High current from microcontroller

[Electronics4you]

Hi there,

I need some inputs on how to supply a high current from a microcontroller. The setup is a high current darlington transistor (Ic, max=150A, beta=100) and a microcontroller to control it. With the transistor gain I need to supply 1.5 A. Can this be done using either darlington driver arrays or power transiors/MOSFETs without using small-Ohm high power resistors that eats up all the power?

Thanks

 

[cowboybob]

Hi there,

I need some inputs on how to supply a high current from a microcontroller. The setup is a high current darlington transistor (Ic, max=150A, beta=100) and a microcontroller to control it. With the transistor gain I need to supply 1.5 A. Can this be done using either darlington driver arrays or power transiors/MOSFETs...

You sort of answer your question with the question.

I assume you mean to control a high current event with a signal from the uc. That being the case (since the uc can't possibly supply that current level), then driving a Darlington pair, etc., is a logical choice.

... without using small-Ohm high power resistors that eats up all the power?

Thanks

Not sure what your asking concerning the resistor.

 

[Electronics4you]

What I'm really asking is what to use and how. I'm preferring the MOSFET/BJT approach, but with e.g. an N-BJT, you need a high-end high-power resistor to the collector of the BJT in order to supply the high current to the base of the power darlington. That I want to avoid - if possible

 

[KeepItSimpleStupid]

(Ic, max=150A, beta=100)

Are you off an order of magnitude? 15A not 150 A? Darlingtons can have beta's on the order of thousands.

MOSFETS are preferable for a variety of reasons. Voltage is used to turn them on, but they have other sorts of issues. Logic level MOSFETS were developed to be turned on with TTL logic.

Depening on what kind of load your trying to turn on, there might be other ideas or recommended practices.

 

[Electronics4you]

I want to use power darlingtons to drive to load, because the ones I have in stock can handle the needed 150 A, but unfortunately the gain is only 100, so I need the large base current. My question is therefore if I can use either a MOSFET or a BJT to create that large base current while controlling it logically from a microcontroller? The load of the circuit is inductive.

 

[Nigel Goodwin]

I want to use power darlingtons to drive to load, because the ones I have in stock can handle the needed 150 A, but unfortunately the gain is only 100, so I need the large base current. My question is therefore if I can use either a MOSFET or a BJT to create that large base current while controlling it logically from a microcontroller? The load of the circuit is inductive.

Of course you can - just use a driver transistor to feed the output one - resistors required depend on your exact circuitry, and on what you're trying to do.

 

[Electronics4you]

@Nigel: It just that, if the base of the power darlington requires 1.5 A the collector resistor power dissipation will be ~4W, so I just thought that there would be a smarter solution

 

[Electronics4you]

Circuit

The circuit could be like the one attached

 

[MikeMl]

When switching 150A of inductive load from a PIC, you are crazy if you think that it is ok to share a common ground between the PIC and the power circuit.
I would be looking at an opto-isolator between the pic port and the input of your Darlington booster, with galvanic isolation between the grounds.

Just search these forums for all the "PIC unexpectedly resetting when I switch my load" threads. There are dozens of them....

 

[Electronics4you]

@MikeMl: I'm using an optocoupler, but didn't include it in the schematic - nice spotted.

But will the low ohm collector resistor really do the job here?

 

[Nigel Goodwin]

The circuit could be like the one attached

Use a PNP driver transistor instead, but the heat has to go somewhere (perhaps better with an FET?) - but as Electronic4you has mentioned, it's fraught with problems - entirely separate supplies and an opto-isolator (or a relay) would be a better way to go.

 

[Brevor]

Maybe several P channel Fets in parellel to switch the load, that way you dont need all that base current. Along with an optically isolated N channel Fet to pull down all the gates. Do you really have 5 Volts at 150 Amps ?

 

[ChrisP58]

What is the voltage to the load?

What is the load? Inductive, resistive, ?

What is the switching frequency?

What is the typical duty ratio of the switching?

 

[MikeMl]

What is the voltage to the load?

...

To which I add:

What voltage is the PIC running on? 5V?

What voltage is the 150A power supply?

 

[Electronics4you]

I'm not going to run 150 Amps continuously through the circuit. The high current is simply a possibility, if the load current (in this case a motor) peaks for a shorter period of time. This could be the case at start-up of the motor, or if the motor is temporarily blocked. I'm going to run in the 20 A range at 18 V thus staying below the 700 W range of the component.

@ChrisP58: I haven't measured the load inductance yet, so I cannot determine the optimal switching frequency at the moment. The duty cycle will be very small (not considering that the signal may be inverted), since I need a low angular velocity. I think the no-load velocity of the motor is ~4000 rpm.

@MikeMl: The PIC circuit is 5V, and all interfacing between the control and the motor side will be optically isolated.

@Nigel Goodwin: I'll consider switching to P-channel MOSFETs to control the darlington base in order to avoid power resistors. Instead, the MOSFET can be mounted with a suitable heat sink.

The new circuit attached shows the idea

 

[KeepItSimpleStupid]

Do youknow about these devices: https://www.pwrx.com/summary/IGBT/IGBT.aspx?gclid=CMfxg9SvorYCFcxlOgod_WQATg IGBJT's.

 

[Electronics4you]

Yes, I have worked with them before, and I would probably use them instead of the darlingstons in another situation. For now though, I already have a bunch of the high power darlingtons, so I'll try to use them first. I know they can do the job, since they used to drive the motors of a 6-DOF industrial robotic manipulator.