Suggestion for reverse diode over coil?

Guest · Posted: Tue Dec 22, 2009 6:15 am

I want some diode(fast, cheap, and small) to take the reverse current from a coil, but what diode to select?

Forward current in the coil is is about(not peak!) 3,5A max without PWM! and power supply is 24VDV.

The NFET is controlled by a PIC(PWM 10Khz duty 30%)

---24---FET_Drain---coil---FET_Source---GND.
The diode is placed _over_ the coil, the FET is a IRF7413_NFET.

Hints?

Ttelmah · Guest

The energy that could be in the coil, depends on it's inductance. You need to find this out. There are quite a few other factors that acn come into play (internal DC resistance limits the maximum current, and the operating termperature/colling, affects how close to it's limit you can push the diode). However at this relatively low current these are probably not significant.

Best Wishes

Guest · Posted: Tue Dec 22, 2009 8:57 am

Yes thats right with the inductance, but I don't know because it must fit a lot difference coil type.
This diode is used in switch mode power but is large UF5408.
What I need is a general diode for my solenoid actuator Smile

mbradley · Joined: 11 Jul 2009 Posts: 118 Location: California, USA

If I may suggest, use a transorb, purpose built for this:

http://se.mouser.com/Search/Refine.aspx?Keyword=Transil
_________________
Michael Bradley
www.mculabs.com
Open Drivers and Projects

mkuang · Joined: 14 Dec 2007 Posts: 257

Ttelmah · Guest

Yes.
For small circuits, 1N4148. For slightly larger, 1N4004.
Part of the answer depends on the nature of the coil, and what it is used for, so (for instance), it may be worth considering using a diode and a Zener, to allow faster switch off (but at extra cost), and then the main drive transistor, needs to be able to handle the supply voltage pus the zener voltage.
For power switching, Schottky diodes give better efficiencies, but at higher cost.

Best Wishes

mkuang · Joined: 14 Dec 2007 Posts: 257

FvM · Joined: 27 Aug 2008 Posts: 2337 Location: Germany

I thought the answer would be obvious, but the discussion shows that it isn't.

Operating a coil with 10 kHz PWM is equivalent to operating a regular buck converter for most actuators,
because you get a continuous DC current with some ripple. That means, that the diode's current rating must
be at least Idc * (1-duty cycle). A fast 1 A diode would be sufficient, 1N914 or 1N4148 wouldn't, 1N400x
too slow. A 1A schottky diode gives the least power dissipation. To operate the coil over the full duty cycle
range with 24 V supply would at least require a 2A diode.

mkuang · Joined: 14 Dec 2007 Posts: 257

Is the OP turning the coil on and off at 10Khz ? What kind of coil is that ? Then you need a Schottky, neither cheap nor small and leaky also then. I suggest drop the frequency to 1khz and use a cheap diode.

Guest · Posted: Tue Dec 22, 2009 12:27 pm

This is an interest discussion about diode. I would never choice a 1N4148 or similar for that. But a transorb will not fit in because the coil can be 5V or 24V it depend on the solution where it will be used. I thing a fast schottky will be what i select.

Maybe some one have a favor to use in this situation. Spec can be 60V 3A, fast, small, SMD is ok but not an issue:)

mkuang · Joined: 14 Dec 2007 Posts: 257

Something that handles 3A continuously cannot possibly be small. They are also very leaky, something on the order of a mA or so. I have used the MBR3100G (rated for 3A continuous and 100V ) as part of a switching regulator. It costs $0.57 when you buy in small quantities so I guess it is pretty cheap. It is a thruhole part (the SM version also exists) and if you want to conserve board space you can stand it up on one end.

bungee- · Joined: 27 Jun 2007 Posts: 206

You did not write what means small diode. With your data I would select something like that: http://uk.farnell.com/international-rectifier/30bq100pbf/diode-schottky-3a-100v/dp/1463216

Ttelmah · Guest

The 'point' about the zener approach, is that you don't trap the coil at the diode voltage, but allow the voltage in the coil to go significantly negative. This allows lower currents to be used, reducing both resistive losses in the coil, and in the whole trap circuit. At high powers, this becomes 'essential', however at low currents it is generally not worth worrying about.
Funnily enough, the 1N4148, is a surprisingly 'good' small trap diode, provided you are happy with the fairly high voltage that you will see across the coil, because it's recovery is faster than most cheap rectifier diodes.
Remember that the trap diode does not handle current continuously.
Personally, given the indeterminate nature of the coil, I'd 'overrate' the diode, to handle the 'worst cae', and a bit more. The cost of this will be tiny, compared to the FETs. Remember also, that heat dissipation in the FET, will be massively dependant on how well itis driven. A PIC output, able to deliver 25mA, is not suitable to switch a FET with perhaps 1000pF gate capacitance at any reasonable speed...

Best Wishes

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

If you measure the inductance of a solenoid coil - STATIC - it will NOT be the same as if the slug (core whatever U call it ) is in MOTION - as this will alter - usually RAISE the effective back EMF generated that you are wanting to supress. IN the process making the inductor a show a higher DYNAMIC value of 'L' than the value then you measure w/o slug motion.

In general I always use a well over rated Absorbtion diode
for such fun as this question.

Guest · Posted: Thu Dec 24, 2009 3:25 am

[quote="Ttelmah"A PIC output, able to deliver 25mA, is not suitable to switch a FET with perhaps 1000pF gate capacitance at any reasonable speed...
Best Wishes[/quote]

Now I'am confused, what will you then suggest. I use a N-FET (low side). Will you use P-FET (placed high side) and then a NPN transistor to drive the FET?

A little explanation please?