Re: DIY speed control of ~1hp PM DC motor?

Grant, I realize you're trying to put one together yourself, but here's what I just got in the mail:
http://www.surpluscenter.com/item.asp?UID=2006060718553670&item=11-2269&catname=electric
Pretty nice unit for the money.

I'm using it to replace the driver board that used to be in my just bought Bridgeport's x feed motor. I thought I was going to put it together last weekend; this was before I realized that not only was the original pc board gone, so were the microswitches that the direction lever actuated. Maybe this weekend.

Steve

Grant Erwin wrote:

Back in 1996, an article appeared in the magazine "Home Shop Machinist". The article contained a simple circuit for speed controlling a small permanent magnet DC motor: http://www.tinyisland.com/images/temp/HSM-DC.jpg

People took exception to the design, and an Intel guru took pity on us and set it up in a circuit simulator and, with the benefit of his experience, he recommended some circuit changes. I pulled these off of google:

"I have finally had a chance to examine the PM-DC motor speed control article that appeared in the May/June 1996 and the January/February 1997 issues of HSM (recall that someone complained about the HSM speed control not working well - my attempts to contact the original poster have been completely unsuccessful).

The speed control is in fact a closed-loop system (early information that I received suggested otherwise but the topology is not typical for this type of speed control). The following deficiencies and recommendations are based on experience and circuit simulations using a particular Spice circuit simulator.

1. Rheostat R1 (a potentiometer wired as a variable resistor) should have some resistance in series with it. Without significant resistance the SCR could be destroyed by excessive gate current if power were applied when the line voltage was at peak value and the speed control happened to be at the maximum possible motor speed (minimum resistance). Under these particular conditions the motor, diode D2, diode D3 (a diac) and the SCR gate are in series across the peak valued full-wave rectified line voltage. I'd use a 470 ohm (2 watt) resistor since that is large enough to limit gate current to a reasonable value and small enough to have little influence on the motor speed control range.

2. The speed control potentiometer is too large of a value. The value is listed as 250K ohms and this combined with the 200 nF timing capacitor results in too large of a time constant at maximum resistance (minimum motor speed). An excessively large time constant will cause the motor to skip-cycle at minimum motor speed because it won't allow the timing capacitor to reach the firing voltage of the diac (28 volts) in 8.33 mS (1/120 Hz). So it takes more than one cycle of input between each SCR firing causing the SCR to fire every other cycle (or even less often).

The optimal value for the potentiometer is 150K ohms but don't use that potentiometer value as a fix for the problem (further explained below).

3. The speed control is quite non-linear due to the fact that the timing capacitor is being charged with a 120 Hz full-wave source. This means that the speed control is too sensitive near low motor speeds (at high R1 resistance) and not sensitive enough near high motor speeds (at low R1 resistance). Both this problem and problem #2 above are best solved simultaneously by placing a 360K resistor in parallel with resistor R1 (the speed control rheostat). The effect that this has is to limit the maximum value of the speed control resistance to about 148K ohms and to make the speed control less sensitive at low speeds. The end result is a near perfect linearity and no more motor skip-cycling at low speeds.

4. There should be a pull-down resistor on the gate of the SCR to prevent self- triggering from power-line noise and/or high ambient temperatures (which causes anode-gate leakage current increases). A typical value to use for this is 100 ohms from the gate to the cathode of the SCR.

5. The current ratings of the diode bridge and the SCR are 10 amps but the motor being speed controlled is rated at 1.5 HP. If 1.5 HP represents the maximum output power of the motor (motors are frequently rated this way) then the RMS current would be about 20 amps (all PM-DC motors are close to 50% efficient at maximum output power). If the maximum output of the 1.5 HP motor is larger than 1.5 HP then the current drain may be considerably lower than 20 amps at rated load due to higher efficienc but it will still exceed 10 amps RMS (at 80% efficiency the current is about 12 amps). I would use at least 20 amp devices for 1.5 HP motors.

6. The HSM author makes some very questionable recommendations with regard to trouble-shooting (increasing the timing capacitor value) and remote control operation (routing high power AC through a remote control cable instead of using a local power switching relay)."

OK. That's the background information. Now I have a question for the electronics guys among us. What is the function of diodes D1 and D2 in this circuit? Also, if one were to insert a 470 ohm 2 watt resistor as suggested, exactly where should it be located?

Also, if I were going to implement this circuit using a stud-mount SCR, using a 4" electrical box, how should I implement a heat sink? Replace the steel box lid with aluminum, and use that as a heat sink? Hate to have wires connected to the box lid ..

Finally, there is an alternate circuit out there in the public domain. It is Figure AN1003.16 in the app note on the littelfuse.com Web site:
http://www.littelfuse.com/data/en/Application_Notes/AN1003.pdf

Does anyone have any comments on why one circuit is better than the other?

Thanks,

Grant Erwin
.. just scrounged another dirt cheap electric lawn mower motor ..
.