12 volt / 10 amp PWM from a micro controller?

Go To Last Post
12 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi guys,

I'm going to need some help on this one. I have a voltage 12v source that can source more than the 10 amps I need (a PC power supply)... and I have an AVR micro controller...

The only thing I'm aware of that would enable this is if I used a relay, or other magnetic/mechanical switching mechanism. But I think that's out of the question because it would be clicking constantly, and super noisey.

How would one accomplish something like this? What type of parts would they use? Would an opto-isolator work (or am I smoking crack?)? Will I need heat sinks/cooling?

My goals here are safety, low cost, and low noise.

Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What are you going to control with this PWM?

A relay is not well suited for PWM. Usually,they are far too slow. You MAY be able to do it with transistors or FETs, depending on the kind of load.

Opto-isolators tend to be relatively slow, though there are some faster ones out there. They might be a big help in keeping the big currents out of your micro (if you use the right opto).

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm attempting to control the heater block for a 3D printer. I don't think the speed needs to be great. I'm just planning on measuring a thermistor and adjusting the duty cycle. I think the 2-10 Hz range would be fine.

Ultimately, I want to get pretty fine control on the temperature (hoping to expose temp control to software), but since heat is slow moving... again, I think low speed is OK. (?)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If it is a 12V resistive heater, a single MOSFET with heat sink should work. One connection of the heater would be fed by +12V, the MOSFET will sink the other connection to ground. A logic level MOSFET can be driven by an AVR pin with a 220 Ohm resistor in series with the gate. This is called a low side driver. Research that. The grounds would be common. The PWM is a way to vary the duty-cycle. It is likely the PWM frequency will need to be higher, to reduce granularity in the control.

A simple proportional, integral control with a temperature sensor will aid in regulation. This is done with AVR software and likely ADC for reading temperature sensor. There are other ways too.

Not sure a PC supply will work, may need more load on the 5V, and a minimum load on the 12V, to get the supply to regulate properly.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would think that a low frequency PWM signal would work fine. The thermal time constant is orders of magnitude longer than even a slow PWM frequency, and at the lower frequency the mosfet will spend less time in transition, thereby generating less heat. I think, perhaps.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Much depends on the heater wattage, thermal inertia, load and regulation requirements. The system may have a high load since material is being added, and possibly at a variable rate.
Data logging of temperature, control output, and other parameters are helpful in the development process.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Since it is simply a heater, you can control either high side or low side. Low side is much easier to drive directly from a micro.

Low switching rate should lead to cooler operation. But, you do need to choose a device that can be saturated hard, with the drive you have. Some FETs have a threshold of 6-8V gate-source, but take 10-12V to turn them on, hard. You would NOT want to use such a device with 5V system.

Earlier, you mentioned opto-isolator. While its a very good idea to keep power out of your control system, opto-isolators do not have a very strong drive capability, in general. This tends to suggest that you really want some sort of gate driver on the isolated side. It can be powered from your 12V power, but you are very likely to need something. Then, you can happily use an opto-isolator.

Personally, I would use a switching frequency in the 10-100Hz range. You can get there, dead easy, with counters and prescalers with almost any reasonable CPU clock frequency.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Also pay close attention to the Rds(on) of the FET you choose at whatever gate drive voltage you have available. Since the power dissipated is I²R, where I is current through the transistor and R is Rds(on), power dissipation can get quite high even with fairly low Rds(on).

Using a higher drive voltage as Jim suggests will help to reduce Rds(on), even with a "logic level" FET. There are a number of FET driver ICs out there that make it quite easy to amplify a logic level signal to the voltage and current required to drive a FET efficiently. (The more current you can drive into the gate during switching, the faster it will saturate, which reduces I²R losses incurred while turning on and off. Note that getting that drive current in practice generally means an extra, low ESR cap right at the driver.)

Using a higher drive voltage also gives you a much larger selection of FETs to choose from, and you may be able to pick a lower Rds(on) part than if you were limited to parts with a logic-compatible gate threshold.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A big fet and a heatsink.

The largest known prime number: 282589933-1

It's easy to stop breaking the 10th commandment! Break the 8th instead. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

To add to the point about Rds(on), suppose that it is 0.1 ohms. Thats a little high for a well-driven power FET, but there are lots out there in this range.

Power is R *Square of I. So, 10A squared is 100. Times 0.1 ohms makes 10W. That is pretty substantial. Common running lights on a car are 2.5W. Those get pretty hot when they have been on. Thus, you would need real care to get that heat removed. If the on resistance were 0.01 ohms (they are available), then you have "only" 1W but that is still more than what a circuit board, by itself, would probably handle.

One mitigating factor is that the power is dissipated only when the switch is on. So, if it is running at 50% duty cycle, then the 10W becomes 5W. That said, you need to design for 100% duty cycle because it will be that when doing a cold start.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Another thing you might think about is putting two FETs in parallel. The resulting total power dissipated in the FETs is reduced by half and each individual FET only dissipates 1/4 of what you would see using a single FET.

Greg

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd use a non logic level super low rds on fet and since you have a 12v rail I'd drive it with that.