Why OH why are there so many N channel mosfet's ???
As the Title suggest..
I'm looking to build a high power H-Bridge ..coupled to Arduino ...controlled by 2.4GHz RC.
My question(s)is / are
Where can I find a source for complementary Mosfet's ,in the order of 300w ?
Or will it take old fashioned Leg work..?
What parameters should I take into consideration?
Also I'm having trouble grasping in plain English a dual mosfet driver that has an inverting output..
Does that mean , as an example ..the non-inverting side handles the - flow
inverting side handles +flow?
Which would reduce the need to invert a PWM ,or voltage signal to the driver IC.
I'm getting the idea that I can use TTL on the Driver to load up the power mosfet's if the mosfet has a VGS of 10v ..and these should be ratiometric & symmetrical in resolution ?
Mosfets can be so much more than a basic transistor.
Because 99% of the time we are working with positive voltages (and current flow in one direction). Its the same reason you see a lot more NPN transistors than PNP transistors.
Also I'm having trouble grasping in plain English a dual mosfet driver that has an inverting output..
Does that mean , as an example ..the non-inverting side handles the - flow
inverting side handles +flow?
Which would reduce the need to invert a PWM ,or voltage signal to the driver IC.
Yeah... Normally, you want a "speed" input and a "direction" input. So, the PWM is inverted in hardware. (And the direction input inverted in hardware to one-half of the bridge.)
I'm getting the idea that I can use TTL on the Driver to load up the power mosfet's if the mosfet has a VGS of 10v ..and these should be ratiometric & symmetrical in resolution ?
TTL is 5V, like the Arduino. If you use a MOSFET that requires 10V on the gate, you need a transistor or another FET/MOSFET (and of course a 10V or higher voltage source) to boost the drive voltage.
Where can I find a source for complementary Mosfet's ,in the order of 300w ?
Or will it take old fashioned Leg work..?
What parameters should I take into consideration?
MOSFETs are usually characterized by the voltage that they will be subjected to, the current that they must carry, and their resistance when turned on. Also the gate voltage needed to achieve full turn on.
There are many other parameters, but these are probably the most meaningful at the start of your search.
Chris1448:
Why OH why are there so many N channel mosfet's ???
First off - I'm not an expert on mosfets. I have read that the reason behind more n-channel mosfets than p-channel is that p-channel fets are more difficult to construct (maybe in high-amperage versions)? Maybe I mis-read or mis-remember.
Chris1448:
I'm looking to build a high power H-Bridge ..coupled to Arduino ...controlled by 2.4GHz RC.
Why? This is an important question.
If you are doing do for learning how - then go for it. If you are doing so because a mosfet h-bridge in the amperage you need is expensive, well...
...you might find yourself blowing up ($$) parts (the fets) as you learn to construct your h-bridge, and find yourself at the same amount of money or more as it would have cost to purchase one. Just keep that in mind.
Chris1448:
Where can I find a source for complementary Mosfet's ,in the order of 300w ?
Also know that you can parallel mosfets to increase current handling (at the expense of greater gate capacitance, which effects switching speed - mainly switching off - IIRC?). I think you can counter that by getting low RDS(on) value mosfets (as low as you can go - not that generally mosfets with extremely low on-resistance are logic level mosfets).
As far as "300W" is concerned - you don't mention what you mean by that? Are you wanting to control that much power, and if so - at what voltage?
Chris1448:
Also I'm having trouble grasping in plain English a dual mosfet driver that has an inverting output..
I'm not sure what you mean by this, so I can't really answer it for you - but I would suggest looking into and purchasing what are called "half-bridge n-channel mosfet drivers". They are made for building an h-bridge, generally using 10V gate mosfets (ie - the most common you find). You can then use all n-channel fets, plus a few other parts (generally a cap for the boost voltage needs on the high-side - note, btw, that such bridges -must- be PWM driven, and you can drive the bridge at "full 100% voltage", because the PWM switching action is needed to keep the boost supply working). You need two of these drivers to build a full bridge.
On the mosfet side of things, you can gang up (parallel) multiple fets on each leg (make the numbers the same on each leg of course) to increase current handling capacity. Select the fets to use based on the datasheet and your needs. Finally - try to purchase them by the tube or reel or whatever - so you can get a "batch set" which all should be fairly well matched.
Or you can try another option - though it may be more expensive in the long run. You could try a "hybrid" h-bridge.
This is one where you construct the h-bridge switching using relays (and transistor/mosfet drivers) - generally a couple of SPDT relays (Bosch-style automotive relays can get you up to 40-60 amps @ 12vdc - if you need more, look into reversing contactors meant for off-road winches - those can get you over 100 amps).
Then - on the low side (ground) of the relay h-bridge, you put your n-channel mosfet(s) and switch those (use logic-level here, it will be easiest).
Just make sure you set the PWM to 0 before switching the contacts of the relays or contactor (otherwise you'll weld the contacts together); so before switching direction, set the PWM to 0, then wait a couple of milliseconds for the contact to "settle", then set your PWM to cause the motor to move in the opposite direct (also - don't "slam" the PWM - ramp it up and down gracefully).
Again though, depending on your needs, you may find pre-built h-bridges to be the better option. For instance, low-cost (from China) BTS7960-based module:
Can handle up to 43A and 32 volts - and they are dirt cheap on ebay and elsewhere. Each module can control one motor. You can even find dual-module versions to control two motors. The BTS7960 can also be bought by itself, but you'll end up spending more trying to build your own driver with it, than you would if you had just purchase the pre-built PCB.
Based on your limited specs you have given - one of these should work for your needs (but give us the full scoop before rushing out to purchase something).
So I can run different motors ,mostly automotive stuff ,12V DC.
Mainly 12 volt because 12volt alternators are easy to come by, 24volt not so common ..but available
12volt work batteries ..not so common ..12v starting batteries ..very common.
Very Heavy..either way.
. Hence ..lower voltage trades of to increased current .
I just would like to RC my lawn mower and my snow blower.. and not have to build two different systems of propulsion. Independent left and right propulsion ..Tank Style I guess its called.
My snow blower is significantly heavier than my lawn mower..Of course...
And before everyone freaks out. An emergency shut down routine ..is my first concern..
Learned along time ago not to start something I cant put a stop to.
Depending On which way a guy reads the spec sheets things get silly.
IRFP1405..55v , 95 amp , with a 310 watt power rating
IRFP4004pbf 40v 195amp 380 watts power rating....
the motors I'm looking at are 1/5th HP permanent magnet gearheads that consume 210watts @ 12 volt or 17.5amps
What the stall current is I don't know.. More over ..because these heads are oil filled, and greased motor bearings ..what the stall current would be @ -15C .. I don't know.
For all I know it may take a 100amp to get them moving and 50 amp in the first few revolutions just to get the juice moving.. before finally setting in to its groove .. hopfully not tripping the breakers to early :o .
Chris1448:
Why OH why are there so many N channel mosfet's ???
Because in silicon electrons are three times more mobile than holes, so n-channel FETs are
3 times better than p-channel FETs. That's a huge performance factor.
As the Title suggest..
I'm looking to build a high power H-Bridge ..coupled to Arduino ...controlled by 2.4GHz RC.
My question(s)is / are
Where can I find a source for complementary Mosfet's ,in the order of 300w ?
H-bridges are typically built from all n-MOSFETs. High side MOSFET gate supply is bootstrapped
from the output switching waveform.
Or will it take old fashioned Leg work..?
What parameters should I take into consideration?
Also I'm having trouble grasping in plain English a dual mosfet driver that has an inverting output..
part number?
Does that mean , as an example ..the non-inverting side handles the - flow
inverting side handles +flow?
Which would reduce the need to invert a PWM ,or voltage signal to the driver IC.
Inverting the signal is a red-herring, the microcontroller can do that, what relevant is getting
a suitable high-low MOSFET driver for each 1/2 H-bridge.
I'm getting the idea that I can use TTL on the Driver to load up the power mosfet's if the mosfet has a VGS of 10v ..and these should be ratiometric & symmetrical in resolution ?
No, MOSFET driver chips drive MOSFETs, that's why they exist (there are 1000's available).
MOSFET drivers often provide very important protection functions such as under-volt shutdown,
shoot-through prevention (deadtime).
Have a good read of the datasheets for chips like the IRS2001, HIP4081A, FAN7388, many more
are available.
Mosfets can be so much more than a basic transistor.
Power MOSFETs are always(*) used for switching, all power electronics is switching today.
IRFP1405..55v , 95 amp , with a 310 watt power rating
IRFP4004pbf 40v 195amp 380 watts power rating....
310 W and 380 W are Absolute Maximum ratings. I hope that you are not anywhere close to this or reliability will suffer. Those power ratings take into account current, resistance when on, turn on time (and associated resistance) and turn off time (and associated resistance). Those power ratings are for the MOSFETs, NOT your motors!
The power ratings are important, but you are unlikely to be equipped to do this sort of calculation. For now, concentrate on the current that your motors use, the voltages that you are dealing with, the resistance of the MOSFETs when on, how you are going to get rid of the heat, how you are going to drive the MOSFETs, how you are going to avoid exceeding the voltage rating of the gate to source voltage (+/- 20 volts Absolute Maximum, which is pretty common), and how you are going to avoid shoot-through (which most drivers handle).
Typically you'd run MOSFETs at 2 to 10% of their maximum power ratings these days, to avoid
expensive large heatsinks and fans. Just pick one with a low enough on-resistance to be easy
to cool. 0.001 ohm is available these days, that can handle 50A and dissipate only 2.5W when
on. The maximum power and maximum current ratings are often the same thing in datasheets,
and are often quoted just for the die, not the package, so you'll see TO220 devices quoted as
handling 170A (at which level the leads melt). The sensible approach is to calculate heat
dissipation from the on-resistance and keep it reasonable.
Switching losses may dominate in practice, not on-resistance losses, if using PWM. For high power
bridge design you've got to keep a handle on both on-resistance losses and switching losses,
which means doing some simple math and choosing appropriate gate driver currents. Protection
is everything at high power levels - its extremely easy to pop MOSFETs when the supply is
capable of kW's of instantaneous power - choose good drivers.
For high voltage systems you'd use IGBTs, not MOSFETs, as these are much more robust with
large voltage swings involved.
Explains why I couldn't get my first circuit to work.
It should be noted ..I'm a scavenger of sorts..buying parts usually happens when I either making something I need and cant find the part or I'm finalising the circuit..
This is a hobby afterall..
Now..that said
R&D
Ive scored some boards out of a tread mill..which have two IRFP250N mosfets onboard.
Those are 200V MOSFETs - on-resistance performance is much worse for high voltage MOSFETs, though
that device is actually extremely good (for a 200V MOSFET - a 30V MOSFET of the same generation/size
would be a few milliohms or so).
So you'll need some heatsinking for 17.5A as that's 5W or so dissipated, and a MOSFET driver to boost
logic signal upto 12V for the gate - word of warning the 12V supply for the MOSFET driver must be
clean, don't use the main 12V directly as switching voltage spikes/transients on that could fry things, so
some circuit to produce a cleaner 12V for the gate driver is really a good idea. (simple as a resistor and 15V
zener will do). You can also add 15V zener across gate+source to further protect the MOSFETs themselves.
If you need an H-bridge you'll need high/low gate drivers and 4 MOSFETs - for one direction only
a single MOSFET will be enough, though for active braking 2 are needed in a 1/2-H-bridge.
Stall current is nothing to do with the grease in the gearboxes! However you are right in that stall current
is larger at lower temperatures because the winding resistance drops a bit with lower temperature.
Yes I do need bi-directional..for now a 20 amp polarity reversing switch will do...If need be 12v automotive realys could do the same..
But for now I have the reversing switch out of a 120v electric impact gun.
Heat sinking isn't a problem, have any shape or size,, aluminum or copper, brass if it comes down to it.
I do have mica isolators ,as well as the soft "rubbery" type and dielectric thermal grease..Clean up old amplifiers in my spare time.
And yes ..grease dosent build into stall current ..but I can just see the cold lubricant as Drag the motor must overcome in tandem to its own static inertia..Try pulling over a 20HP engine in summer and the same in the dead of winter after its been frozen for a month.Never want to relive that episode..
After examining the boards , I notice theres a 100k resistor (approx. 1watt) on each of the mosfet gates.
There is also ,as you said , a zener diode and another Resistor :gate to source.
And two transistors the driver NPN/PNP ..which the manufacturer so kindly ground the face's off.. actually most of the through hole components involving PWM and driver circuit have been defaced..
Theres also a custom made "sister" board installed in the path..probably has to do with speed and torque?
I found another treadmill with a IRFP250M .. either 2 more N's or one more M should set me up for a H bridge
Chris1448:
And yes ..grease dosent build into stall current ..but I can just see the cold lubricant as Drag the motor must overcome in tandem to its own static inertia..Try pulling over a 20HP engine in summer and the same in the dead of winter after its been frozen for a month.Never want to relive that episode..
No, grease adds dynamic friction, which is neither static friction nor inertia. Only static friction needs
torque to overcome to get the motor turning in the first instance - the inertia affects how fast it can
accelerate upto speed, the dynamic friction (+ static) affects the power needed to run at a particular speed,
and thus the efficiency and top speed.
Typical DC/universal motors have a lot of static friction, and most of this is due to the brushes and
commutator. This limits efficiency at low output torque, and is one of the reasons brushless motors are
becoming dominant (the other two being low-maintenance with no brushes to wear out, no sparks).
True , very True .
If I had the coin I would use brushless motors possibly even stepper motors ..
The amount of power they can generate is wild..
Have 4 brushless motors on my quad-copter.
Out and about today I got a board that has 4-four : fqp50n06
Well any motor can generate lots of power if you force cool the windings and run it at high rpm...
The limiting factor is the properties of silicon-steel, copper and neodymium magnets, whatever the
motor configuration. Not having brushes does help, but good brushes can be found (copper-filled).
The highest performing electric motors are superconducting ones, BTW - gets rid of the copper-losses
entirely.