Maker Pro
Maker Pro

Need help driving two IPS1031 Low-Side MOSFETs from a PIC

The IPS1031 is a Low-Side MOSFET Driver that drives its own MOSFET. I
am planning on incorporating it into my design, but most of its spec
sheet is total gibberish to me.
I am driving two of them off of a single PIC output which can sink
3.7-5V @ 25mA.
The IPS1031 datasheet:
http://www.irf.com/product-info/datasheets/data/ips1031.pdf

Attached is a schem I used for its simulations purposes. Through the
use of "OR" diodes, only two IPS1031s are on at any single time; Q2 is
saturated when either Q1 or Q3 is saturated. S1 and S2 signafy two
different PIC Outputs.
Schematic Image located At:
http://www.myfilehut.com/userfiles/155002/Switch3.JPG
 
W

Winfield Hill

Jan 1, 1970
0
[email protected] wrote...
The IPS1031 is a Low-Side MOSFET Driver that drives its own MOSFET.
I am planning on incorporating it into my design, but most of its
spec sheet is total gibberish to me. I am driving two of them off
of a single PIC output which can sink 3.7-5V @ 25mA.
http://www.irf.com/product-info/datasheets/data/ips1031.pdf

Attached is a schem I used for its simulations purposes. Through
the use of "OR" diodes, only two IPS1031s are on at any single
time; Q2 is saturated when either Q1 or Q3 is saturated. S1 and
S2 signafy two different PIC Outputs.
http://www.myfilehut.com/userfiles/155002/Switch3.JPG

You are driving these MOSFET's with 5V logic signals, right?

OK. What about the spec sheet is gibberish to you? This part
looks like an ordinary MOSFET with a well-controlled avalanche
voltage of 34V or more, and a shutoff-protection feature that
operates if Id > 18A or Tj > 165 degrees, typical. The FET's
protection circuitry considerably slows down the its response
time, but it makes it virtually indestructible.

What's your question? Why did you choose this part?
 
S

Spehro Pefhany

Jan 1, 1970
0
The IPS1031 is a Low-Side MOSFET Driver that drives its own MOSFET. I
am planning on incorporating it into my design, but most of its spec
sheet is total gibberish to me.
I am driving two of them off of a single PIC output which can sink
3.7-5V @ 25mA.
The IPS1031 datasheet:
http://www.irf.com/product-info/datasheets/data/ips1031.pdf

Attached is a schem I used for its simulations purposes. Through the
use of "OR" diodes, only two IPS1031s are on at any single time; Q2 is
saturated when either Q1 or Q3 is saturated. S1 and S2 signafy two
different PIC Outputs.
Schematic Image located At:
http://www.myfilehut.com/userfiles/155002/Switch3.JPG


Q2 isn't guaranteed to see the 4.5V minimum datasheet "on" voltage
with a 5V nominal supply on the PIC.

As Win asked, why did you choose the IPS1031 rather than a more
jellybean part?


Best regards,
Spehro Pefhany
 
Spehro said:
Q2 isn't guaranteed to see the 4.5V minimum datasheet "on" voltage
with a 5V nominal supply on the PIC.

As Win asked, why did you choose the IPS1031 rather than a more
jellybean part?


Best regards,
Spehro Pefhany

Spehro, I was looking for the Voltage required to turn on. The spec
sheet said less than 4.5V could be used, but it didnt specify beyond
that, or did it? Thats what confused me. The reason I picked this, over
a a more "normal" device such as a 2N2222, was because I was told that,
"a dozen of these" could be controlled by a PIC output. As you were, I
was also concerned that in combination of the voltage drop from the
PIC, as it does when increase mA sourced, and the voltage drop from the
diode, the MOSFET would not see the required voltages
Perhaps you could suggest another MOSFET or transistor?
 
L

Luhan

Jan 1, 1970
0
The IPS1031 is a Low-Side MOSFET Driver that drives its own MOSFET. I
am planning on incorporating it into my design, but most of its spec
sheet is total gibberish to me.
I am driving two of them off of a single PIC output which can sink
3.7-5V @ 25mA.
The IPS1031 datasheet:
http://www.irf.com/product-info/datasheets/data/ips1031.pdf

Attached is a schem I used for its simulations purposes. Through the
use of "OR" diodes, only two IPS1031s are on at any single time; Q2 is
saturated when either Q1 or Q3 is saturated. S1 and S2 signafy two
different PIC Outputs.
Schematic Image located At:
http://www.myfilehut.com/userfiles/155002/Switch3.JPG

Use 3 outputs from the PIC. Eliminate the diodes, do the OR function in
the PIC.

Run the PIC on 6 volts to make sure you can turn on the FETS. This may
violate the PIC specs (just a bit), but I've run them this way with no
problems.

Luhan
 
L

Luhan

Jan 1, 1970
0
Luhan said:
Use 3 outputs from the PIC. Eliminate the diodes, do the OR function in
the PIC.

Run the PIC on 6 volts to make sure you can turn on the FETS. This may
violate the PIC specs (just a bit), but I've run them this way with no
problems.

Luhan

Actually, most PICs are specified at 6.5 to 7.5 volts. Just don't try
to program them at that voltage.

Luhan
 
Spehro said:
The VIH parameter is given as 4.5V. It says to refer to the
application note for operation outside of those conditions. When you
look at the application note, they say it's not recommended:
http://www.irf.com/technical-info/appnotes/an-1083.pdf

But actually it will work okay under normal conditions since you're
hardly drawing any current- 2.4mA on that output.



For Q2 and Q3 you can use a couple of 2N4401's. You can use 4.7K
series base resistors on Q2 (two pieces, one to each of the two PIC
outputs) and Q3. You don't need the diodes or the 100 ohm resistors or
the resistors to ground. I don't know what the relay coil current is,
so I can't tell you exactly what would be good for that output. For a
small relay (15-20mA) another 2N4401 with a 4.7K base resistor will
work fine. For up to 50 or 60mA, just reduce the base resistor to 1K.

Your BOM (for the drivers alone) in 100 quantity will drop from about
$5 to less than $0.50 and the parts count will drop from 11 to 7. They
are also multiple-sourced, small and easily available. OTOH, you lose
the short-circuit protection.


Best regards,
Spehro Pefhany

Thanks for the follow-up Sphero. However, those 2 diodes are needed.
The PIC is pre-progged. I don't know the exact specs of the 2N4401, I
will go look them up. However you believe that a single PIC output
could run two of them?
My entire design started out with 2N2222s, guess my original design
would have most likely worked.
 
S

Spehro Pefhany

Jan 1, 1970
0
Spehro, I was looking for the Voltage required to turn on. The spec
sheet said less than 4.5V could be used, but it didnt specify beyond
that, or did it?

The VIH parameter is given as 4.5V. It says to refer to the
application note for operation outside of those conditions. When you
look at the application note, they say it's not recommended:
http://www.irf.com/technical-info/appnotes/an-1083.pdf

But actually it will work okay under normal conditions since you're
hardly drawing any current- 2.4mA on that output.

Thats what confused me. The reason I picked this, over
a a more "normal" device such as a 2N2222, was because I was told that,
"a dozen of these" could be controlled by a PIC output. As you were, I
was also concerned that in combination of the voltage drop from the
PIC, as it does when increase mA sourced, and the voltage drop from the
diode, the MOSFET would not see the required voltages
Perhaps you could suggest another MOSFET or transistor?

For Q2 and Q3 you can use a couple of 2N4401's. You can use 4.7K
series base resistors on Q2 (two pieces, one to each of the two PIC
outputs) and Q3. You don't need the diodes or the 100 ohm resistors or
the resistors to ground. I don't know what the relay coil current is,
so I can't tell you exactly what would be good for that output. For a
small relay (15-20mA) another 2N4401 with a 4.7K base resistor will
work fine. For up to 50 or 60mA, just reduce the base resistor to 1K.

Your BOM (for the drivers alone) in 100 quantity will drop from about
$5 to less than $0.50 and the parts count will drop from 11 to 7. They
are also multiple-sourced, small and easily available. OTOH, you lose
the short-circuit protection.


Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
Thanks for the follow-up Sphero. However, those 2 diodes are needed.
The PIC is pre-progged.


Like this, no diodes: 4K7
___
S1 o-|___|-+
|
|
___ | to 5K load
S2 o-|___|-+
| |
4K7 | |/
+-----| 2N4401
|>
|
|
===
GND
I don't know the exact specs of the 2N4401, I
will go look them up. However you believe that a single PIC output
could run two of them?

Yes. Even with an extremely conservative design you only need 1/20 of
the collector current at the base. So you could drive lots of 5K
(2.4mA) loads since each one only would need 120uA source (not sink)
current from the PIC output. Fewer 50mA loads.
My entire design started out with 2N2222s, guess my original design
would have most likely worked.

The guaranteed gain is better on the 2N4401, IIRC, but probably it
would have worked.


Best regards,
Spehro Pefhany
 
Spehro said:
Like this, no diodes: 4K7
___
S1 o-|___|-+
|
|
___ | to 5K load
S2 o-|___|-+
| |
4K7 | |/
+-----| 2N4401
|>
|
|
===
GND


Yes. Even with an extremely conservative design you only need 1/20 of
the collector current at the base. So you could drive lots of 5K
(2.4mA) loads since each one only would need 120uA source (not sink)
current from the PIC output. Fewer 50mA loads.


The guaranteed gain is better on the 2N4401, IIRC, but probably it
would have worked.


Best regards,
Spehro Pefhany

Thanks for the quick reply again Spehro. Your diagram isnt readable,
sorry. How do you use the transistors, Hfe, in order to find the needed
base amperage? Sorry I kinda feel out of the loop on the transistory
termanology.
 
Thanks for the quick reply again Spehro. Your diagram isnt readable,
sorry. How do you use the transistors, Hfe, in order to find the needed
base amperage? Sorry I kinda feel out of the loop on the transistory
termanology.
Ok I "decypherd" your schematic LOL. But in your schem your mixing the
signals, If i were to do this, the relay would be on whenever a signal
was enterd. The 2 diodes allow for the relay to go on when only one
specific signal is enterd.
 
S

Spehro Pefhany

Jan 1, 1970
0
Thanks for the quick reply again Spehro. Your diagram isnt readable,
sorry.

Change your viewing font to a fixed-width font such as Courier and it
will be readable. You'll also find many other posts that read better
that way, so it's recommended.
How do you use the transistors, Hfe, in order to find the needed
base amperage?

You don't use it directly, as you want the transistor to be really
well saturated. If you have the full data sheet, look at the minimum
hFE over temperature, the curves of typical Vce(sat) vs. base current
and collector current etc. Usually Vce(sat) is specified at Ic/Ib =
10, but larger numbers are quite practical in most cases.

http://www.fairchildsemi.com/ds/2N/2N4401.pdf

The base resistors can be calculated using Ohm's law - supply voltage,
minus output drop of the PIC when sourcing the desired current minus
base-emitter voltage of the transistor, divided by the desired base
current. With Q2 you have 10-15% of current bled off by the resistor
to the 2nd output, when one is on and the other is off.



Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
Ok I "decypherd" your schematic LOL. But in your schem your mixing the
signals, If i were to do this, the relay would be on whenever a signal
was enterd. The 2 diodes allow for the relay to go on when only one
specific signal is enterd.

The PIC outputs actively drive low as well as high*, and the 130uA or
so that the "low" relay drive output sinks when it is off, and the
other output is on, is not high enough to cause problems. The base
resistors could be increased from 4.7K to 10K, 20K or 30K without
problems.

If it really bothers you, it would be better to use two 2N4401s in
parallel to drive the 5K load than to add diodes (one less part etc.),
but it's not necessary.

*Unless deliberately tristated, but that's an unlikely situation.


Best regards,
Spehro Pefhany
 
Spehro said:
The PIC outputs actively drive low as well as high*, and the 130uA or
so that the "low" relay drive output sinks when it is off, and the
other output is on, is not high enough to cause problems. The base
resistors could be increased from 4.7K to 10K, 20K or 30K without
problems.

If it really bothers you, it would be better to use two 2N4401s in
parallel to drive the 5K load than to add diodes (one less part etc.),
but it's not necessary.

*Unless deliberately tristated, but that's an unlikely situation.


Best regards,
Spehro Pefhany

The pic is not tristated. The pull-down resistors were only for use by
the simulator. I don't know if I will even use them on the actual
circuit. Could you clerify your use of two extra 2N4401s in place of
the two diodes??
Thanks, Jed
 
S

Spehro Pefhany

Jan 1, 1970
0
The pic is not tristated. The pull-down resistors were only for use by
the simulator. I don't know if I will even use them on the actual
circuit. Could you clerify your use of two extra 2N4401s in place of
the two diodes??
Thanks, Jed

One unnecessary 2N4401:

5K Load to 12VDC
o
|
|
+--+--+
30K | | 30K
___ |/ \| ___
o-|___|---| |---|___|-o
S1 |> <| S2
| |
+--+--+
|
|
===
GND



Best regards,
Spehro Pefhany
 
F

Fred Bloggs

Jan 1, 1970
0
The IPS1031 is a Low-Side MOSFET Driver that drives its own MOSFET. I
am planning on incorporating it into my design, but most of its spec
sheet is total gibberish to me.
I am driving two of them off of a single PIC output which can sink
3.7-5V @ 25mA.
The IPS1031 datasheet:
http://www.irf.com/product-info/datasheets/data/ips1031.pdf

Attached is a schem I used for its simulations purposes. Through the
use of "OR" diodes, only two IPS1031s are on at any single time; Q2 is
saturated when either Q1 or Q3 is saturated. S1 and S2 signafy two
different PIC Outputs.
Schematic Image located At:
http://www.myfilehut.com/userfiles/155002/Switch3.JPG

Re-arrange it so the active low outputs at the IPS1031's are ORed to
activate the IPS1031 driving the 5K load:
View in a fixed-width font such as Courier.

..
.. 12V
.. .------------+------------+-------------.
.. | | | |
.. - | [150] |
.. |\| | | |
.. RLY4 |\| | | [5K LOAD]
.. + |\| .-+-. | |
.. D1 - | | | |
.. | |< >| | |
.. +--[47K]-| 2x3906|-[39K]--+ |
.. | |\ /| | |
.. | | | - |
.. | '-+-' V D2 |
.. | | - |
.. | | | |
.. |D | D| |
.. +----+----+ | +----+----+ |
.. G| | | | |G |
.. S2>---| ips1031 | | | ips1031 |---< S1 |
.. | | | | | |
.. +---------+ | +---------+ |
.. | |
.. | |
.. | +---------+ |
.. | G | |D |
.. '-------| ips1031 |--------'
.. | |
.. +---------+
..
..
 
F

Fred Bloggs

Jan 1, 1970
0
Just noticed that thing wants logic level inputs:
View in a fixed-width font such as Courier.



12V
.-----------------+-------+-------------.
.. | | |
.. | | |
.. | | |
.. | 5V | |
.. - | [150] |
.. |\| .----+ | |
.. RLY4 |\| | | - [5K LOAD]
.. + |\| [10K]|< V D2 |
.. D1 - +--| - |
.. | | |\ | |
.. | [4.7K] | | |
.. | | | | |
.. | | | | |
.. +--|<|--+----|----|>|-----+ |
.. | | | |
.. | | | |
.. |D | D| |
.. +----+----+ | +----+----+ |
.. G| | | | |G |
.. S2>---| ips1031 | | | ips1031 |---< S1 |
.. | | | | | |
.. +---------+ | +---------+ |
.. | |
.. | |
.. | +---------+ |
.. | G | |D |
.. +-------| ips1031 |--------'
.. | | |
.. | +---------+
.. [4.7K]
.. |
.. |
.. ---
.. gnd
..
..
 
Top