Hi all,

http://www.ortodoxism.ro/datasheets/irf/ir2112.pdf

i have this thing in stock, what i want to do is to let it control 6 fets (IRF1104 , N channel) in parrallel (the gate) wich are in the high side configuration.

also the low output needs to control 1 fet (for a bit of regen braking).

Input voltage will be 12/14volts. (car power).

The chip has some pins i have no idea what to do with them.

VSS
VCC
VB
VS

also there are caps / resistors drawn can anyone help me a bit?

with their values or simple explain me where there are for?

In the heigh side configuration we need more gate voltage then suplly voltage right??

Hope you all can help me abit with some values to test it with.

Tks

 

I would take a look at the Application Notes that are provided by I.R. , they are well done and I find them to be very well written and easy to follow.

Start here, http://www.irf.com/technical-info/appnotes.htm

You could also e-mail your local IR representative office, and see if they have any application notes that directly pertain to the IR2112 and/or the design you are trying to achieve. In my past experience, IR are a very helpfull bunch.


The datasheet explains what the pins you listed are used for.

Maybe a little research of a general nature into high-side switching, low-side switching, and motor control involving regenerative braking would explain alot. At the power levels you are considering, there is the potential to destroy alot of components without knowing what went wrong, or the time to prevent it.

 

Yeah the datasheets say names, descriptions but we need to untherstand that names. and i don't do that.

Basicly mosfets is new for me, and i never went to college to learn about programming micro's etc. etc. etc. but i know that with a bit of explanation i can make this.

its not difficult to just build it to the sheme you see on the first page. its just that i need to know the translation of the pins?

Will search for examples altough i hope you can explain a bit the pins..

Tks

 

Try this application note, it will answer alot of your questions.

http://www.irf.com/technical-info/designtp/dt04-4.pdf

 

I think you should at least try to understand the documents instead of following the pictures. As zevon8 said if something goes wrong you won't have a clue how to fix it unless you do some legwork.

VSS is the pin that is conneced to Logic Ground
VCC is a positive supply voltage to which the gate of the lower MOSFET is pulled, called the low side supply
VS is the voltage on the Source of the upper MOSFET, called the high side floating supply return.
VB is voltage which is generated on the chip to drive the gate of a high side N-channel MOSFET to a voltage above VS, called the high side floating supply.

Right there on page 4 of the datasheet, underneath the functional block diagram, is a description of the pins. If you're having this much trouble then I strongly suggest you switch to a simpler project and learn to crawl and walk before trying to do the iron-man tiathalon.

 

Incidentally, IGBT's are crosses between Mosfets and BJT's - typically used for high voltage devices. Personally I've never used them...

So the general idea is that to turn a N-channel mosfet on, you need to put a positive voltage on the gate. Ignoring some of the finer details - the higher the voltage you can put on the gate, the lower the resistance of the mosfet. Obviously check the data sheet for the max gate voltage (Vgs). Exceeding this voltage - typically ~20 - 30V even for a nanosecond will cause the mosfet to fail, usually by short-circuit and lots of smoke. The gate has a capacitance of ~1 - 10nF which is large considering how fast you want it to change between on and off - which is why they make dedicated gate driver chips.

So if you look at the internals of the IR2112, you can see what it does is connect the gate of a mosfet to either a positive voltage or to it's drain - on or off. I assume the lower mosfet is pretty self explanitory - the tricky part is the upper mosfet. If you want the upper mosfet on, you need to connect it's gate to a voltage that is ~12V (or whatever your Mosfet rating) above the load supply - in your case 14+12V. The usual way of doing this is with a "bootstrap" capacitor and diode connected to the switched node - check pg 15 of the design note. One very important thing to keep in mind is that this bootstrap approach will only work if you are PWMing the signal. If the PWM signal ever disappear, or there is some delay on startup, those upper mosfets may suddenly overheat as they turn partially on and dissipate a lot of power.

James

 

Thanxs guys this starts to make sense!

You are right that i need to know to walk before i can make it to run.

But i have build serveral good projects, in fact my locost car is one complete project. suddenly on of them is the reverse and that needs to be electric since the fact that i use a bike engine as powerplant. (no reverse gear).

I wanted to use a startermotor to do the reverse job, i know that they aren't designed for this, but mainly i only need them when they test/aprove my car.

in the UK they just hookup everything an use a simple make button to fly backwards.. i thought that would be allot more abuse for the motor than PWMing 6 mosfets from a HWPWM micro. In my setup i need to look at the type of motor i have (series wounded rotor and stator i guess) because of the regen braking future, i mean to say to be able to cope with the voltage generated when the motor is turning when no PWM signal.

I had 2 things in mind:
- another MOSFET shorts both motor leads?
- a diode after the upper mosfet to mosfet supply (=Vdd) (think its this one)
- a diode after the upper mosfet to ground (in blocking position) doesn't make sense????

The lower mosfet won't be there in my design because i cant see a reason to fit it? in fact i use this chip only to create the High gate voltage and to transform the 5volts pwm signal to mosfet gate signal..

Ok, this said i only have 2 question/doubt left.

I see that between Vdd and Vss is a cap this is there for only smooting out purposes right?? i will fit a 1uF there.

Then the diode i guess a 1amp 0,6volts 1N4001 should do the job. (its rated for my voltage).

then comes the last part the cap wich generates for me the bootstrap (higher then supply voltage) wich value should i take for that? my feeling says 1uF should do the job since there is basicly no difference between supply voltage and controllig voltage.

Comments please?

Tks

p.d. gonna open up the motor to see if i need to care about regen/breaking/ that i have a type of motor wich creates voltage when its turned freely..

 

Well its series wounded

i have 1 connection to positive and ground is case.

It series wounded because i have opened it up and there are no magnets just coils..

Do i need regenbreaking?? / Feuture to minimize generator effect??

Tks

 

I think you will find that regenerative braking is not simply shorting the motor with a FET. Maybe a small toy motor, but a motor of any size is going to detonate your FETs. It can only end in tears.

You need a fairly decent sized resistor to dissipate the heat. It will all depend on the time required to stop the motor, the rate at which it is going to be stopped, etc.

Perfect example, many subway systems use series/parallel configured DC motors, series to get moving, parallel for speed. The braking is handled in a similar way, progressive increase in the resistive load. Braking is regenerative, with many banks of spiral wound metal resistors the size of your forearm. They look like heating coils, and essentially become that. The heat generated is used for heating the train cabin in winter. Even in industrial settings where motors use regenerative breaking, the resistors are big.

 

One thing to note about series wound DC motors... they cannot control their own speed. If for whatever reason there is no load, or it gets disconnected accidentally or otherwise, they can, and often will, accelerate until they destroy themselves. This can happen very quickly by the way, and is generally not enjoyed by anyone.

 

Typically the caps are ceramic and placed as close to the chip as possible. The 1N1004 should be okay for the bootstrap.

I'm not sure if the bootstrap approach will work without the lower mosfet. You should test it out by attaching a meter on the bootstrap voltage and just running a small DC motor to see if it stays above the rail voltage. I think if the motor back-emf approaches the rail voltage, the bootstrap voltage will decrease to just a little bit over the rail voltage.

Putting in reverse biased "freewheeling" diodes is a reasonable idea in any case - the motor will spark, and it's reasonable to shunt them away from the motor. The diodes should be rated for the max expected motor current (but for short durations). The idea is that if you have schottky's with lower Vf's, they'll end up dissipating that extra power instead of heating up the mosfets even more.

The other thing to be aware of is that if this thing gets pushed or goes down a big enough hill, the motor may start generating enough power to exceed 12V, in which case, the freewheeling diodes/parasitic diodes in the mosfets will cause the supply rail to rise above what you'd expect. I think the usual thing is to have another mosfet and a resistor bank setup to dissipate the extra power on the supply rail in order to keep it from blowing out all the parts.

James

 

Mhh ok,

well its a reverse, sow i never will go down hill backward.. and if i do it i apply brake or i select neutral. Sow in fact i need minimal regen / brake / generator feauters in this application on this controller.

Do i need to put the diodes from motor feed to Motor supply (in fact like the mosfets ones are)??

Or do i need to put them accros the motor terminals (in fact from motor feed to ground in blocking direction)

Also did my idea for the caps values seem reasonable?

Wil drawup a sheme when i can.

Tks

 

Ok,
I'm designing the scheme with the Free LT spice program.

I was thinking isn't it easyer to convert 14volts to 10+14 = 24volts????
shouldn't that do the job instead of the bootstrap?

Tks

 

Well hope that someone comes up with a good idea or with a easyer alternative to be able to hook it up!

Attached Images

IR2112StarterScheme.JPG (31.9 KB, 91 views)

 

What do i do? Any help maybe possible guys?? opinions??
if someone knows a better way of pulsing the six mosfets...

come on you guys only need 5min to show up a better possibility..

Tks