P-channel, 60 amp mosfet?

[ClydeCrashKop]

I am trying to drive 0.2 ohm motor coils with 12 volts AC and no more than 10 amps. Ohms law says that would be 60 amps. (But not for very long) I am making the AC under computer control and just swapping polarity. That part is working fine except that I need a logic level component on the positive side of the 12 volts. I tried the logic level IRL3103 mosfets that are good for 64 amps and 30 volts. With a resistor in line to keep the amps down, they worked fine on the ground side but the ones on the positive side blew right away. I know they weren’t getting +5 v above their ground, oscillations also.
A P-channel, logic level, 60 amp mosfet would be perfect but there doesn’t seem to be any such animal in captivity. Is there another device that I can use? Maybe an SCR? Can I switch a voltage regulator on and off?
Also, if I send a 10% duty cycle, at a higher frequency than swapping polarity into this coil, will it be the same as 6 amps instead of 60?
I am a digital kind of guy and trial and error can get expensive.
Mosfets don’t have much magic smoke in them.

 

[dknguyen]

PMOSs are half as efficient as NMOSs so at those power levels with the power dissipation involved, it's really hard to find a PMOS because it would be so huge and expensive.

Try using an NMOS and a gate driver IC. It will allow you to switch a high-side MOSFET with an MCU signal.

 

[Oznog]

Ohms law says that would be 60 amps. (But not for very long) I am making the AC under computer control and just swapping polarity. That part is working fine except that I need a logic level component on the positive side of the 12 volts.

Well you don't need to use Ohm's law here. You need to have enough current drive at Vgs=5v, not the thermal rating of the transistor, to ensure you can start the motor and with a healthy margin.

MOSFETs can be paralleled very effectively if need be.

Please clarify with a schematic. I am not even clear how you have a 0 to 5v control signal and are switching the low side of 12v with an NMOS and also on the high side of 12v with a PMOS. For example if you had a PIC sharing 12v ground, the PMOS is going to see Vgs of -7v to -12v which won't turn it off.

 

[Hero999]

It depends on the inductance of the motor coils, the DC resistance may only be 0.2ohms but at AC the impedance it likely to be a lot higher.

How big is this motor?

Does it have a power rating?

 

[ClydeCrashKop]

I am driving this right from LPT1 data bits with 1K resistors for protection. The gate driver IC sounds perfect. Can you recommend one?

I was hoping AC the impedance might work in my favor but I am trying to run it like a stepper motor so turning slow, I think it would be resistive again. Would a pulsed current of the same polarity keep AC the impedance going?

I started to make a schematic but being a Virgo, it might take a while.

 

[dknguyen]

Search them up. Maybe one of these. You will need to know the gate voltage threshold required to switch the MOSFET. Make sure to search and read application note on Gate Driver ICs.

Just for the high-side NMOS (low-side NMOS driven directly by MCU pin)
**broken link removed**

Each IC drives one high-side NMOS and PMOS
**broken link removed**

 

[ClydeCrashKop]

Thanks dknguyen
I am sure one of those will work even if I get mosfets to fit their drivers.

Any ideas about keeping the amps below 10 with PWM? I have never tried this before.

 

[dknguyen]

Use a current sensing resistor and modify the PWM accordingly with code. If you decide to use a high-side resistor rather than low-side you need a current sense IC to produce a ground-referenced voltage or current to measure with the MCU's ADC.

When you say AC, do you mean:
1. Simply you are not driving the motor with a direct DC voltage but with a unipolar PWM signal
2. You are actually driving the motor with an AC voltage (doesn't seem to make a lot of sense to me since you are using MOSFET switching, but you do say AC and you do say swapping polarity).

And when you say swapping polarity, do you actually mean you just want to control the direction that current flows into the motor to control the motor's direction (like with an H-bridge?)

I am not entirely sure what you are trying to do.

 

[Nigel Goodwin]

I was hoping AC the impedance might work in my favor but I am trying to run it like a stepper motor so turning slow, I think it would be resistive again. Would a pulsed current of the same polarity keep AC the impedance going?

Trying to run a motor that slow is fraught with problems, most sizeable motors include fans to keep them cool - at slow speeds the fan doesn't cool it enough and the motor overheats.

You also give no indication of the type of motor?, you mention AC and DC?, but this confuses more than helps.

If you're wanting stepper type performance?, why not use a stepper?.

 

[ClydeCrashKop]

I will have to research the current sensing resistor and IC and H-bridge. The object of this game is to create a fractional horse power stepper motor from a small 3 phase motor. I want to control machinery with it. The steppers that I have been playing with are too slow and have no balls. Anything bigger is cost prohibitive. I think the best way to describe it is like a 3 phase Bi-polar (tri-polar?) stepper motor. I did actually have it running for a minute or two but with a resistor in line, it was slow and weak. The scope pattern looked like a good design for a water slide, an oscillating down staircase. The mosfets blew when I gave it more amps but still under 10 amps. Even if it doesn’t perform as well as I hope, I am at least amusing myself.

 

[Nigel Goodwin]

I will have to research the current sensing resistor and IC and H-bridge. The object of this game is to create a fractional horse power stepper motor from a small 3 phase motor. I want to control machinery with it. The steppers that I have been playing with are too slow and have no balls. Anything bigger is cost prohibitive. I think the best way to describe it is like a 3 phase Bi-polar (tri-polar?) stepper motor. I did actually have it running for a minute or two but with a resistor in line, it was slow and weak. The scope pattern looked like a good design for a water slide, an oscillating down staircase. The mosfets blew when I gave it more amps but still under 10 amps. Even if it doesn’t perform as well as I hope, I am at least amusing myself.

The whole point of a stepper is that it doesn't actually 'rotate', it moves in small discrete steps - by stepping fast enough you can get the illusion of rotation.

On the other hand a motor rotates, and doesn't 'step' - are you hoping for steps from it?, they will certainly be VERY coarse ones assuming you can make it work.

Generally if you can't get something to work it may be because you're going about it the wrong way? - what EXACTLY are you trying to do?.

 

[ClydeCrashKop]

I do want incremental steps in both directions. I want to control the travel on a big lathe and a milling machine. Gear reduction can make big steps into small ones. Printer motors need gear reduction to move anything heavy which makes small steps into tiny ones and no speed.
On my first attempt with resistance in line, it was working with only about 2.5 volts and 5 amps getting to the motor. With dknguyen’s help using a high side driver, I can get 12 volts and ?? amps.
When Edison was working on the light bulb, someone said he must be getting discouraged because he wasn’t making any progress. Edison said he’d been making great progress, because now he knew twenty thousand materials that you couldn’t use for a filament.

 

[Nigel Goodwin]

I do want incremental steps in both directions. I want to control the travel on a big lathe and a milling machine. Gear reduction can make big steps into small ones. Printer motors need gear reduction to move anything heavy which makes small steps into tiny ones and no speed.

Now you've explained what you're doing it becomes far more clear!.

H-Bridge - DC motor - feedback encoder - gear box.

Or feedback encoder after the gear box.

No stepper required, and no chance of a missed step!.

 

[dknguyen]

Ah, I see. Okay, so it's that you are using AC in the signal processing sense that you are not driving the motors with a purely DC signal. You are pulsing a PWM from a DC source. Yeah, everything I said before can still apply.

EXCEPT you will have to IC-sense chip to convert the voltage read across the current resistor into a ground referenced voltage (in ALL cases regardless of whether you go low or high side current resistor) because the way your circuit is set up, the same current resistor will alternate between high-side and low-side depending on which MOSFETs are on. Also, your ADC will be destroyed if you connect it directly to measure the resistor voltage because it can't handle the negative voltage across the current resistor because the current can travel through the resistor in both directions (assuming you place the resistors in all three T-intersection, rather than than at each ground terminal...for symmetry sakes.

If you place it at each ground terminal, then the above doesn't matter.

About the milling machine, I think they all use steppers because the milling motors move really REALLY slow and steppers move best at slow speeds. To use a DC-servo motor, it would take a serious amount of gearing to move motors of that size at that speed, or you would need ginormous motors (and still have gearing, but less).

 

[ClydeCrashKop]

I will have to study that. Maybe I can put the current resistor on the main supply. It would only see pulsating DC.

 

[dknguyen]

I will have to study that. Maybe I can put the current resistor on the main supply. It would only see pulsating DC.

Haha, oh yeah. It never occured to me from staring at the diagram that the power would only enter and exit through two terminals of the battery in a single direction...hehe.

 

[ClydeCrashKop]

Is that current sensing resistor similar to a remote ammeter that uses a shunt on the mains?
You have no idea how much I have learned while researching all of the help you guys have given me today. I found a lot of good info on all of the above subjects at http://www.irf.com/technical-info/appnotes.htm including this one as a future project: AN-1045: AC TIG Welding: Output Inverter Design Basics PDF
Thanks a lot

 

[dknguyen]

Is that current sensing resistor similar to a remote ammeter that uses a shunt on the mains?
You have no idea how much I have learned while researching all of the help you guys have given me today. I found a lot of good info on all of the above subjects at http://www.irf.com/technical-info/appnotes.htm including this one as a future project: AN-1045: AC TIG Welding: Output Inverter Design Basics PDF
Thanks a lot

I think so (I am soley drawing on the definitions you are using, I believe I have seen others refer to a current sense resistor as a shunt before.). It's just a very low resistance resistor so that a small voltage drop forms across it depending on the current flowing through it. You can measure this voltage to figure out the current flowing.