H-Bridge help: What transistors/MOSFETS to use for 500ma to 5A

Status
Not open for further replies.
S

Souper man

Guest
I am going to build a H bridge for my next robot, but I dont know what transistors or MOSFETS to use. They need to be able to power a motor that ranges from:

Minium requirements:
100ma no load
800ma stalled

Max requirements

500ma no load
5A stalled

I need some MOSFETS or transistors that can handle the minium requirements. I was thinking about using the 2N2907A and 2N2222A, but when it comes to MOSFETS, im stumped.

Thanks
 
I recommend the IRL540 - cheap & reliable.

What are you using for the high side driver?

Are you planning to use a bootstrap capacitor or are you going separate higher voltage supply or using p-channel MOSFETs?

Did you ever manage to figure out how my circuit works in the other thread?
 
Lol, no i did not manage to figure out how your circuit works in the other thread. I dont know what the high side driver is. I dont know what a bootstrap cap is. The schematic Is shown in the attachment.

It is hard to see, but I am going to make another computer generated schemmy later.
 

Attachments

  • scan0001.jpg
    423.1 KB · Views: 444
You have the symbols of NPN and PNP transistors mixed up.
Your transistors don't have resistors to turn them off and no resistors to limit the base current into their bases.

A Mosfet is different from a junction transistor. The transistor needs 0.7V to 1V with a medium current at its base for it to turn on. A Mosfet needs to have 10V at its input to turn on.
The bootstrap capacitor doubles the supply voltage so that the gate of the Mosfet can be driven 10V above the normal supply voltage for the high-side N-channel Mosfets to turn on.
 

So as not to confuse the audience.

Mosfets don't need 10V to turn on. They usually require between 2-4V to turn on. You can even get 1V max turn on voltage parts. However, for them to meet their RDS(on) specification, that usually does mean a Vgs of 10V. Check the datasheet for your particular device.
 
Optikon said:
Mosfets don't need 10V to turn on. They usually require between 2-4V to turn on.
Click to expand...
Do you gamble with the spec's?
Are you going to spend weeks testing buckets full of Mosfets to find a few that turn on well with only a few volts?
What if a big company ordered and received all the lowest turn-on ones before you ordered yours?
Then all your circuits won't work!

I feed 10V to the gates of my ordinary Mosfets and they all work well.
 

I don't gamble. Mosfet turn on voltages are production tested.
If your design doesnt have a 10V supply available, it would be silly to add one just for doing this job (in some cases) since you can buy them with low turn on voltages. This is what the ever-popular "logic-level" mosfet is.

If you require them to meet their RDS(on) spec, then yes, you usually have to drive them that high (10V). I've never had a single instance of a 2V-4V Vth mosfet not turn on with 4V. (it's guaranteed by MFG)

Turn on voltage and minimum RDS(on) are two different specifications.
 
An ordinary Mosfet is guaranteed to conduct only 250uA at its max Vth (usually 4V)rating. It is its threshold.
If your design uses ordinary mosfets that are guaranteed to conduct well when their gate is at 10V then your circuit should have 10V available.
 
I have never seen a logic-level Mosfet. The datasheet for the IRF3711Z shows that it is good with a gate voltage of 4.5V and it is much better with a gate voltage of 10V. Its max threshold voltage is only 2.45V.
 
if you could referr me to a better H-bridge that closely resembles mine, please let meh know!
 
audioguru said:
I have never seen a logic-level Mosfet. The datasheet for the IRF3711Z shows that it is good with a gate voltage of 4.5V and it is much better with a gate voltage of 10V. Its max threshold voltage is only 2.45V.
Click to expand...

We have to agree to disagree. There is no doubt, the higher you can drive the gate, the better the RDS(on) spec will be. Often, the RDS(on) is specified at a level much lower than 10V (4.5V sometimes) and just as often, that level works out fine in the design. The threshold level is the turn on level. We usually say that a bipolar turn on at ~0.65V or so, and likewise, a Mosfet turns on at ~2-4V.

There are many designs that work with driving mosfets at a 5V (logic) level and they are not gambling with specs.

If you are unfamiliar with logic level fets, have a look:
Logic level is understood to mean that the mosfet has transitioned through its turn on threshold with a 5V VGS. I.e. guaranteed to be on.

**broken link removed**

https://www.vishay.com/docs/72124/si4800bd.pdf

This last one has 30 milli-ohms guaranteed max RDS(on) at only 4.5V

Not too shabby eh? who needs 10V ??
 
There are some Mosfets with a very low threshold voltage. They turn on just by looking at them.
 
I myself am plannig a self navigating on wheels robot. Ill mount or add a microcontroller and at least a fan/filter plus some sort of heat so perhap its all than encased. What it will do is drive around avoiding obstacles but also filter and heat the air for its surroundings. A mobile air filter/heater, for a room lets say or hallway.

Wonder if how Ill power it and where to get the juice, nevermind where/how to build it.
 
Last edited:

Heat the ambient air as it travels? How will you accomplish that? In order to heat the air substantially, you will need to dissipate quite alot of power and for a mobile device that runs on batteries, that spells disaster. You will need a seriously large reservoir not to mention how long those batteries will last.

If you do not use batteries, them why be mobile?

Why not use a stationary space heater?
 
Well unless I have a 120VAC cabl erunning from it and hopefully it doesnt tangle it self.
So that the thing, maybe no heating involved if bateries cant deliver the power.
 
Souper man said:
Lol, no i did not manage to figure out how your circuit works in the other thread.
Click to expand...
What do you want this h-bridge to do?

Do you just want to reverse the motor or do you want full speed control as well?

Using a full h-bridge for speed control is more difficult than a simple low side driver which is what my circuit is.

I dont know what the high side driver is. I dont know what a bootstrap cap is. The schematic Is shown in the attachment.

It is hard to see, but I am going to make another computer generated schemmy later.
Click to expand...
I think you're diving in at the deep end a bit here. I'm not meaning to discourage you if you're confusing bipolar transistors with MOSFETs and don't understand elemental terms like high side diver and bootstrap capacitor. I strondly recommend that you read some tutorials on MOSFETs, bipolar transistors and PWM before continuing any further.
 
Wow, lotta hatred for FET's 'round yonder.

Don't read the columns of the data sheet to find out if a FET is useable in your app, lots of app notes don't tell you a whole lot in the column section. You HAVE to look at the curves. Specifically, Id vs Vds @ a particular Vgs. I've used FET's that turn on HARD by only 3V Vgs. Their on resistance at that point is mere milliohms.

Here's a great example. It's a 20V, 23 amp n-channel MOSFET in a little SOIC8 package:

**broken link removed**

Click on the data sheet and look at Figure 1 on Page 3. With a mere 1.8V Vgs level, you can almost max out the current capability of the FET losing a mere 100-150 millivolts across it. That's about 5 milliohms on resistance, at a 1.8Vgs gate drive level.

Problem is you pay for it. To get performance like this, you generally have to look for FET's with a lower maximum allowable Vds. You're also likely to only find them in smaller packages if not in surface mount only packages. Gate capacitance can be big too.

But IMO, FET's are usually the better device to use in pure switching applications like H-bridges.
 
I must be very difficult soldering to the drain of the tiny mosfet on the bottom of its package. You might need to solder a piece of wire to its bottom.
 
Yeah, they're doing that a lot more these days to heatsink the SMT stuff. You pretty much are forced to reflow it. At work we put on the solder paste by hand, seat the component, and put it in our reflow oven (aka toaster oven). It works good enough for prototypes, never had a failure. We're looking into hot skillets since the folks over at Spark Fun Electronics really seem to like those.

This is also our regular procedure for anything else with bottom pads like QFN's and whatnot.
 
Status
Not open for further replies.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…