# Mosfet/H-bridge Vgs low speed control

So, I'm planning on making a very basic h-bridge (like so https://i.stack.imgur.com/s9Uvc.png) to control a 12Vdc/11 amp max motor via a teensy 3.6.

The teensy's analog pins only output up to 3.3v. The power source is an 11.1v lipo so really the voltage will run between 9v and 12.6v.

Here is the mosfet I plan to use: IRLR4132TRPBF Infineon / IR | Mouser Europe

It has a Vgs of 1.35v.

Does this mean that at 1.35v/3.3v is the minimum ratio of pwm I can get? So I'd be able to do 104/255 or 4.9v out of 12 on the motor? Will it just suddenly turn on once the pwm is at 104, or will I be able to get nice, slow speed control?

I'm having a hard time clarifying whether this will be the case, so any info is appreciated.

1. It has a THRESHOLD Vgs of 1.3V. That's the level at which it is just about starting to conduct. It isn't fully switched on. Fortunately that MOSFET probably will switch on fully by 3.3V which is the voltage on the output pin.

2. Sorry but that's a complete misunderstanding of PWM. When you write a PWM signal it isn't a varying voltage you get. It's a pulsed frequency and what varies are the on and off times. So the PWM ratio which controls the speed of a motor is nothing to do with voltages. It is the amount of TIME that the signal is ON compared to the TIME it is OFF. So e.g. on for 1ms, off for 9ms = slow, on for 9ms, off for 1ms =fast.

Steve

snowskijunky:
Here is the mosfet I plan to use: https://www.mouser.com/productdetail/942-irlr4132trpbf

It has a Vgs of 1.35v.

Hmm, I couldn't find an actual datasheet for that part, I only found the IRLB4132

As slipstick says:

It has a THRESHOLD Vgs of 1.3V. That's the level at which it is just about starting to conduct.

I would think it is similar to this:
Diagram:

See where the On Resistance is at 3.3v - off the top of the graph.
So you might get away with it, but it would be better to use another transistor to switch the MOSFET.

slipstick's covered the PWM thing nicely.

Yours,
TonyWilk

Actually the threshold is 1.8V nominal, not 1.3V, that's the minimum value for it.

It is rated at 4.6 milliohm max at Vgs=4.5V, so its good for 5V drive, the datasheet does not imply it
will work at anything less than 4.5V of gate drive, and in fact thats far too close to the nominal plateau
voltage.

Use part number IRLB4132 to find the datasheet, the IRLR is not a good search term.

slipstick:

1. It has a THRESHOLD Vgs of 1.3V. That's the level at which it is just about starting to conduct. It isn't fully switched on. Fortunately that MOSFET probably will switch on fully by 3.3V which is the voltage on the output pin.

2. Sorry but that's a complete misunderstanding of PWM. When you write a PWM signal it isn't a varying voltage you get. It's a pulsed frequency and what varies are the on and off times. So the PWM ratio which controls the speed of a motor is nothing to do with voltages. It is the amount of TIME that the signal is ON compared to the TIME it is OFF. So e.g. on for 1ms, off for 9ms = slow, on for 9ms, off for 1ms =fast.

Steve

Thank you for the clarification. I was thinking that the voltage would be averaged or something...not really sure how a mosfet works on the inside, but this makes more sense that its based on the time the signal is on.

TonyWilk:
Hmm, I couldn't find an actual datasheet for that part, I only found the IRLB4132

See where the On Resistance is at 3.3v - off the top of the graph.
So you might get away with it, but it would be better to use another transistor to switch the MOSFET.

MarkT:
Actually the threshold is 1.8V nominal, not 1.3V, that's the minimum value for it.

It is rated at 4.6 milliohm max at Vgs=4.5V, so its good for 5V drive, the datasheet does not imply it
will work at anything less than 4.5V of gate drive, and in fact thats far too close to the nominal plateau
voltage.

Use part number IRLB4132 to find the datasheet, the IRLR is not a good search term.

I've used this guy from sparkfun to control a small fan from a 5V source using teensy: https://www.sparkfun.com/datasheets/Components/General/RFP30N06LE.pdf

When controlling the fan, it wouldn't come on until I would write (using analogWrite) 100/255, so I think thats part of why I was originally confused about how mosfets are turned on. Perhaps the motor just needs close to 5v to move the fan properly. Also, for this motor at 12V and 11A max (more like 2-4A) I'd like the 140W dissipation capability just in case.

The RFP30N06LE says Vgs is 1-2v where the IRLB4132PBF says 1.35v to 2.35v in the same location on the sheet. Now, I see what you mean on the first page it also says Vgs=4.5v so I don't know what to believe haha.

snowskijunky:
The RFP30N06LE says Vgs is 1-2v where the IRLB4132PBF says 1.35v to 2.35v in the same location on the sheet. Now, I see what you mean on the first page it also says Vgs=4.5v so I don't know what to believe haha.

Many datasheets are a right pain in the a***, the front sheet seems to be written by the sales and marketing men and features like:
Look... really low Vgs !!!
and
Wow... very low ON resistance
are often very misleading because those two things do not happen at the same time.
(i.e. the Vgs threshold voltage does NOT result in the features ON resistance)

For MOSFET parts that are designed to switch several amps, what you really want to know is What is the Vgs which will turn on the MOSFET enough to pass X Amps AND have the stated low ON resistance.

For that you have to dig... In the snippet from the IRLB4132 datasheet I posted earlier, there is that nice graph showing what ON resistance you can expect for Vgs. The datasheet for the RFP30N06LE doesn't seem to have that, it has a graph of Gate to Source voltage Vgs vs. Drain to Source Current and you have to look at the line in the Electrical Specifications to see

``````Drain to Source On Resistance (Note 2) rDS(ON) ID = 30A, VGS = 5V, Figure 9 - - 0.047 Ω
``````

If you are using a teensy at 3.3v I think you will have to switch the MOSFET with another transistor to make sure it's really conducting when you have an 11amp load.

Yours,
TonyWilk

snowskijunky:
The RFP30N06LE says Vgs is 1-2v where the IRLB4132PBF says 1.35v to 2.35v in the same location on the sheet. Now, I see what you mean on the first page it also says Vgs=4.5v so I don't know what to believe haha.

Lets repeat this to be clear - the threshold voltage is utterly irrelevant to driving a MOSFET ON, its
about when its OFF. Below 1.35 the device is OFF, or rather its down to a small leakage current given
in the small print. Between 1.35V and 4.5V the device is in transition from high resistance to low resistance.

The Vgs=4.5V at Rds(on) = 0.0046 ohms is the specification for the device being on. Trust me from
the nominal plateau voltage shown in the graphs that device is not designed to work at less than 4.5V
drive, that really is the limit for this device.

Typically all gate voltages have a large manufacturing spread between devices (several volts in some
cases), so you cannot even rely on testing, you have to look at the min and max specs in the datasheet.
Also gate voltages can drift with device age too.

Ok, I think the graph initially posted makes more sense now. So, basically, I could technically use the IRLB4132 but I wouldn't really be able to fully drive it on which may be an issue at higher loads. Could I use an op-amp (such as this) to amplify my 3.3v signal? I would only need two of them for the h-bridge, I suppose.

The RFP30N06LE seems to behave similarly in that it is not fully on until 4.5v, yet sparkfun makes the claim that it is compatible with 3.3v logic, so it must at least work in some capacity.

snowskijunky:
Ok, I think the graph initially posted makes more sense now. So, basically, I could technically use the IRLB4132 but I wouldn't really be able to fully drive it on which may be an issue at higher loads.

Yes, exactly.

Could I use an op-amp (such as this) to amplify my 3.3v signal? I would only need two of them for the h-bridge, I suppose.

You'd be better off simply using a transistor like:
Diagram:

Outputting a HIGH from the Arduino turns on the transistor Q1 which turns off the MOSFET, a LOW turns off Q1 and R1 pulls the MOSFET gate high to turn it well on.

The extra resistor R2 is there to turn on Q1 and the MOSFET OFF when the Arduino is reset and it's pins are not yet defined as outputs.

The RFP30N06LE seems to behave similarly in that it is not fully on until 4.5v, yet sparkfun makes the claim that it is compatible with 3.3v logic, so it must at least work in some capacity.

Yes, with lower current loads it may be ok - but in this case if 3.3v on the gate leaves the device with an ON resistance of (for the sake of argument) 100 milliohm, then at 11 amps, there would be (V=IR) 1.1volts across the MOSFET which would develop (W=VI) 12.1 watts. So it would get hot, really quick.

Yours,
TonyWilk

TonyWilk:
Yes, with lower current loads it may be ok - but in this case if 3.3v on the gate leaves the device with an ON resistance of (for the sake of argument) 100 milliohm, then at 11 amps, there would be (V=IR) 1.1volts across the MOSFET which would develop (W=VI) 12.1 watts. So it would get hot, really quick.

Probably not particularly great for battery life either if its just wasting power as heat. Thanks again!

Also, I forgot to note in the original post (not sure i this matters) but I have been using this MD10C controller to control the motor, but the main reason I'm trying to move away from that is because it just takes up too much space. Its rather long and has tall capacitors + pwm connector on top that aren't ideal for the space I have. With the mosfets I can easily fit them into the space I have with some perf board.

That particular board has a low threshold of about 0.5v, but it does get to about 100deg F or 38C.

snowskijunky:
Also, I forgot to note in the original post (not sure i this matters) but I have been using this MD10C controller to control the motor, but the main reason I'm trying to move away from that is because it just takes up too much space. Its rather long and has tall capacitors + pwm connector on top that aren't ideal for the space I have. With the mosfets I can easily fit them into the space I have with some perf board.

Ah, I forgot you originally mentioned doing your own H-bridge, pretty tricky making your own driver for an 11A motor - lots of chances for unwanted excitement, like smoke.

You will also need to switch the MOSFETs at the top of the 'H', making sure the top and bottom ones at each side don't turn on at the same time.

The board you mention uses a couple of IR2184 Half-Bridge driver chips, have a look at the Infineon page for the IR2184, scroll down and read the data sheet and all the application notes, manual and articles.

Even with a known-good circuit diagram, it will not be easy making this work on perf board.

Yours,
TonyWilk

P.S. Even experienced engineers, if they were honest, would not be so bold as to only order 4 MOSFETs if they were prototyping something like this

Yes, a good point, always order 10 to 20% spares unless there's a good reason not to - the delay and postage
costs of having to re-order them when you break something and didn't have spares is not fun. Of course if the
part costs more than the postage you might be more parsimonious...

Yes, high power H-bridge mistakes tend to lead to this: Mosfet explosion filmed @ 600fps - YouTube

Wear eye protection?

Another good datasheet to read is the HIP4018A H-bridge driver - this device has a separate charge pump
as well as bootstrapped high side supply, which can be more easier to use.

For a simple life look at some of the higher current Pololu MOSFET drivers using the VHN modules.

Hi,
As you are only controlling a fan, it does not need to reverse.
So you only need one MOSFET to control your fan.

@TonyWilk suggestion will do the job.

As pointed out the transistor between the Arduino and the MOSFET will invert the PWM, so if you have no power on the teensy the fan will run full-speed.

The use of an opto-coupler would fix problem and you would not have to use a logic level MOSFET.

Tom...

MarkT:
For a simple life look at some of the higher current Pololu MOSFET drivers using the VHN modules.

Great suggestion; just looked up VNH5019 Motor Driver Carrier - didn't know about that, looks nice!

Yours,
TonyWilk

TonyWilk:
Great suggestion; just looked up VNH5019 Motor Driver Carrier - didn't know about that, looks nice!

Yours,
TonyWilk

Looks really good as it has current feedback and a stack of protection including OVERCURRENT.
If you biuld your own on perf board and want it small, how are you going to get the perf board to conduct 11A?
What is the application of this fan?
What does it cool or ventilate?
Tom...

TomGeorge:
Looks really good as it has current feedback and a stack of protection including OVERCURRENT.
If you biuld your own on perf board and want it small, how are you going to get the perf board to conduct 11A?
What is the application of this fan?
What does it cool or ventilate?
Tom…

Sorry for just getting back to this, but the fan is a separate thing that I was using a mosfet for. What the H-bridge would be used for is a model train motor, hence the need to switch directions. My apologies for the confusion, I should have explained further.

TonyWilk:
Great suggestion; just looked up VNH5019 Motor Driver Carrier - didn’t know about that, looks nice!

Yours,
TonyWilk

That looks like it might work, thank you for the suggestion. Pololu seems to make good stuff, I’ve ordered some really tiny voltage regulators from them in the past.