Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

mosfet amp transistors getting HOT!

Status
Not open for further replies.
The datasheets for your Mosfets show that the N-channel one conducts 18.6 times better than the P-channel one so the P-channel ones are much warmer.
The piezo current is not high so the "shoot-through" current is what causes the heating of the P-channel Mosfets.

You cannot bolt both P-channel Mosfets to the same heatsink unless they have mica or polymer electrical insulation.
A heatsink inside a closed box is almost useless unless the box is huge. Use the surface of an aluminum box as the heatsink. The air outside the box will cool it.
 
perhaps look for a faster P channel mosfet or?
I am under the imprssion a similar circuit is available on a single chip?
need to look over all the past posts to find reference.
 
Would an H bridge controller chip prevent heating?
not ral sure if available but a one chip h bridge?
 
perhaps look for a faster P channel mosfet or?
Mosfets switch extremely fast. It is the high capacitance at their gate that slows down their driving circuit.
You have a 1k resistor that takes a long time to charge or discharge the gate capacitance. a lower value resistor or an active transistor will produce a higher current which will charge and gischarge the gate capacitance quicker.

I am under the imprssion a similar circuit is available on a single chip?
There are Mosfet driver ICs available. I don't know any and don't know if they will work in your circuit.
 
off to radio shack for resistor to go parallel with the smd that I have in the present circuit.
ben doing some reading and found that the N channel mosfets have a lower Rson than the P channel. this is why they get warm.
Mentioned an N channel H bridge ?
lets see, two 1 k resistors in parallel equal a mess but this is a proto-type.
you mentioned going lower like 330 ohms?
 
did the math and a 500 ohm will do the deed. thinking of going in steps, 500 ohm then try the 333.33 ohm.
not sure what I am going to do. probally go with the 500 in parallel w/ 1k = 333.33ohms
recess!!!
 
I have never used surface-mount resistors. They are small so do not dissipate much heat. With 12V across 500 ohms half the time then the heating is half of (12 squared/500 ohms)= 0.144W. Is 500 ohms a standard value? 470 ohms is a standard for ordinary resistors.
 
smd resistors I assume are the same standard values. my plan is to use a 1/4 watt metal film resistor with leads then just solder to pcb traces. no holes etc. don't have desolder braid etc. only a soldering iron (on vacation in Calif. 1200 miles from home) and not sure how big that soldering iron is??
hopefully get this resistor in before I leave Washington state in 12 days (3000 miles round trip in 3 weeks, lots of driving)
 
Redesigning the CRITTER RIDDER

using all SMD except for the MOSFETS. Mounting the P channel mosfets onto TO-220 heat sinks on top of the enclosure using two seperate heat sinks. Going to use 330 ohm in place of the 1k.
Now the question is concerning this quote
"The datasheets for your Mosfets show that the N-channel one conducts 18.6 times better than the P-channel one so the P-channel ones are much warmer.
The piezo current is not high so the "shoot-through" current is what causes the heating of the P-channel Mosfets."

If I can find a faster P channel then we would be better concerning the heat build up of the P channel mosfets BUT what am I looking for to determine how the mosfets conducts.
Thinking resistance or trans-conductance? from the data sheet.
looking at these two mosfets from Allied Electronics.
https://www.electro-tech-online.com/custompdfs/2011/03/273-1489PDF.pdf
https://www.electro-tech-online.com/custompdfs/2011/03/503-0577PDF.pdf
 
One of the most important specs for Mosfets is the on-resistance. If the on-resistance is very low then most of the power goes to the load and only a small amount of power heats the Mosfet. If the on-resistance is high then much of the power that should go to the load heats the Mosfet instead.

The drive circuit for a Mosfet must switch fast, then the Mosfet will switch fast. Then the shoot-through current lasts a very short amount of time then the heating is low.
Mosfets have a high input capacitance that takes a lot of current to quickly charge and discharge.

The Mosfets you selected have a P-channel that works fairly well with a 4.5V or 10V gate voltage. Its max on-resistance with a 10V gate voltage is 0.105 ohms when cold and its gate charge is 47nC max.

The N-channel Mosfet is spec'd only with a 10V gate voltage and its on-resistance is 0.028 ohms max when cold and its gate charge is 67nC max.
 
So if I locate mosfets with low low on resistance then hating will be less.
Need to locate a lower resistance and low gate voltage. I think Allied had some really low on resistance mosfets.
Reason for using allied is lower shipping costs and some of the parts I need are lower cost.
I shouldn't have to change anything in the H bridge amp circuit I hope.
 
Your kind of stuck. You either need to slow down the turn on time of the fets or limit the shoot through current. To minimize rework I think I would try to limit the shoot through current by placing a resistor (1 ohm would probably do it) in the 12 volt line feeding the top fets. The 330 ohm resistors still need to go straight to +12. The rating on the resistor will depend on the frequency you are running the piezo at and the current through the piezo.
 
Like I tried to say about 30 posts ago; that H-bridge circuit is brain-dead because it has NO provision to prevent shoot through!!!!
 
Last edited:
PCboard revised

IT WORKS GREAT!!
Just need to revise design as I want to go w/ all smd except the 4- Mosfets, two of which (P channel) are to be mounted on the top of the enclosure w/ heat sinks.
Trying to figure out heat sink data as well
The P channel Mosfets are the only ones getting warm/hot.
It was suggested by Audio Guru to replace R1 & R2 w/ 330 ohm instead of 1K.
Going w/ parts from Allied Electronics and all smd the cost of parts is less.
the Mosfets I am looking at are
N channel--http://www.alliedelec.com/Images/Products/Datasheets/BM/VISHAY_SEMICONDUCTOR_ITALIANA/503-0577.PDF

P channel--http://www.alliedelec.com/Images/Products/Datasheets/BM/INTERNATIONAL_RECTIFIER/273-1489.PDF
the on resistance is a little lower
 

Attachments

  • smd cr.png
    smd cr.png
    34.3 KB · Views: 200
Hi Mr DEB,
Your new schematic software produces an awful negative image instead of a positive image like most are. But I am glad that the lines are straight.
It also produces a simple box instead of a Mosfet.

The schematic has no values for the resistors so it might not work.
 
revised design

Trying to cut down the amount of heat.
The addition of a 1 ohm resistor might work but going to try the 330 ohm first as the circuit board is already to etch. Just inserting a 330 in place of the 1K (R1 & R2)
PLUS I am using heat sinks instead of a short piece of aluminium for each P channel TO-220 Mosfet.
The circuit works great. The deer, Elk, as well as birds run for cover within 100 yard distance from vehicle.
Presently have unit in passenger compartment but plan to mount under the hood by the grill for air flow.
 

Attachments

  • smd cr.png
    smd cr.png
    34.3 KB · Views: 245
Hi Mr DEB,
Your circuit works great since it shoos away animals and birds that are 100 yards away.

It is too bad that your new schematic software makes a negative image.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top