Replace Schottky with Mosfet

Go To Last Post
22 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

1. I have a application in which I have schottky connected at input. At input a solar panel is connected & at output battery is connected along with charge controller. Schottky serves two purposes; one is protect the circuit in case reverse connection of panel are connected & second prevents reverse flow of current from battery to solar panel known as dark current.

 

2. Max current from panel is 10A, so I have connected two parallel schottky PDS1040. Datasheet of PDs1040 state max Vf drop of diode around 0.55V at 10A(wrost case Tj=-65) & 0.52V at 5A. 
Max panel voltage goes to around 21V.

 

3. These two diodes gets heat up very quickly & I want to replace them with Mosfet. Now I was looking for how I can replace it with Mosfet.

 

4. First option is to use ideal diode. I have checked linear LTC4358, which is for 5A. Operating voltage range is 9V-26.5V. I am thinking I will connect these two in parallel. One problem is its lower voltage range is 9V so from datasheet I didn't understand what will happen when voltage falls below 9V, will it turn off, or it may partially tun on which can increase the voltage across mosfet & heat it up. Second it very costly around $5.38 unit price.

 

5. Other option I found here ( https://www.google.co.in/search?... )

It uses p channel mosfet & BCM856DS. Its an doubled matched transistor. 

 

6. Any other circuit which I can use?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Of course the diodes get hot - combined you've got 0.55V times 10A. That equals 5.5W. According to the specs, the best C/W is 50C/W, so each diode is getting to 125C over ambient. You want much better heatsinking. 

Let's do some numbers:

Assume the mosfet has an on resistance of 0.01R. At 10A, that is a 0.1V drop or 1W. This could be good or bad depending on the package and heatsinking. Say a TO220 package - around 62C/W, so it will get to 62C above ambient, so at a nominal 25C ambient, it will be too hot to the touch at 87C. If you choose a surface mount package, the results will be most likely be worse, so you need some form of heatsinking. Most mosfets have an intrinsic substrate diode which will spoil your day for reverse current, so you need a diode in series!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But a mosfet could be chosen based on a much lower on resistance. 

For example the AUIRF7739L2TR has a max on resistance of 1 milli ohm. So at 10 amps the power dissipation would be 100mW which would be ok.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

... and Findchips reports zero stock of AUIRF7739L2TR anywhere. devil

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Most mosfets have an intrinsic substrate diode which will spoil your day for reverse current, so you need a diode in series!

When used for polarity protection, the MOSFET itself is 'reverse' biased, so the substrate diode is part of the magic.

http://www.ozqrp.com/rev_pol.html

 

With normal polarity, the substrate diode is initially forward-biased, which puts the source about 1V lower than the drain.  With the gate pulled low via the 33K (and protected by the zener), the MOSFET switches on full.  With reverse polarity, the substrate diode is reverse biased, and the gate is pulled to within VF of the zener, turning off the MOSFET.

 

If you can tolerate low-side switching, you can use an N-channel with a lower RDS(ON).  For example, and IRF1404 has RDS(ON) of 0.004R, and max IF of 202A.  At 10A, VF would be 40 mV, and power dissipation would be 0.4W.  With a low junction-to-ambient thermal resistance of 62 °C/W, (and junction-to-case of 0.45 °C/W) even with no heatsink the junction would keep about 25 °C above ambient.  TJ(max) is 175 °C.

 

 

 

 

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

Last Edited: Sat. Mar 19, 2016 - 08:12 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I thought the idea of using a mosfet as a diode replacement paid off when 10A * .55V diode drop = 5.5W is more than drain current * Ron? They use this trick in synchronous rectifiers, right?

 

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

and diodes in parallel - to work, they need low ohm series resistor for each. High wattage.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

joeymorin wrote:

When used for polarity protection, the MOSFET itself is 'reverse' biased, so the substrate diode is part of the magic.

http://www.ozqrp.com/rev_pol.html

 

 

Assume the source (S) is connected to a 12V battery, and the drain (D) is connected to a solar array that is now at 8V (limited light).

Won't the FET still be conducting (G=0V, S=12V), allowing the current from the battery to flow to the solar array?

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, this was the point i was trying to make. I chose 10milliohm as a starting point. Then there is the issue of vgs vs rdson that needs to be considered.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Won't the FET still be conducting (G=0V, S=12V), allowing the current from the battery to flow to the solar array?

Ah, yes, I see what you're getting at, the 'dark current' as the OP has said.  Well, the gate could be overridden by, say, a comparator with inputs on S and D.  When D < S, force the gate off.  You'll need to protect against the panel being wired up in reverse, though.

 

As for the transfer function, you select a device which has a transfer function which results in lower dissipation that a diode would, over the range of expected voltages and currents.

 

 

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

joeymorin wrote:

Well, the gate could be overridden by, say, a comparator with inputs on S and D. 

 

Below is the circuit linked to in post#1:

 

As you can see, the dual transistor is functioning as a comparator, shutting the FET off if the source voltage is greater than the drain voltage.

 

The problem with the above circuit is that the FET specified has a VSD of only 12V, a max ID of 4.3A, and a 50 mili-ohm resistance.

Also, the dual transistors have a base-emitter reverse breakdown voltage of only 5V, which is a problem if the battery voltage is greater than 5.6V.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

https://www.google.ca/search?q=ideal+diode+comparator&tbm=isch

 

A number of ideas.

 

 

 

 

Linear tech have a range of "ideal diodes" based on this idea.

look at LTC4358 for example.

here's an article on them

http://cds.linear.com/docs/LT Magazine/LTMag-V18N02-14-LTC4355_57-MeilissaLum.pdf

they have a 9-26volt, 5A range, and cost $2-$3.

PDF link is broken, but you can get if from the horse's mouth:

http://www.linear.com/product/LTC4355

http://www.linear.com/docs/26512

 

EDIT:  Although the above suffers from the same 9V minimum mentioned in the OP, a functional equivalent could be constructed from a combination of discrete components and a low-power OP with more favourable limits.  While the 4358 has a 5A maximum, but the 4355 (and 4357) work with external FETs.

 

But isn't that 9V minimum for the VDD of the device, rather than the input side?  As such, if you're charging a 12V battery you should be fine so long as the >>battery<< doesn't dip below 9V.  The PV can drop to 0 and the device itself will still operate correctly.  Once the PV voltage again rises above the battery voltage, the device will conduct.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

Last Edited: Sun. Mar 20, 2016 - 03:10 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I want reverse polarity protection fro solar panel only & reverse current going into panel.

Currently I have decided that I will place three PDS1040 in parallel & large coper pads with vias on it on PCB for 10A current & will see the result.

 

Since ideal diode from linear is around $5.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I came across this http://www.instructables.com/id/Reverse-polarity-protection-for-your-circuit-with/ a few days ago, might be usefull, He replaces diode with a mosfet,

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Read above, this has already been covered. The problem is when you have a solar cell and battery. The current can flow both ways.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

1. Can I connect 4 diodes in parallel. Max total current is 10A & volatge is 35V I am using PDS1040.

2. However I was reading on internet, then diode Vf decreases as its get more heat up & this may cause current hogging. One solution suggested that add series resistor for each.
3. Can I add 0.010 resistor in front of each diode?
So that max dissipation across in worst case is 10*10*0.010 = 1W.

4. Will 10 milliohm is ok for this application? ( http://www.digikey.com/product-d... )

Ckt is attached.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry Ross, was just an example of what one can use.

There are many choices out there but my whole point was regarding rds on wink

 

Paul

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I use an LTC4412 driving a p-channel fet si7141DP in my solar controller for this purpose at 10 amps.   Works well.

 

Jim

 

 

 

(Possum Lodge oath) Quando omni flunkus, moritati.

"I thought growing old would take longer"

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That is a sweet little chip. Thanks for the tip.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have made the PCB. Will test it.
But I want to know whether 0.010 ohm resistor is ok to balance the current in case of mismatch
 or should I use higher value resistor.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You need to read the diode datasheet carefully. What is the max variation of the voltage drop?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But in any case how to select which resistor value is suitable. I have attached the pic of If Vs Vf for PDS1040.
In my application maxx current is 10A total with 35V max.
Pic shows max Vf mentioned in red line. So how to calculate most optimal resistor value.

Attachment(s):