Searching \ for '[EE] Looking for low RDSon N-ch Mosfet' 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=looking+low+rdson
Search entire site for: 'Looking for low RDSon N-ch Mosfet'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Looking for low RDSon N-ch Mosfet'
2010\07\05@151716 by Dwayne Reid

flavicon
face
Good day to all.

I'm looking at a buddy's project where his requirements have changed slightly.

He's currently using a IRF1405 N-channel Mosfet with a data-sheet
rating of RDSon = 5.3 Milli-Ohms.

He's finding that the temperature rise is higher than he likes and is
therefore looking for a Mosfet with lower RDSon.

The circuit operates at less than 30 Vdc and he has at least 10V of
gate drive available.

He'd prefer to stay with the current TO-220 package if possible.

Any suggestions?

Recap: N-channel Mosfet, TO-220 package, Vds 30V max, Id 30A max,
RDSon < 3 Milli-Ohms, > 10V gate drive available.

Many thanks!

dwayne

--
Dwayne Reid   <spam_OUTdwaynerTakeThisOuTspamplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2010\07\05@153941 by mikin

flavicon
face
IRF1404 RdsON= 4 mili
or
CSD17312Q5 RdsON= 1.4 mili from TI no need TO220 for 30Amps!
I worked with such low miliOhms Rds and the wires have more mili
than the RdsON so in case of something works bad the tracks/wires
will burn, not the MOS :))) If the Mos will be burned there is other
problems in the design (times constants, bad caps, bad Vgs, etc).


I hope will help you

Cheers
Miki


{Quote hidden}

> -

2010\07\05@154158 by Ing. Marcelo Fornaso

picon face
In continuous regime now you have 30 Amp. X 0.0053 Ohm = 159 mW.

It doesn't seem to be too much to heat the mosfet.

But maybe you are switching the mosfet. You didn't tell anything about switching frequency.
In that case the scenario is different.

Cheers,

Marcelo Fornaso






________________________________
From: Dwayne Reid <dwaynerspamKILLspamplanet.eon.net>
To: pic microcontroller discussion list <.....piclistKILLspamspam.....mit.edu>
Sent: Mon, July 5, 2010 4:17:15 PM
Subject: [EE] Looking for low RDSon N-ch Mosfet

Good day to all.

I'm looking at a buddy's project where his requirements have changed slightly.

He's currently using a IRF1405 N-channel Mosfet with a data-sheet
rating of RDSon = 5.3 Milli-Ohms.

He's finding that the temperature rise is higher than he likes and is
therefore looking for a Mosfet with lower RDSon.

The circuit operates at less than 30 Vdc and he has at least 10V of
gate drive available.

He'd prefer to stay with the current TO-220 package if possible.

Any suggestions?

Recap: N-channel Mosfet, TO-220 package, Vds 30V max, Id 30A max,
RDSon < 3 Milli-Ohms, > 10V gate drive available.

Many thanks!

dwayne

--
Dwayne Reid   <EraseMEdwaynerspam_OUTspamTakeThisOuTplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2010\07\05@155854 by Alan B Pearce

face picon face
> In continuous regime now you have 30 Amp. X 0.0053 Ohm = 159 mW.
>
> It doesn't seem to be too much to heat the mosfet.
>
> But maybe you are switching the mosfet. You didn't tell anything about
switching
> frequency.
> In that case the scenario is different.

That was my initial reaction too, switching losses rather than IIR
losses. What is he driving the gate with? At any reasonable switching
frequency he is probably going to need a MOSFET driver chip to get a
couple of amps gate drive, which you definitely won't get from an MCU
pin.
--
Scanned by iCritical.

2010\07\05@160536 by jim

flavicon
face
Run 3-4 in parallel

-----Original Message-----
From: piclist-bouncesspamspam_OUTmit.edu [@spam@piclist-bouncesKILLspamspammit.edu] On Behalf Of
Dwayne Reid
Sent: Monday, July 05, 2010 2:17 PM
To: pic microcontroller discussion list
Subject: [EE] Looking for low RDSon N-ch Mosfet

Good day to all.

I'm looking at a buddy's project where his requirements have changed
slightly.

He's currently using a IRF1405 N-channel Mosfet with a data-sheet
rating of RDSon = 5.3 Milli-Ohms.

He's finding that the temperature rise is higher than he likes and is
therefore looking for a Mosfet with lower RDSon.

The circuit operates at less than 30 Vdc and he has at least 10V of
gate drive available.

He'd prefer to stay with the current TO-220 package if possible.

Any suggestions?

Recap: N-channel Mosfet, TO-220 package, Vds 30V max, Id 30A max,
RDSon < 3 Milli-Ohms, > 10V gate drive available.

Many thanks!

dwayne

--
Dwayne Reid   <KILLspamdwaynerKILLspamspamplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2010\07\05@160658 by Olin Lathrop

face picon face
Dwayne Reid wrote:
> He's currently using a IRF1405 N-channel Mosfet with a data-sheet
> rating of RDSon = 5.3 Milli-Ohms.

That sounds pretty good already.

> He's finding that the temperature rise is higher than he likes and is
> therefore looking for a Mosfet with lower RDSon.

So go look, what's the problem?  If I was looking for this I'd check the
usual suspects like IR, Fairchild, and other transistor manufacturers.  I'm
not sure what you're expecting from us short of doing the search for you.

The other thing to consider is paralleling two or more FETs.  5.3mOhms
already sounds pretty good.  To get much better than that probably requires
multiple devices.

> The circuit operates at less than 30 Vdc and he has at least 10V of
> gate drive available.
>
> He'd prefer to stay with the current TO-220 package if possible.

30V at 5.3mOhms is 4.8W, which a TO-220 with reasonable heat sink should be
able to handle well enough.  The real cause is likely not the on resistance
of the FET, but slow switching edges or inadequate heat handling.  Having
something that deals with 30V and 30A dissipate 5W doesn't sound out of
line, so a better description of what exactly the limiting factor is and how
much heat he can really handle is required.  At first glance, things don't
add up.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2010\07\05@160934 by Olin Lathrop

face picon face
Ing. Marcelo Fornaso wrote:
> In continuous regime now you have 30 Amp. X 0.0053 Ohm = 159 mW.

Not even close.

Note that your units don't ballance.  Amps times Ohms is not Watts.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2010\07\05@161836 by RussellMc

face picon face
> He's finding that the temperature rise is higher than he likes and is
> therefore looking for a Mosfet with lower RDSon.
> The circuit operates at less than 30 Vdc and he has at least 10V of
> gate drive available.
> He'd prefer to stay with the current TO-220 package if possible.

> Recap: N-channel Mosfet, TO-220 package, Vds 30V max, Id 30A max,
> RDSon < 3 Milli-Ohms, > 10V gate drive available.

http://www.digikey.com
Excellent parametric selection guide.

If price is no object (and still only $US3.88 in 1/s)

PSMN1R6-30PL,127
568-4900-5-ND
30V, 100A, TO220
1.7 m.ohm nominal Rdson
http://www.nxp.com/documents/data_sheet/PSMN1R6-30PL.pdf

Even better [tm]  AND cheaper (but may not sell in 1's).
IPP015N04N G
IPP015N04N G-ND
40V, 120 A
1.5 milliohm
$2.33/500
http://www.infineon.com/dgdl/IPB015N04N_rev2+0.pdf?folderId=db3a3043156fd5730115c7d50620107c&fileId=db3a30431689f4420116c474de8f0845


OR

Search Digikey , selecting parameters as required.

              Russell

2010\07\05@165805 by Ing. Marcelo Fornaso

picon face
Ups!

Gross mistake! I'm sorry.



________________________________
From: Olin Lathrop <RemoveMEolin_piclistTakeThisOuTspamembedinc.com>
To: Microcontroller discussion list - Public. <spamBeGonepiclistspamBeGonespammit.edu>
Sent: Mon, July 5, 2010 5:09:48 PM
Subject: Re: [EE] Looking for low RDSon N-ch Mosfet

Ing. Marcelo Fornaso wrote:
> In continuous regime now you have 30 Amp. X 0.0053 Ohm = 159 mW.

Not even close.

Note that your units don't ballance.  Amps times Ohms is not Watts.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2010\07\05@184844 by Dwayne Reid

flavicon
face
At 02:18 PM 7/5/2010, RussellMc wrote:
> > He's finding that the temperature rise is higher than he likes and is
> > therefore looking for a Mosfet with lower RDSon.
> > The circuit operates at less than 30 Vdc and he has at least 10V of
> > gate drive available.
> > He'd prefer to stay with the current TO-220 package if possible.

Many thanks for all the suggestions.  I'm checking into them right now.

The problem is that the original spec was for max Id of 10A.  So: the
board was laid out without provision for any heat-sinking other than
soldering the tab of the device to a copper flood area.  Max
temperature rise was predicted to be less than 30C and was actually
measured to be a bit more than 20C.

Now the customer wants to run somewhere between 15A to 30A through
that device.  Unfortunately, the customer doesn't understand (yet)
that changing the spec AFTER the board design has been finished,
accepted, and is now certified costs money.  Real money.

The easiest way to fix this is to select a different Mosfet - one
that has lower RDSon.

Thanks again!

dwayne


{Quote hidden}

--
Dwayne Reid   <TakeThisOuTdwaynerEraseMEspamspam_OUTplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2010\07\05@192136 by Michael Watterson

face picon face
Dwayne Reid wrote:
{Quote hidden}

If the tracks are not big enough for x3 current, then changing the FET
won't help.

Calculate your track resistances from width, length and thickness.


2010\07\05@200929 by Dwayne Reid

flavicon
face
At 05:19 PM 7/5/2010, Michael Watterson wrote:
>
>If the tracks are not big enough for x3 current, then changing the FET
>won't help.
>
>Calculate your track resistances from width, length and thickness.

Hi there.

I didn't specifically mention track width and length but I alluded to
them when I mentioned "copper flood area".  Copper current capacity
is NOT an issue here.  The Mosfet getting hotter than I like IS the
issue - the fix is to reduce the RDSon.

Thanks, though.

dwayne



--
Dwayne Reid   <RemoveMEdwaynerspamTakeThisOuTplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2010\07\05@223016 by RussellMc

face picon face

Solution - see comments at end re Infineon MOSFET - then try one :-).
_______

Simplistically, to maintain temperature as-is, *IF* I^2.Rdson is the main
heating  component then:

20A  needs:

5.3 x (10/20)^2 =~ 1.325 milliohohm Rdson

30 A needs

5.3 x  (10/30)^2 = 0.6 milliohm Rdson

Thoughts:

Spec sheet quoted Rdson is almost invariably for highish frequency very low
duty cycle conditions - die gets little time to heat wrt its thermal time
constant and a long time to cool compared to heating. ie they are cheating
wrt most actual uses. . Real world conditions increase die temperature and
so working Rdson is higher than quoted. Replacement FETs are in same
situation BUT some FETs may have lower Rdson rise with die temperature than
others so this is worth looking at on a a case by case basis.

Saving grace may be extra allowed rise of 30 C rather than actual current
20C.

BUT Rdson increases with die temperature so more rise brings more Rdson
relative to quoted value.

FET with a lower maximum Vgs will be driven harder into enhancement by a
given level of gate drive. Logic level FTs usually have  a lower Vgsmax so
better enhancement per Volt of drive. Very low Vgsth FETs often have a
Vgsmax around 12 Volt so that can be living dangerously depending on how
controlled drive voltage is.

Attention to detail while staying within certification MAY provide useful
gains. Some of the following is "moderately desperate" but moderate
desperation sounds like it may be required.

- Is there mounting thermal grease or a mounting insulator.
- Can the grease be changed in type or increased in quantity.
- Can the thermal washer be changed in type or reduced in thickness (30V
breakdown requires far thinner actual washer material than will be used in
practice - mechanical aspects predominate.
- Will attention to mounting produce better results? - scrupulous surface
preparation, board condition etc.
- Can airflow be altered subtly without violating certification. This could
be a major gain. If physical mounting layout of PCB and environment is not
speciifed then it may be possible to substantially modify airflow in crucial
areas.
- Can radiating surfaces be persuaded to radiate better?
- Is PCB copper thickness controlled by certification?
- Effect of solder mask may be non trivial thermally. Type, thickness,
pattern.
- You said that TO220 is preferred. Can you go to an eg TO247 package
without certification change (sound unlikely but ...).

- Rthjc can vary between devices although it is liable to be 'quite low' in
all cases at this
At 30 A you get about 1 Watt per milliohm.

_______

Best hope ?:

The IPP/IPB015N04N<www.infineon.com/dgdl/IPB015N04N_rev2+0.pdf?folderId=db3a3043156fd5730115c7d50620107c&fileId=db3a30431689f4420116c474de8f0845>
I
mentioned has a Rthjc of 0.6 K/W so only a few degrees of rise junction to
case.
http://www.infineon.com/dgdl/IPB015N04N_rev2+0.pdf?folderId=db3a3043156fd5730115c7d50620107c&fileId=db3a30431689f4420116c474de8f0845


The minimal footprint Rthja for the SMD version of this device is 62 K/W so
you could only achieve ~0.5W at 30C rise at about 1 milliohm Rdson :-(.
Even with a fairly unlikely 6 cm^2 PCB area available * it is 40 C/W or ~~
0.75 W at 1 milliohm for30 C rise.

Note that the Rthjc and Rthja specs are for the SMD version but there is
also a TO220 version available. The SMD is the typical cranked lead and
truncated tab version of the TO220 BUT the extxra metal on the TO220 tab may
help here.

As this device is 1.2 mohm typ, 1.5 mohm max at 100A (!) it's higher than
desireable but extremely good by most standards. Also, the ratings APPEAR
vto be at full DC and not pulsed !. It has some excellent spec sheet curves
including Rdson across temperature  with both typical and 98% curves.

Vgsmax is a nice 20V but all current curves have reached optimum by 10V
Vgs.

All up this looks like a very attractive device to trial in the
circumstances.

(* Device on 40 mm x 40 mm x 1.5 mm epoxy PCB FR4 with 6 cm2 (one layer, 70
μm thick) copper area for drain
connection. PCB is vertical in still air.)




         Russell McMahon

>
>

More... (looser matching)
- Last day of these posts
- In 2010 , 2011 only
- Today
- New search...