Need high side logic-level mosfet

[Anthony Fremont]

I've been tinkering with some N-channel logic-level mosfets (NTE-2985).
These are really nice, but I have an application where I need to switch
the high side from a microcontroller. Since it's being retrofitted into
an existing automotive application, I don't have the luxury of just
wiring it up to switch on the low side.

Does anyone know of a complimentary part to the 2985 that doesn't cost
$20.00?

In case anyone is cares, here is the jist of the project. My motorcycle
has self-canceling turn signals triggered by two momentary switches.
The rear signals are single filament bulbs so there are no running
lights, just signals. The existing controller module has failed on one
side, consequently it won't pass inspection. Instead of paying $85.00
for a new module, I wish to build my own and make some modifications.
(repair is out of the question as it is all potted in epoxy) While I am
at it, I want to PWM the rear lights so that they act as running lights
without having to change to dual filament bulbs. I also will be
modifying the pattern of the signal flash so that they flash 3 times (at
approx 3Hz) and then pause for approx 1 second (and then repeat) for
maximum attention grabbing.

 

[James Meyer]

On Sat, 15 May 2004 13:22:44 GMT, "Anthony Fremont" <spam@anywhere.com> posted
this:

I've been tinkering with some N-channel logic-level mosfets (NTE-2985).
These are really nice, but I have an application where I need to switch
the high side from a microcontroller. Since it's being retrofitted into
an existing automotive application, I don't have the luxury of just
wiring it up to switch on the low side.

Does anyone know of a complimentary part to the 2985 that doesn't cost
$20.00?

Take a look at:

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=1085197&Row=264049&Site=US

It's the IPS521S "high-side switch" from International Rectifier. It's
$4.82 each in one's quantity. It was designed specifically for automotive
applications.

Jim

 

[Fred Bloggs]

Anthony Fremont wrote:

I've been tinkering with some N-channel logic-level mosfets (NTE-2985).
These are really nice, but I have an application where I need to switch
the high side from a microcontroller. Since it's being retrofitted into
an existing automotive application, I don't have the luxury of just
wiring it up to switch on the low side.

Does anyone know of a complimentary part to the 2985 that doesn't cost
$20.00?

In case anyone is cares, here is the jist of the project. My motorcycle
has self-canceling turn signals triggered by two momentary switches.
The rear signals are single filament bulbs so there are no running
lights, just signals. The existing controller module has failed on one
side, consequently it won't pass inspection. Instead of paying $85.00
for a new module, I wish to build my own and make some modifications.
(repair is out of the question as it is all potted in epoxy) While I am
at it, I want to PWM the rear lights so that they act as running lights
without having to change to dual filament bulbs. I also will be
modifying the pattern of the signal flash so that they flash 3 times (at
approx 3Hz) and then pause for approx 1 second (and then repeat) for
maximum attention grabbing.

View in a fixed-width font such as Courier.

..
..
.. VBATT (filtered)
.. |
.. 1K |
.. +--/\/\----+
.. | |
.. | ||--+
.. +------||>-+ NTE2371
.. | ||--+
.. | |
.. ON | |
.. +--- | +---> TO LIGHTS
.. OFF | |
.. ---+ |
.. ||--+
.. FROM UPROC>---+-------||<-+ 2N7000
.. | ||--+
.. | | (NTE491)
.. | 10K |
.. +-/\/\------+
.. |
.. ---
.. ///
..
..

 

[Anthony Fremont]

"James Meyer" <jmeyer@nowhere.com> wrote

"Anthony Fremont" <spam@anywhere.com> posted

Does anyone know of a complimentary part to the 2985 that doesn't
cost
$20.00?



http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=1085197&Row=264049&Site=US

It's the IPS521S "high-side switch" from International Rectifier.
It's
$4.82 each in one's quantity. It was designed specifically for
automotive
applications.

Thanks for the pointer, those are pretty slick. Doesn't look too bad
for my app, but I could possibly use a bit more continuous current
headroom. BTW, I really like the built in protection and feedback
features for determining a shorted/open output.

thanks again

 

[Jan Panteltje]

On a sunny day (Sat, 15 May 2004 15:07:53 GMT) it happened Fred Bloggs
<nospam@nospam.com> wrote in <40A63244.7010100@nospam.com>:

Anthony Fremont wrote:
I've been tinkering with some N-channel logic-level mosfets (NTE-2985).
These are really nice, but I have an application where I need to switch
the high side from a microcontroller. Since it's being retrofitted into
an existing automotive application, I don't have the luxury of just
wiring it up to switch on the low side.

Does anyone know of a complimentary part to the 2985 that doesn't cost
$20.00?

In case anyone is cares, here is the jist of the project. My motorcycle
has self-canceling turn signals triggered by two momentary switches.
The rear signals are single filament bulbs so there are no running
lights, just signals. The existing controller module has failed on one
side, consequently it won't pass inspection. Instead of paying $85.00
for a new module, I wish to build my own and make some modifications.
(repair is out of the question as it is all potted in epoxy) While I am
at it, I want to PWM the rear lights so that they act as running lights
without having to change to dual filament bulbs. I also will be
modifying the pattern of the signal flash so that they flash 3 times (at
approx 3Hz) and then pause for approx 1 second (and then repeat) for
maximum attention grabbing.


View in a fixed-width font such as Courier.

.
.
. VBATT (filtered)
. |
. 1K |
. +--/\/\----+
. | |
. | ||--+
. +------||>-+ NTE2371
. | ||--+
. | |
. ON | |
. +--- | +---> TO LIGHTS
. OFF | |
. ---+ |
. ||--+
. FROM UPROC>---+-------||<-+ 2N7000
. | ||--+
. | | (NTE491)
. | 10K |
. +-/\/\------+
. |
. ---
. ///
.
Given that Vbatt can be quite high, and you do not want Vgs breakdown,

would not Winfrieds circuit recently posted in this group,
with a current source as driver more safely limit Vgs in the top FET?


/c
-----| NPN
\e
|
| | R so ( (Vdrive - 0.7) / R) * 1k = smaller then Vgs max
|
gnd

?

 

[Graham W]

Anthony Fremont wrote:

I've been tinkering with some N-channel logic-level mosfets
(NTE-2985). These are really nice, but I have an application where I
need to switch the high side from a microcontroller.
[...]
I also will be modifying the pattern of the signal
flash so that they flash 3 times (at approx 3Hz) and then pause for
approx 1 second (and then repeat) for maximum attention grabbing.

In the UK at least, the flash pattern is controlled by legislation and
is subject to the vehicle examiner's testing schedule. Check with your
local DOT office for what is acceptable.


--
Graham W http://www.gcw.org.uk/ PGM-FI page updated, Graphics Tutorial
WIMBORNE http://www.wessex-astro-society.freeserve.co.uk/ Wessex
Dorset UK Astro Society's Web pages, Info, Meeting Dates, Sites & Maps
Change 'news' to 'sewn' in my Reply address to avoid my spam filter.

 

[Winfield Hill]

Jan Panteltje wrote...

would not Winfrieds circuit ...

Winfried? Ahem.

Thanks,
- Win

(email: use hill_at_rowland-dot-org for now)

 

[Anthony Fremont]

"Fred Bloggs" <nospam@nospam.com> wrote in message
news:40A63244.7010100@nospam.com...

Anthony Fremont wrote:
I've been tinkering with some N-channel logic-level mosfets
(NTE-2985).
These are really nice, but I have an application where I need to
switch
the high side from a microcontroller. Since it's being retrofitted
into
an existing automotive application, I don't have the luxury of just
wiring it up to switch on the low side.

Does anyone know of a complimentary part to the 2985 that doesn't
cost
$20.00?

View in a fixed-width font such as Courier.

.
.
. VBATT (filtered)
. |
. 1K |
. +--/\/\----+
. | |
. | ||--+
. +------||>-+ NTE2371
. | ||--+
. | |
. ON | |
. +--- | +---> TO LIGHTS
. OFF | |
. ---+ |
. ||--+
. FROM UPROC>---+-------||<-+ 2N7000
. | ||--+
. | | (NTE491)
. | 10K |
. +-/\/\------+
. |
. ---
. ///

Thanks Fred, I'm sure that would work fine, but I'm really intrigued by
these "Intelligent Power Switches" and the added benefit of being able
to detect a short or open circuit. Lack of proper protection is why I'm
having to re-engineer the factory module. :-( I'll have to seal
everything in epoxy due to vibration and moisture, so I'd like to use
some parts that can't be easily killed.

 

[Anthony Fremont]

"Graham W" wrote:

Anthony Fremont wrote:

I also will be modifying the pattern of the signal
flash so that they flash 3 times (at approx 3Hz) and then pause for
approx 1 second (and then repeat) for maximum attention grabbing.

In the UK at least, the flash pattern is controlled by legislation and
is subject to the vehicle examiner's testing schedule. Check with your
local DOT office for what is acceptable.

Every thing that I can find (USDOT/TXDOT) indicates that this is
acceptable. State law here only says that it has to flash and that's
about it. USDOT regs seem to indicate an ideal rate of 90
flashes/minute with 60-120 being acceptable. No wording about duty
cycle or flash pattern is indicated.

I'd like to find a copy SAE J590B if anyone knows where to find it (for
free of course ;-)

 

[Anthony Fremont]

"Graham W" wrote:

Anthony Fremont wrote:

I also will be modifying the pattern of the signal
flash so that they flash 3 times (at approx 3Hz) and then pause for
approx 1 second (and then repeat) for maximum attention grabbing.

In the UK at least, the flash pattern is controlled by legislation and
is subject to the vehicle examiner's testing schedule. Check with your
local DOT office for what is acceptable.

Every thing that I can find (USDOT/TXDOT) indicates that this is
acceptable. State law here only says that it has to flash and that's
about it. USDOT regs seem to indicate an ideal rate of 90
flashes/minute with 60-120 being acceptable. No wording about duty
cycle or flash pattern is indicated.

I'd like to find a copy SAE J590B if anyone knows where to find it (for
free of course ;-)

 

[Jan Panteltje]

On a sunny day (15 May 2004 09:27:57 -0700) it happened Winfield Hill
<Winfield_member@newsguy.com> wrote in <c85ged081d@drn.newsguy.com>:

Jan Panteltje wrote...

would not Winfrieds circuit ...

Winfried? Ahem.

Thanks,
- Win

(email: use hill_at_rowland-dot-org for now)
Sorry about that, I will try to remember now

JP

 

[Rich Grise]

"Anthony Fremont" <spam@anywhere.com> wrote in message
news:lGspc.102292$NR5.98925@fe1.texas.rr.com...

"Graham W" wrote:
Anthony Fremont wrote:
I also will be modifying the pattern of the signal
flash so that they flash 3 times (at approx 3Hz) and then pause for
approx 1 second (and then repeat) for maximum attention grabbing.
In the UK at least, the flash pattern is controlled by legislation and
is subject to the vehicle examiner's testing schedule. Check with your
local DOT office for what is acceptable.
Every thing that I can find (USDOT/TXDOT) indicates that this is
acceptable. State law here only says that it has to flash and that's
about it. USDOT regs seem to indicate an ideal rate of 90
flashes/minute with 60-120 being acceptable. No wording about duty
cycle or flash pattern is indicated.

Remember those 3-lamp units that came on in sequence? I wonder what ever
happened to them.

Cheers!
Rich

 

[Anthony Fremont]

"Rich Grise" <null@example.net> wrote

Remember those 3-lamp units that came on in sequence? I wonder what
ever
happened to them.

Penny pinching PHB's probably eliminated them due to the "outrageous"
extra expense. ;-)

On topic:
I was playing with my logic level mosfets (n-channel) driving some bulbs
{I don't need the high side switches for bench testing} and I ran into
an interesting problem while working on some code. I had some stuff
working using an automotive electrical tester as my "bulb". When I dug
out a real 1156 bulb (like I'll be using in real life) and hooked it up
for a real smoke test, all kinds of interesting things started happening
with my PIC. It was spontaneously rebooting and parts of the code
weren't working any more. I put a resistor between the PIC and the gate
of the mosfet and lo it all started working again.

Q: Why did I need to do that? I'm surprised that the high impedance
gate could kick back at the PIC that hard. I also find it quite
intriguing that the bulb acted so inductively(?) as if it was causing
large voltage spikes that shot right thru the gate of the transistor.
Admittedly I was trying to hold little test leads to the bulb since I
didn't have a proper socket for it, so they were connecting and
disconnecting intermittently causing the drain to float. Could it have
been due to the inductance of the test leads, or is it that in
combination with the drain floating intermittently?

 

[Anthony Fremont]

"Rich Grise" <null@example.net> wrote:

Sorry if this is all wrapped wrong, I had to repost due to my ISP's
incompetent news server randomly tossing messages away.

Remember those 3-lamp units that came on in sequence? I wonder what
ever
happened to them.

Penny pinching PHB's probably eliminated them due to the "outrageous"
extra expense. ;-)

On topic:
I was playing with my logic level mosfets (n-channel) driving some bulbs
{I don't need the high side switches for bench testing} and I ran into
an interesting problem while working on some code. I had some stuff
working using an automotive electrical tester as my "bulb". When I dug
out a real 1156 bulb (like I'll be using in real life) and hooked it up
for a real smoke test, all kinds of interesting things started happening
with my PIC. It was spontaneously rebooting and parts of the code
weren't working any more. I put a resistor between the PIC and the gate
of the mosfet and lo it all started working again.

Q: Why did I need to do that? I'm surprised that the high impedance
gate could kick back at the PIC that hard. I also find it quite
intriguing that the bulb acted so inductively(?) as if it was causing
large voltage spikes that shot right thru the gate of the transistor.
Admittedly I was trying to hold little test leads to the bulb since I
didn't have a proper socket for it, so they were connecting and
disconnecting intermittently causing the drain to float. Could it have
been due to the inductance of the test leads, or is it that in
combination with the drain floating intermittently?

 

[Hal Murray]

On topic:
I was playing with my logic level mosfets (n-channel) driving some bulbs
{I don't need the high side switches for bench testing} and I ran into
an interesting problem while working on some code. I had some stuff
working using an automotive electrical tester as my "bulb". When I dug
out a real 1156 bulb (like I'll be using in real life) and hooked it up
for a real smoke test, all kinds of interesting things started happening
with my PIC. It was spontaneously rebooting and parts of the code
weren't working any more. I put a resistor between the PIC and the gate
of the mosfet and lo it all started working again.

Q: Why did I need to do that? I'm surprised that the high impedance
gate could kick back at the PIC that hard. I also find it quite
intriguing that the bulb acted so inductively(?) as if it was causing
large voltage spikes that shot right thru the gate of the transistor.
Admittedly I was trying to hold little test leads to the bulb since I
didn't have a proper socket for it, so they were connecting and
disconnecting intermittently causing the drain to float. Could it have
been due to the inductance of the test leads, or is it that in
combination with the drain floating intermittently?

Incandescant lamps take roughly 10 times the normal current when
they are cold. My guess on the rebooting problem is that your
power supply was glitching when the lamp turned on.

The gate is not really high impedance. It's high resistance, but
the capicatance can be "high" too. Adding a resistor might slow
the turn on enough (simple R-C time constant) so that the lamp
turns on slowly and never draws a huge current spike.

Might be fun to investigate with a scope.

--
The suespammers.org mail server is located in California. So are all my
other mailboxes. Please do not send unsolicited bulk e-mail or unsolicited
commercial e-mail to my suespammers.org address or any of my other addresses.
These are my opinions, not necessarily my employer's. I hate spam.

 

[Rich Grise]

"Anthony Fremont" <spam@anywhere.com> wrote in message
newsoBpc.102398$NR5.95009@fe1.texas.rr.com...

On topic:
I was playing with my logic level mosfets (n-channel) driving some bulbs
{I don't need the high side switches for bench testing} and I ran into
an interesting problem while working on some code. I had some stuff
working using an automotive electrical tester as my "bulb". When I dug
out a real 1156 bulb (like I'll be using in real life) and hooked it up
for a real smoke test, all kinds of interesting things started happening
with my PIC. It was spontaneously rebooting and parts of the code
weren't working any more. I put a resistor between the PIC and the gate
of the mosfet and lo it all started working again.

Q: Why did I need to do that? I'm surprised that the high impedance
gate could kick back at the PIC that hard. I also find it quite
intriguing that the bulb acted so inductively(?) as if it was causing
large voltage spikes that shot right thru the gate of the transistor.
Admittedly I was trying to hold little test leads to the bulb since I
didn't have a proper socket for it, so they were connecting and
disconnecting intermittently causing the drain to float. Could it have
been due to the inductance of the test leads, or is it that in
combination with the drain floating intermittently?

I agree with what Hal Murray said, but I'd also add that I've seen
applications where they do something like shunt the FET so that
there's enough current flowing through the filament to keep it hot,
but not incandescent. I can see where that would be a pretty good
idea, if you don't mind the constant current drain. (then again,
I guess "running lights" accomplish pretty much the same thing. )

Cheers!
Rich

 

[Jan Panteltje]

On a sunny day (Sun, 16 May 2004 03:42:26 GMT) it happened "Anthony Fremont"
<spam@anywhere.com> wrote in <CqBpc.102399$NR5.3725@fe1.texas.rr.com>:

"Rich Grise" <null@example.net> wrote:

Sorry if this is all wrapped wrong, I had to repost due to my ISP's
incompetent news server randomly tossing messages away.

Remember those 3-lamp units that came on in sequence? I wonder what
ever
happened to them.

Penny pinching PHB's probably eliminated them due to the "outrageous"
extra expense. ;-)

On topic:
I was playing with my logic level mosfets (n-channel) driving some bulbs
{I don't need the high side switches for bench testing} and I ran into
an interesting problem while working on some code. I had some stuff
working using an automotive electrical tester as my "bulb". When I dug
out a real 1156 bulb (like I'll be using in real life) and hooked it up
for a real smoke test, all kinds of interesting things started happening
with my PIC. It was spontaneously rebooting and parts of the code
weren't working any more. I put a resistor between the PIC and the gate
of the mosfet and lo it all started working again.

Q: Why did I need to do that? I'm surprised that the high impedance
gate could kick back at the PIC that hard. I also find it quite
intriguing that the bulb acted so inductively(?) as if it was causing
large voltage spikes that shot right thru the gate of the transistor.
Admittedly I was trying to hold little test leads to the bulb since I
didn't have a proper socket for it, so they were connecting and
disconnecting intermittently causing the drain to float. Could it have
been due to the inductance of the test leads, or is it that in
combination with the drain floating intermittently?
I dunno, but lightbulbs have a R cold about 1/10 of wha tyou may expect,

so the peak current can be 10x that based on the wattage.
That may pull you supply down... If the PIC is on the same supply.
Resistor in series gate is goed idea anyways, but I have seen HV MOSFET
oscillate at many MHz with just a 40 W bulb...
What cured that was a 1nF across drain - source! (at 230V DC).
This oscillation can kill the MOSFET really fast if any real current flows.
JP

 

[Anthony Fremont]

"Jan Panteltje" <pNaonStpealmtje@yahoo.com> wrote

I dunno, but lightbulbs have a R cold about 1/10 of wha tyou may
expect,
so the peak current can be 10x that based on the wattage.
That may pull you supply down... If the PIC is on the same supply.

This was my initial thought, but the bulb has a separate 12V power
supply from the PIC's 5V supply. Both supplies have the ability to
deliver more than 20A. Since I'm playing on solderless breadboards,
they share a common ground path. At any rate, it was the resistor that
made the difference so I really feel that it has something to do with
the gate to PIC connection.

Resistor in series gate is goed idea anyways, but I have seen HV
MOSFET
oscillate at many MHz with just a 40 W bulb...
What cured that was a 1nF across drain - source! (at 230V DC).
This oscillation can kill the MOSFET really fast if any real current
flows.

What would cause the oscillation?

 

[Anthony Fremont]

"Jan Panteltje" <pNaonStpealmtje@yahoo.com> wrote

I dunno, but lightbulbs have a R cold about 1/10 of wha tyou may
expect,
so the peak current can be 10x that based on the wattage.
That may pull you supply down... If the PIC is on the same supply.

This was my initial thought, but the bulb has a separate 12V power
supply from the PIC's 5V supply. Both supplies have the ability to
deliver more than 20A. Since I'm playing on solderless breadboards,
they share a common ground path. At any rate, it was the resistor that
made the difference so I really feel that it has something to do with
the gate to PIC connection.

Resistor in series gate is goed idea anyways, but I have seen HV
MOSFET
oscillate at many MHz with just a 40 W bulb...
What cured that was a 1nF across drain - source! (at 230V DC).
This oscillation can kill the MOSFET really fast if any real current
flows.

What would cause the oscillation?

 

[Spehro Pefhany]

On Sun, 16 May 2004 14:46:46 GMT, the renowned "Anthony Fremont"
<spam@anywhere.com> wrote:

"Jan Panteltje" <pNaonStpealmtje@yahoo.com> wrote

I dunno, but lightbulbs have a R cold about 1/10 of wha tyou may
expect,
so the peak current can be 10x that based on the wattage.
That may pull you supply down... If the PIC is on the same supply.

This was my initial thought, but the bulb has a separate 12V power
supply from the PIC's 5V supply. Both supplies have the ability to
deliver more than 20A. Since I'm playing on solderless breadboards,
they share a common ground path. At any rate, it was the resistor that
made the difference so I really feel that it has something to do with
the gate to PIC connection.

Resistor in series gate is goed idea anyways, but I have seen HV
MOSFET
oscillate at many MHz with just a 40 W bulb...
What cured that was a 1nF across drain - source! (at 230V DC).
This oscillation can kill the MOSFET really fast if any real current
flows.

What would cause the oscillation?

Inductance in the source.

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com