On the information page for a MOSFET I'm looking at (http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_1561408_-1) it lists the gate voltage at +/- 20. Does this mean the gate voltage will need to be [VOLTAGE TO MOTOR] + 20?

EDIT: The info you linked to listed a motor driver for high voltages. In the driver's PDF they list a circuit on the top of the 2nd page http://www.intersil.com/content/dam/Intersil/documents/fn36/fn3659.pdf. Would this circuit work on a high voltage system without PWM (I noticed it only used NMOS- no PMOS, and the driver is only rated for a few Amps (but it doesn't look like the big current goes through the driver)).

 

You need to look at the IRF8010 Datasheet linked by the product for full information.

RDSON is 15mΩ @ 40 Amps with a VGS of 10V

To use, you'll need to supply the gate with 10V, preferably 15V for faster switching. This is the purpose in life of Gate Driver ICs that have charge pumps.

You'll want to note the turn on and turn off times in the datasheet to prevent shoot-through (mentioned in paper I linked earlier). With a 10V VGS, minimum pulse time is 10mS, with 15V, pulse time is reduced to 1mS.

See page 6 of the datasheet for thermal characteristics. A heatsink will be needed for 30 Amps due to resistive mode between off and on during switching, which is dissipated as heat..

There is some math involved, which is explained in the paper I linked.

 

On the information page for a MOSFET I'm looking at (http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_1561408_-1) it lists the gate voltage at +/- 20. Does this mean the gate voltage will need to be [VOLTAGE TO MOTOR] + 20?

EDIT: The info you linked to listed a motor driver for high voltages. In the driver's PDF they list a circuit on the top of the 2nd page http://www.intersil.com/content/dam/Intersil/documents/fn36/fn3659.pdf. Would this circuit work on a high voltage system without PWM (I noticed it only used NMOS- no PMOS, and the driver is only rated for a few Amps (but it doesn't look like the big current goes through the driver)).

That would be a Gate Driver IC that your microcontroller "talks to", it, in turn, will drive the 4 IRF8010 MOSFETs with a high enough VGS.

N Channel MOSFETS are preferred in this application as they typically have a lower RDSON than P-Channel MOSFETS.

 

If you only need 15V to switch (asuming it's not 18+15), then couldn't you use a smaller transistor to step it down to 5V?

EDIT:
As long as that schematic should work I'll give it a try. It's within my budget, and easy enough for me to understand at my current level.

 

Which schematic?

Before ordering anything, make sure you aren't adding apples to oranges!

Post the complete schematic of what you intend to put together here first. People have bought a mish-mash of parts, only to find out they wouldn't work together and wished they asked first.

Please post a schematic as a .PNG file here that has labeled values and part numbers so others can review it. I am not always correct, you know...

 

Here's the schematic http://switchit001.com/motor4.png

 

So I posted to Yahoo Answers, and a few people said it may not work without PWM, which I don't have enough experience to go into at this point :/

I also showed them the 10A tutorial and a few people suggested using the same system, but putting a few extra MOSFETS in a series. (http://answers.yahoo.com/question/i...V0yPNtXsy6IX;_ylv=3?qid=20130215075713AANsl78) Do you think a system like that would work?

EDIT: As far as the heat sink goes I'm thinking of just buying a 10ft copper pipe and attaching all the mosfets to it. Would that dissipate enough heat? Also would it work better if I filled it up with water?

 

Did you read the PDF I linked several posts back?

I don't think you'll need that amount of thermal mass, thermal dissipation is what you require, which involves large surface areas, commonly realized as fins in metal.

That PDF has a proper schematic in it.

 

Yes, it seemed like the concept in the PDF was very similar to the concept behind the H-Bridge in the tutorial. The only conflict I see in the PDF is the MOSFETS require 12V over the through current, however I read that you can open the MOSFET enough with only about 20V at the gate so I'm not sure which method is correct.

Also in another tutorial I read it looks like they're turning on a 12V MOSFET with only 5V. http://www.seas.upenn.edu/~ese206/labs/MOSFET/MOSFETHBridge.pdf (Page 3), but I may be missing an important concept

 

There are many types of MOSFETs, the ones that can be switched at 5V are in the "Logic Level" family. They may not have the current ability you'll need.

When making a full H-Bridge with N-Channel MOSFETs, the high side gates need to be a full VGS (3-20V, depending on MOSFET) above the power supply voltage to conduct, this is the purpose of the charge pump, to get, say, 28V to the gate of the high side N-Channel MOSFETs to switch them.

If you look in the IRF8010 datasheet I posted above, there is a rating for RDSON, look at the VGS and IDRAIN for that RDSON measurement. if rated at the current and voltage you need to switch at, and the resistance is very small, you have one of many appropriate choices.

To get more specific, you need the full specifications of your motor, inductance, peak (startup/stall) current, free running current, min/max voltage, etc. Then you can put your motor literally somewhere on each plot in the datasheet to be sure you are in the safe operating area.

It's harder to make a "General Purpose" 30A H-Bridge than it is to make one for 2 Amps. Too many things can vary. That is part of the reason there are so many MOSFETs to choose from. You may pick an N-Channel and P-Channel Logic Level set to drive from the Arduino, at the loss of switching speed and more heat from higher RDSON.

Generally, design for 50% over your anticipated max voltage and 25% to 50% over your anticipated max current. That will allow different motors with similar characteristics to be used with the same circuit. If you design too far over, you waste money in components, as well as finding performance at the low end of the the MOSFETs, rather than upper middle of safe operating area.

 

Ah, that makes more sense now.

I didn't want to get into PWM since it increases the chance of me making a mistake and shorting my battery, but it looks like I'll have to. I found another schematic using the driver listed in the PDF that lays everything out in a very readable format (http://www.hvlabs.com/Images/dcmotorcontroller.jpg). I'll try reading over it tonight to see if I can get a good understanding of how each section relates to the explanations in the PDF.

 

So I just studied the circuit, and I understand most of it, but there are a few questions I have

1. Does the sub-circuit in the top left corner act as a charge pump?
2. I see he has the charge pump going into the high side current inputs (pin 2&7). A shoot through won't be a problem since the low side pins override the high side pins, but would the voltage coming out of the charge pump be over the 15V limit?
3. Out of pins 20, 18, 13, and 11 he has a loop with a diode and resistor before the MOSFET. Is this used to decrease oscillation/protect the MOSFET?
4. I don't fully understand pins 12 & 19, but they look like they are used to return the oscillation to the IC
5. Why do are pins 1,10, and 15 all used to drive the high side gates? Do those just need a larger current than the low side

Sorry for asking so many questions, but I would like to actually know why a schematic works instead of just putting parts together and thinking they magically work.

 

The diodes that are "backwards" (cathode to positive) are for spike supression, inductors try to keep the current flowing through them at a steady rate. When an external current source is removed, the magnetic field around the inductor collapses, creating a high voltage in attempting to continue current flow. That is very simplified, but the general idea.

Motors are made of windings that are essentially inductors/have inductive properties, so when the power is switched off, either when turning off the motor, or rapidly in PWM, double and triple the supply voltage spikes will be seen at the motor terminals, these voltages exceed the breakdown voltage of the MOSFET. To prevent destroying the MOSFET, the external diode is a fast switching diode which shunts the current back to the supply, bypassing the MOSFET. These diodes should be across all 4 diodes, not just the low side, though the low side will see a higher voltage spike. The internal "body diode" formed as a byproduct of how MOSFETs are made isn't fast enough to absorb the spikes, so schottky or other fast diodes are used.


As for the driver, they work by using capacitance as a charge pump. Charge two capacitors in parallel to 18V each, then switch them in series and there is 36V output. Charging/switching fast enough gives a 36V output from an 18V input (or different ratios). They are a form of switching power supply.

The low side MOSFETs can be switched on with the supply voltage, which is often Vgs in higher voltage circuits. There needs to be an additional transistor to switch the 5V signal from the microcontroller to be the 18V of the supply, but no charge pump is required.

The upper left hand circuit is a 7812, a linear voltage regulator for a steady supply to the IC. This allows the IC to be ddesigned for 12V operation inside, with outputs rated for higher voltages. Trying to make an IC capable of being fast and capable of high voltages is contradictory, as a large potential difference on the IC die will cause arcing/shorts inside the IC itself.

From another thread here, IR2110 Driver Datasheet has an internal block diagram and more details.


Does that help clear things up?

 

Thanks, I have a much better understanding of the circuit now. Also how would you calculate the surge the diodes would need to protect against? Would it be higher than the current, or smaller than it?

 

The current will be about the same as the motor has running, since the inductors want to keep that current flowing. The voltage builds up quickly, which is why you want a fast diode to be the first thing to "dump" the current back to the source, rather than allowing the voltage to reach a level that can destroy components.

 

So If I used this diode would I have to parallel 2 to get to a safe level of 50Amps


Also is this the correct calculation to find the watts of the resistor?
P=V^2/R

so for a 10,000 Ohm resistor with 18 V going through it...
P=18^2/10000

 

For the flyback diodes, the peak current is the rating to look at, not average/RMS current.

Vishay SB260 Example Datasheet 2A Avg, 60A Peak, max reverse voltage 60V.

50 cents ea. @ Digikey

 

Ok, so I just got a few parts in, and have made some slight modifications to the schematic to fit the system I will be using.
http://switchit001.com/h-bridge_new.png

I have outlined the "dangerous" lines in red. I will keep these off the breadboard, and use 16 gauge wires. I have changed the voltage source to a single 18V battery, and changed the voltage regulator portion so it is similar to how the actual part looks. After I get the voltage regulator set up I will try testing it at a lower voltage to make sure everything works, then add onto the system from there.

 

Make sure you have heatsinks on your MOSFETS, old CPU or VGA heat sinks work well for that if you drill some holes, you can mount them to one large one if you use the insulating washers and a mylar gasket with heatsink grease.

The Tab is often connected to source or drain, so if they are all connected to a common heatsink without insulation, the battery would short out through them and they'd go poof (and the battery might explode if it is lithium and doesn't have overload protection).

Note on that schematic states to put a heatsink on each MOSFET and have active cooling (a fan) moving air over them. You'll decide that once it's running, usually better to start out too cool than too hot.

 

I have a few old xBox heatsinks so those will probably work well while testing the H-Bridge.

I tested the voltage regulator and it worked perfectly . I'm now working on the MOSFET section of the H-Bridge, and I've run into a few problems. One of the top mosfets seems to be broken (it lets current through if the gate is attached to negative or positive), I may have let it heat up too much while soldering and luckily I bought a few extra MOSFETs so I'll replace it.

I've also noticed the P6KE68CA diodes seem to be letting 2.5 out of 3V through while testing, will this lead to a shoot through in the live system?

EDIT: After further testing the P6KE68CAs alone seem to let virtualy no current through, I may have the diodes parallel to them flipped.
EDIT2: The flipped diodes were the problem, now I just need to replace the broken MOSFET and make sure the overall system works
EDIT3: Ok, the board works with 3 volts then 18volts, near 0 amps each test. I am now going to try to connect the H-Bridge to the controller and do another test with near 0 amps