TIP142 Stepper motor driver

[musicworld1]

Now, I have try this circuit with 24V DC, 1.7Amp power supply. But still the motor does take any step (forward or backward). I think the value of resister is very much big. What it should be?

[vger]

I see you have posted the same question in many forums. First I must ask if there is a return path to logic ground for the base/emitter current. Second, are your stepper motors really 24V steppers?

Steve

[nhatson.elec]

you can try linistep

[Torchhead]

One thing missing from the TIP142 circuit is recirculating diodes but that is not the reason for no motion. The 4.7K resistors are fine. It should work with any resistor value from 2.2K to 10K. It's operating in the saturated mode. The current gain on the darlington is doing to be 100 or more so if you put 10ma in at the base you can saturate up to 100 times that (1000 ma) on the collector. To test this connect a collector through a 20 ohm 25W resistor and use your scope to observe the wave forms as you send pulses. You should have nice sharp square waves that swing between 24 volts and a couple of volts (low). The problem with darlingtons is they don't turn off nearly as fast as a conventional bipolar transistor (with the proper Base driver circuitry) or a FET. Also make sure your driver is both sourcing and sinking current at the Base connections. You will have to current limit through the +24 to each coil or the transistors will just burn up. You have to select a value that keeps the current through the motor coil at or below the motor specs based on 24V. It needs to be a big large (size/wattage) resistor based on: Power dissipation = I^2 * R (current squared times the resistor value). What you find is about 50% of the power goes up in heat.

Driving large inductive loads (like motors) with transistors is wrought with "gotchas". The power dissipation across the transistors will be higher for Darlingtons because the Vsat of the transistor can be higher. Couple that with slow trun on or turn off and the dissipation quickly can triple.

If you are interested in learning transistor theory and this is a hobby (electronics) for you then continue to go forward. If your goal is to build a machine that moves and cuts something then building your own motor drives from scratch is a false sense of economy.

The Linistepper schematic is interesting. It is basically a stepped voltage regulator (with current limiting) so that it puts a quasi sine wave across the coil. Another way to look at it is it's (similar to) a class D linear amp. This does several things for you.

1. Sine waves don't have nasty harmonics and the need for rapid turn off and turn on times that square waves require. Sine waves through a coil don't produce nasty inductive kickback that turning on and off instantly does. Inductors/transformers/coils "like" sine waves and fight square waves. Semiconductors "like" nice sharp ON/OFF (square) waves and "fight" sine waves.

2. Lowers the parts count for a given design. You don't need the snubbers and kickback diodes.

3. The current limit is provided by the two 1 ohm 5W resistors (equiv of .5 ohm 10W). The total current at the motor is more RMS than peak.

Now for the down side:

Lower efficiency and higher heat dissipation. The voltage drop is performed across the transistor in a linear form so it's no longer just the Vsat voltage times the current but the total voltage drop at the given part of the wave form times the current. So the heat you would lose in the saturated design with a big power current limit on the +12V side is largely moved to the transistor device.

Steppers get RPM from voltage. The power dissipation goes up (a lot) with higher voltages. Running at 48 volts with a 3A motor means lots of heat. Instead of something the size of an ice cube you end up with a driver the size of a cigar box.


It's still a unipolar drive design that is inherently less efficent than a bipolar approach.

You need bigger transistors and bigger heatsinks for them. It's a valid design for smaller motors with lower current needs. It can quickly balloon out of hand when the current needs are several amps. If you count the cost of heatsinks, fans, etc the price differential tends to narrow versus a chopper style drive.

This is not a knock on the design. It's just to point out there are pro's and cons to every power driver approach.

What would be an interesting variation of this would be to measure the emitter current via analog inputs and use a PWM drive to the base that varies the current waveform in a sine wave function.....wait that sounds like a microstepping chopper drive (:-)

TOM caudle
www.CandCNC.com

[RomanLini]

Originally Posted by Torchhead ... The Linistepper schematic is interesting. It is basically a stepped voltage regulator (with current limiting) so that it puts a quasi sine wave across the coil. ...

A minor correction; what is a regulated voltage (as you said) at the base of the power transistor causes a regulated current through the transistor and motor, due to the fact that the Vbe of a darlington remains quite constant regardless of collector current. The overall effect is of controlling current through the motor (regardless of PSU voltage) so it more correctly should be referred to as a "current regulator".

Originally Posted by Torchhead ... What would be an interesting variation of this would be to measure the emitter current via analog inputs and use a PWM drive to the base that varies the current waveform in a sine wave function. ...

Some of my earliest designs were closed-loop using opamps to measure the emitter resistor current and set Vb accordingly to regulate current. I discovered this big increase in complexity was not necessary due to the relatively fixed Vbe of the darlingtons (as I said above) so the total design was simplified to be released as a minimal parts open-source design in the form above.

You are very right about the problem of heat dissipation, the Linistepper is very good with 1 amp stepper motors but is quite unsuitable for microstepping with 3 amp motors etc. There are some considerable benefits in smoothenss because of the "sine" like DC current fed to the motor, as C5 and C6 provide a gentle ramp in current change as each microstep moves to the next microstep. That makes it quite a smooth, quiet, high performance driver for smaller motors moving at low-medium speeds like with small PCB engraving machines especially when in 18th microstep mode (3600 steps/rotation).

[musicworld1]

Originally Posted by Torchhead If your goal is to build a machine that moves and cuts something then building your own motor drives from scratch is a false sense of economy.

Yes, you are right, I am planning to make a CNC machine. I want to select such a good driver that could run about every motor in a good way. So I select TIP142 that can handle 100 VDC & 10 Amp. I am not expert and you can suggest me an excellent driver schematic.

[Torchhead]

[QUOTE]I am not expert and you can suggest me an excellent driver schematic...{/QUOTE]

I don't think "excellent" or even "good" can be used in the same sentence as TIP142 and driver. The ratings can be deceiving. You can do 100V and you can do 10A but not both for very long. You cannot exceed the total power dissipation of the device. The Vce Sat for the Darlington is 2 to 3 volts. The device will dissipate 150W but only if you can keep the junction temp at 25 deg C! For every deg C above that you have to derate the amount of power it can handle. Turn on/off times require rapid rising and falling base drive signals. Base turn off/storage times are in the range of 5 times the turn on time so they will dissipate more current turning off.

FET's are better suited to drive motors. There are plenty of FETS rated at 100Volts and more than 10A. (IRF530 is common one) They are easier to drive and have lower dissipation at voltages below 50 volts.

An 'excellent" design is an H-Bridge so you can run the motor BiPolar. Turning on the high side device rapidly can be a challenge (especially for a conventional transistor)

Then you have to worry about current limiting. You can't just turn on the motor with a transistor (or FET) with the other end of the coil directly connected to + voltage unless you either limit the current with a power resistor or you do it electronically and limit the width of the turn-n pulse. 10A at 100V is 1000Watts. Even at 50V it's 500W.

Power dissipation is the total voltage drop across the device times the current times the duty cycle.

TOM caudle
www.CandCNC.com