta
figure you use has to have a margin to allow for transistor aging due to
radiation effects and so on.

Forced beta is done by determining what the required collector current is,
and supplying the base with Ic/(forced beta figure). This is often the only
way of being able to guarantee a Vce(sat) figure at a given Ic, as roman has
found with the specs on his Zetex transistors.

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@052642 by Roman Black

flavicon
face
Jim wrote:
>
>    "requiring betas as low as 10(!!) or 50 "
>
> Huh?
> Do you mean to say that you were required to
> provide additional "base drive current" in
> order to achive the desired Vcesat (volts collector
> to emitter at saturation) you were looking for?
>
> I'm sorry, but I have to address this issue -
> 'beta' values for a transistor are measured at
> *specified* collector currents and is a parameter
> that varies *greatly* over the range


Exactly my point. The beta specified in the
catalogue as 160, quite reasonable for a low sat
transistor in its happy range.

What annoyed me was the (hidden) fact that YES
the transistor will do 0.3v sat at 3A, but only
at a very poor beta of 10.

Zetex (or Farnell) seem to have deliberately
chosen to show a beta of 160, at a totally
unreasonable current for a 3A transistor. When
operated as a low-sat transistor the beta is
VERY POOR.

Over 20 years experience of having to substitute
hard to get semis in repairing appliances I ALWAYS
test semis. I have many low-sat transistors here,
which WILL give betas of 50 to 100 with high currents
and low sats. A beta of 10 makes the low sat feature
worthless to me.

My gripe is that the Farnell catalogue shows the
Zetex parts as "high gain super low sat" which is
downright misleading when you find that they don't
do either thing *that* well. I don't consider these
transistors to be high gain, and if advertised as
low sat at 3A a more realistic beta should have been
shown, as based on their figures I expected the
Zetex parts to perform similar to the other brands
I had here, with betas around 100 at 1A and maybe 80
at 3A. The Zetex parts have betas of 150 at 200mA
and 10(!) at 3A, gripe gripe gripe.

But they are small.

I also sent back a batch of 1N5819 and 1N5822 schottky
diodes, when I tested them they were the ST brand and
had sat voltages 40% higher(!) than the ones I had in
stock, which were motorola/ON brand. I hate getting ripped
off with semis, and suppliers constantly source cheaper
parts with the same part number, meaning you end up
getting crappier and crappier performance from a part
number you expect to perform as advertised. Grrr.
-Roman

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@054542 by Vasile Surducan

flavicon
face
On Tue, 15 Oct 2002, Roman Black wrote:

> Zetex parts to perform similar to the other brands
> I had here, with betas around 100 at 1A and maybe 80
> at 3A. The Zetex parts have betas of 150 at 200mA
> and 10(!) at 3A, gripe gripe gripe.
>
 Roman, I really didn't understood why are you so surprized,
this is a common situation for power transistors. Take a bunch of
primitive 2N3055 and see the best beta ( or h21e if you prefer ) at
3...4A is 10 to 50 ( 50 just for some exceptions )
Beta is not a real datasheet value. But I'm agree, a serious producer will
mark on his datasheet the correct range of beta for the measured range of
colector current.
However a beta of 150 at 200mA is a good starting point to appreciate
the beta at 2 A ( which can be near 15,  more or less )

Vasile

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@055204 by Roman Black

flavicon
face
Alan B. Pearce wrote:

> True, but often to achieve a transistor saturation a "forced beta" is used.

> Forced beta is done by determining what the required collector current is,
> and supplying the base with Ic/(forced beta figure). This is often the only
> way of being able to guarantee a Vce(sat) figure at a given Ic, as roman has
> found with the specs on his Zetex transistors.

Yep, I use two constant current supplies with
good ammeters for Ic and Ib and a voltmeter for
the Vce sat. I don't have Jim's luxury of high power
curve tracer, although THAT sounds like a cool
application for a PIC. And I NEVER trust published
semi specs, especially wih Asian manufacturers.
-Roman

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@060622 by Roman Black

flavicon
face
Vasile Surducan wrote:

> > Zetex parts to perform similar to the other brands
> > I had here, with betas around 100 at 1A and maybe 80
> > at 3A. The Zetex parts have betas of 150 at 200mA
> > and 10(!) at 3A, gripe gripe gripe.
> >
>   Roman, I really didn't understood why are you so surprized,
> this is a common situation for power transistors. Take a bunch of
> primitive 2N3055 and see the best beta ( or h21e if you prefer ) at
> 3...4A is 10 to 50 ( 50 just for some exceptions )
> Beta is not a real datasheet value.

I knew the beta would be much lower with the
larger currents, but Zetex were much WORSE than
parts I already had! The Zetex were advertised
as "high performance super transistors" and
really did not deliver "super" performance
at all. Try a BD203/BD204 or even some of the
TIP low-sat series which work better and at half
the price. Sure they're bigger but maybe that is
better too, they DO have much greater safe
margins.
-Roman

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@062319 by Michael Rigby-Jones

flavicon
face
> -----Original Message-----
> From: Roman Black [SMTP:EraseMEfastvidspamEraseMEEZY.NET.AU]
> Sent: Tuesday, October 15, 2002 11:01 AM
> To:   @spam@PICLIST@spam@spamspam_OUTMITVMA.MIT.EDU
> Subject:      Re: [EE]: Protecting FETs
>
> I knew the beta would be much lower with the
> larger currents, but Zetex were much WORSE than
> parts I already had! The Zetex were advertised
> as "high performance super transistors" and
> really did not deliver "super" performance
> at all. Try a BD203/BD204 or even some of the
> TIP low-sat series which work better and at half
> the price. Sure they're bigger but maybe that is
> better too, they DO have much greater safe
> margins.
> -Roman
>
I happened to have the datasheet for the BD203, and took a look at the
specs.

Ic = 4A, Vce = 4volts : hfe min=20, hfe max = 100
Ic = 10A, Vce = 4volts : hfe min=5

Ic=4A, Ib=400mA : Vce(sat) max = 1.1volts
Ic=10A, Ib=3.3A : Vce(sat) max = 8volts

That dosen't seem particularly good, in fact the Zetex seems to be better?

Regards

Mike

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@124345 by Jim

flavicon
face
 "Forced beta"

We need to scratch the above term.

No one 'forces' a beta or hfe value. One may be said to
*overdrive* the base of a bipolar device ('device' - that's
what we used to call 'product' when generically referring to
whatever it was we were building/measuring/testing at the time) -

- via 'excess' base current (above that required to obtain
a particular collector current) in order to 'force' all
devices of a particular type number (and range of "betas"
for particular Ic (collector current)) -

- into as low a Vcesat condition as possible.

But no one 'forces' a calculated value (which the parm beta
happens to be: Ib/Ic) so to say you "force a beta value" makes
no sense ...

RF Jim




{Original Message removed}

2002\10\15@143540 by Peter L. Peres

picon face
On Tue, 15 Oct 2002, Jinx wrote:

*>> Does it not go on like:
*>>
*>> BC549   Vceo 30         Vce 5   Ic 5    Beta 120 ...
*>
*>A complete entry is
*>
*>                 Ic/A     Pd/W    Vceo      hfe@      Ic/mA  Vce    fT/MHz
                                             ^^^^^^^^^^^^^^^^^^^^

These three go together, as your rep said. Don't jump to conclusions about
catalogs too quick. These guys have been in business longer than you
probably ;-)

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@154532 by Bob Blick

face picon face
On Sun, 13 Oct 2002, Jinx wrote:
> > There are all sorts of useful app notes and data sheets on the Intersil
> > web site on the HIP408x family

I missed part of the thread, but if you want a good driver for making
mosfet h-bridges, look into the IR2112, they work great and are not
_too_ spendy. But you do need a supply voltage handy for the low side gate
drive.

Cheers,

Bob

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\10\15@174234 by Russell McMahon

face
flavicon
face
> *>                 Ic/A     Pd/W    Vceo      hfe@      Ic/mA  Vce
fT/MHz
>                                               ^^^^^^^^^^^^^^^^^^^^
>
> These three go together, as your rep said. Don't jump to conclusions about
> catalogs too quick. These guys have been in business longer than you
> probably ;-)

The catalogue mentioned is inconsistent and not always sensible, alas.
You can't rely on the above applying in every case.

The company has been in business longer than I have been alive, but I have
probably been in business longer than the people who work for them who
prepared the abominable version two issues back have been alive :-)

       RM

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...