A Pulse Width Modulated Speed Controlled DC Lead Screw Drive for a Small Lathe
By Evan Williams
I have a South Bend 9C lathe. It has change gears and no quick change gearbox. The only time I find this to be a hinderance is when I am doing a finishing cut. I usually have the change gears set up for some common thread like 18 tpi or so. This is much too coarse for a finishing cut and it is a pain to change the gears just to make a single cut.
So, I designed an electric lead screw drive. I found a nice little 24vdc worm gear motor in the junk box. I drilled and tapped the end of the lead screw 5/16 x 18 tpi and attached a cog belt pulley.
The motor and belt can be disengaged in about two seconds by loosening the bottom left bolt to remove belt tension.
To drive the motor I designed and built a pulse width modulation DC motor controller. This is a true PWM drive as the pulse frequency does not vary with pulse width. I chose a fairly low frequency around 250hz because this motor turned out to be rather loud at higher frequencies and not suitable for ultrasonic frequencies. Low frequency also simplifies the design.
The drive will operate the motor to turn the lead screw under very considerable load from 20 rpm down to 1 1/2 rpm. Operation is smooth and very stable. The drive unit is quite overbuilt and could drive a much larger motor. The motor runs quietly although there is a low hum at low speed. No phonographing is visible in a cut on aluminum.
It has reverse capability and is also equipped with an input for an auto stop switch. I recycled an old computer power supply case to put it in.
Technical details :
The drive uses an 18 volt 4 amp transformer which when full wave rectified and filtered provides 27 volts no load and 24 vdc under load. The drive motor is a 24 vdc motor.
Input and output of the drive is fused.
Pulse clock is generated by a 555 timer. This is used to trigger a 555 configured as a monostable time delay.
The time delay is adjustable. This creates the pulse width adjustable drive for a pair of IRF720 power HEXFETS. The IRF 720s are rated at 3.3 amps each with a Vds max of 400 volts and max dissipation of 50 watts each.
I could have used just one FET as the motor only draws about 2 amps at stall but I like having a big safety margin and besides, it is trivial to parallel HEXFETS. In theory this drive could operate a treadmill motor if you ganged up a dozen or so of these HEXFETS and increased the drive voltage and heat sink.
The timers run on their own 12 vdc regulator.
Circuit protection is provided by several items:
The output of the pulse generator is isolated with a diode so that in the event of failure of the gate structure in a FET it cannot feed back 24 volts to the timers.
A 20 amp schottky diode is used to clamp the inductive motor load in the reverse direction.
Waveform shaping is used to control the rise and fall time of the gate drive to about 0.1 ms.
The FET gates use 100 ohm series resistors to provide a time constant with the gate capacitance that suppresses hf parasitic oscillations.
The HEXFETS have an RC snubber across the drain/source to clamp inductive ringing.
Here is the schematic for the drive:
Download the full resolution schematic here
Parts List for Pulse Width Modulated 24 VDC Motor Controller
Unless otherwise stated resistors are 1/4 watt, 10%
Unless otherwise stated capacitors are 50 volt (or more) rated
S1, S2 SPST switch
T1 Transformer, 117 VAC Primary, 18 VAC Secondary, 4 amp output (size for application)
DB1 Full wave diode bridge rectifier, (size for application)
F1, F2 Fuses in holders, (size for application)
LED1 Light emitting diode for pilot light
C1 6800 mfd, 35 vdc This part is not critical. Any reasonable substitute will do.
R1 4.7k,1/2 watt resistor (bleeder resistor)
U1 LM7812 or similar 3 terminal voltage regulator. See note 5.
U2, U3 LM555C timer IC (see figure 1)
C3 100uf, 35 vdc electrolytic capacitor
C4, C6 0.47 uf mylar or polyester capacitor. See note 4.
C5, C7 0.01 uf
C8 0.1 uf mylar or polyester capacitor
R7 100 ohm
R8 10k linear potentiometer
D3 Diode, 1N4004 - 1N4007
Q1, Q2 IRF720B N channel B-FET, 3.3 amps, 50 watts, 400 Vds (figure 2). See note 3.
R3, R4 100 ohm gate resistor for FET. See note 1.
R3 1k, 1/2 watt
C2 0.22uf, 400 volts mylar or polyester
D2 Schottky diode, see note 2.
S3 DPDT switch
The gate resistors should be mounted as close as possible to the gates of the FET power transistors. Snubber resistor R5 and C2 are also shown. See photo below.
The schottky diode is recommended to be rated at least 5 amps forward current and a reverse voltage rating of 40 vdc. An excellent source for this part is a dead computer power supply. They usually have a diode pair that can be paralleled to function as a single diode. Any such diode pair found in a computer power supply should work well. See figure below.
There are many suitable FETs that can be used in place of the IRF720B. A good replacement is the IRF511. It has a slightly higher current rating but the voltage rating is much less at 60 volts Vds.
IMPORTANT NOTE: The FET mounting tabs are at drain potential. Either the devices must be mounted with insulating spacers and washers or the entire heat sink must be insulated from the chassis.
C4 and C6 may be changed together to change the operating frequency of the drive. If they are reduced in value the frequency goes up. A reduction to 0.047uf will result in an operating frequency around 2000 hertz. Reducing to .001 will result in a frequency of about 10khz. If this is done then R7 should be increased to about 470 ohms to maintain the pulse width of the trigger pulse. Also, R11 may be decreased to 470 ohms as well as C8 decreased to 0.01 uf.
The drive may be operated on 12 volts for a 12 volt motor. If this is the case then U1 may be eliminated. C3 should still be used and should be placed close to U2 and U3.
Once the circuit is constructed the operation should be tested on a resistive load. This can be a light bulb or two to suit the voltage of the drive. The most important part of testing the drive is to ensure that when set to maximum power the width of the pulse from U3 is just slightly less than the time between trigger pulses from U2. If the pulse width of U3 exceeds the time between trigger pulses from U2 at some point when the potentiometer is advanced to maximum speed the power output will suddenly fall. This can be seen as a sudden drop in brightness in the test light bulbs or in motor speed.
To make adjustment easier a trim pot may be substituted for R6. A suitable value would be 50K. If this is done a resistor of 1000 ohms should be used in series with the trim pot so that resistance cannot be reduced to zero.
The autostop input may be left unused. If you wish to use it then any source of a switch closure will cause the drive to stop. This input has 12 volts at a few milliamps present and is not static sensitive or noise sensitive. Any switch that can short the autostop input will stop the drive.
S3, the reversing switch, should not be operated unless the drive is turned off at S2 or by the autostop.
Dead computer power supplies are a good source of parts for a project such as this. In particular, heat sinks and diodes as well as mounting hardware are easily obtained. Just go to your nearest computer repair shop and ask for a few dead power supplies, I am sure they will be happy to give you some.
Here is an example of what is often available in a dead supply. These parts will not usually be damaged. Especially useful is the dual schottky diode in this picture. It is rated at 2x15 amps at 100 volts DC.
555 timer IC
IRF720B FET pinout