Announcement

Collapse
No announcement yet.

OT - PWM controller troubleshooting

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by Dan Dubeau View Post
    Question 1: So would the failure mode of that mosfet be to pass full current? That would logically explain why the broke board still had all function, fwd/reverse, except for the speed control pot.

    Question 2: The mosfet is essentially a relay, right? The IC takes the resistance value from the pot, and sends on/off pulses to the mosfet to pass full current to the motor for various durations to vary the speed.?.

    Is my basic understanding of this circuit close?
    Yes to both cases.
    Mosfets fail 99.9% time to short-circuit unless there is such a high current available that the mosfet package literally explodes. (also pretty easy to recognize)
    My bet is on shorted mosfet: Measure ohms between D-S pins.
    Location: Helsinki, Finland, Europe

    Comment


    • #17
      Originally posted by FinnoUgric View Post

      PWM controller is better than a variable resistor (rheostat) because it does not waste power by heating the control element. If there is a resistor in series with the motor, the resistor will heat up because there will be a voltage across it and there is also a current present.
      So, if the resistor would be, let's say 3 ohms and the motor is taking one amp then the voltage across that resistor would be 3 volts. The power generated in that resistor would then be 3 volts times 1 amp so 3 watts.
      The PWM takes an entirely different approach here. It switches the power to the motor on/off in a rapid fashion. The time the pulses stay on in relation how long they are off determines the speed of the motor. The difference here to the resistor approach is the power generated by the switching element. In a perfect world it is zero. There is either no current or no voltage across the switch so zero times something is zero.

      My primary suspect of the failed unit would be the semiconducrtor used as a switch. When power semiconductors fail, they almost always get shorted. It requires quite a jolt to burn that thing open. That would also be seen outside the semiconductor.

      I also suspect that the new unit will have the same destiny as the now deceased unit.
      I suspected that you would say that.Total power consumption cost for the life of this home yarn winder unit shouln't be more than 39 cents. A rheostat should work forever.So instead of WASTING 30 or 40 watts of power on a rheostat, you are useing a board with about 60 solder joints 15 resisters,2 heat sinks, a capacitor and some kind of IC with about a half of a million switches. By the way most wiper motors and heater motors in cars still use simple resisters. Maybe this is a good learning project and of course I NEVER do anything weird like this. Just ask my family. Edwin Dirnbeck

      Comment


      • #18
        Originally posted by Edwin Dirnbeck View Post
        Why is this better than a simple rheostat? ...
        Also, rheostats are harder to choose: you have to know the maximum current for the motor & calculate the resistance range needed for your speed range. Then there's the rheostat wattage needed. The OP's controller is plug-n-play. Just needs to know the voltage & approximate power. These controllers can be cheaper than a rheostat also.

        Comment


        • #19
          Originally posted by MattiJ View Post

          Yes to both cases.
          Mosfets fail 99.9% time to short-circuit unless there is such a high current available that the mosfet package literally explodes. (also pretty easy to recognize)
          My bet is on shorted mosfet: Measure ohms between D-S pins.
          OK thanks. I'll check it in a bit. I don't have a very good foundational knowledge of electronics like this. Just a bunch of self taught randomness I've picked up over the years, and while I can think through a circuit logically sometimes, other times I'm completely lost and a lot of head scratching happens at what various components do, and how they function. I'd love to take an online course that started with the basics and worked up through a variety of simple circuit designs and troubleshooting basics if anybody knows if such a thing exists.

          I don't remember the specs of this board, but found a similar one online that says it's good for 6a. I don't know the current draw of this motor, but I wouldn't be surprised to see it drawing more than that. I'll try and put a meter on it in a bit*. I think this board is probably under gunned, and this new one will probably fail similarly at some point. The old one lasted about 2 years of fairly heavy use. I just need this one to last another couple hours to get her through this order and I won't have to sleep in the garage .

          *Right now it's working great and, she doesn't want me touching it, or poking around on it right now lol.


          https://imgur.com/AvOd4vu

          Comment


          • #20
            These days, the PWM units are WAAAAAY cheaper, since rheostats at any significant power are rarely used, made in low production, and are high in cost as a result.

            The issue of the rheostat in use is that it has a lousy control characteristic. The PWM acts more-or-less like a constant voltage supply at a given setting. The rheostat is a crummy hybrid of a bad constant current, and a worse constant voltage supply.

            Particularly horrible on series motors, but fairly crappy control response as the load changes with any motor.
            4357 2773 5150 9120 9135 8645 1007 1190 2133 9120 5942

            Keep eye on ball.
            Hashim Khan

            Everything not impossible is compulsory

            "There's no pleasing these serpents"......Lewis Carroll

            Comment


            • #21
              Originally posted by Bob Engelhardt View Post

              Also, rheostats are harder to choose: you have to know the maximum current for the motor & calculate the resistance range needed for your speed range. Then there's the rheostat wattage needed. The OP's controller is plug-n-play. Just needs to know the voltage & approximate power. These controllers can be cheaper than a rheostat also.
              I think this controller was around $5 back when I bought it
              https://www.banggood.com/10V-12V-24V...N&rmmds=search

              Comment


              • #22
                Originally posted by J Tiers View Post

                With the multimeter, set on a resistance range that does NOT measure diodes, if possible.... Measure from each motor pin to power supply negative. On each, measure with the probes forst in one polarity , then the other (reverse positions.

                One position may give a low resistance in both polarities.... If so, the switching device is probably bad.

                The switching device is the one on the larger heatsink. The barrel shaped device behind it seems to be the freewheeling diode.

                There are three connections on the MOSFET. As you look at it, they would be "source" (negative) on the right, "drain" (output to motor) in the middle, and "Gate" control input) on the left.

                The measurement you want is from "drain" to "source". looks like that is from negative to the second pin from the right on the 4 pin connector.
                Thank you Jerry, I'll measure those things later tonight if I get a chance.

                Comment


                • #23
                  Originally posted by Dan Dubeau View Post
                  Question 1: So would the failure mode of that mosfet be to pass full current? That would logically explain why the broke board still had all function, fwd/reverse, except for the speed control pot.

                  Question 2: The mosfet is essentially a relay, right? The IC takes the resistance value from the pot, and sends on/off pulses to the mosfet to pass full current to the motor for various durations to vary the speed.?.

                  Is my basic understanding of this circuit close?
                  The Mosfet is just another semiconductor switch device.
                  The basic premise is it is a current switching device controlled by a voltage on the gate, essentially known as a transconductance amplifier similar to the old tube/valve devices of yore.
                  As opposed to the bi-polar transistor which uses a base Current to control a collector Current.
                  IOW the Mosfet places a minimal demand on the device that is operating it.
                  Coming from someone who came up through the valve era, I really took to the Mosfet! 😎

                  Comment


                  • #24
                    Originally posted by Edwin Dirnbeck View Post
                    Why is this better than a simple rheostat? I am not an electrician. Edwin Dirnbeck
                    Along with efficiency that was noted above, the motor will "see" a lower power supply impedance, the result is higher torque at lower speeds.


                    Comment


                    • #25
                      Originally posted by Edwin Dirnbeck View Post
                      Why is this better than a simple rheostat? I am not an electrician. Edwin Dirnbeck
                      PWMing a motor is superior to a rheostat for the same reason that it's better to run a regulator rather than cut a hole in your airline to reduce the pressure coming out of your compressor. Modulate the power rather than just burn off the excess.
                      -paul

                      Comment


                      • #26
                        Originally posted by Edwin Dirnbeck View Post

                        I suspected that you would say that.Total power consumption cost for the life of this home yarn winder unit shouln't be more than 39 cents.<snip>
                        Actually it is not the cost of the electricity. It is the heat itself that must be dissipated somewhere.

                        There has been some talk about the control characteristics of a PWM controller when a DC (Series DC) motor is used as the load. I can tell that there is no noticeable difference between those two. Both behave likewise badly.

                        One of the projects I have done was a acid pump to deliver formic acid for preservation of animal feed. A PWM controller was used to drive the pump motor whose speed was dependent on the amount of incoming feed. There was also a flow sensor to measure the actual flow of formic acid.
                        The problem was that the motor had no torque at lower speeds. First there were no rotation and then suddenly, approximately at 70%, the motor decided to deliver almost all it was able to. This behaviour was finally solved by using a PID (Proportional, Integral, Derivative) controller. After this it worked as designed.

                        Comment


                        • #27
                          Originally posted by FinnoUgric View Post
                          This behaviour was finally solved by using a PID (Proportional, Integral, Derivative) controller. After this it worked as designed.
                          I don't see that being a problem in this application as the operation is manually controlled by the operator (wife) and 'feedback' would be in the form of foot pedal and visual observation!

                          Comment


                          • #28
                            MOSFETs are pretty much generic devices and for an application like this you can't go too far wrong using one with much higher current and voltage ratings. Here are five pieces of IRF540N (33A 100V) for $5 with free shipping from US:

                            https://www.ebay.com/itm/30325364009...0AAOSwNnhhs8h7
                            http://pauleschoen.com/pix/PM08_P76_P54.png
                            Paul , P S Technology, Inc. and MrTibbs
                            USA Maryland 21030

                            Comment


                            • #29
                              "An application like this" means a low speed PWM, with no attempt to do anything fancy.

                              When you get into say, VFDs, then it starts to get very important exactly what MOSFET (or more usually IGBT) is used. The wrong ones will heat up and fail, even though they may be rated for much more current than is needed. (Actually, it is partly BECAUSE they are so high-rated for current).

                              For these, there is some benefit to using one with suitable specs, because it can make a 2 or 3 to one difference in heating. A poor choice of device may have "killed" the failed unit..... Electrical and electronic parts nearly always fail due to overheating. Most types of fault end up causing too much heat.
                              4357 2773 5150 9120 9135 8645 1007 1190 2133 9120 5942

                              Keep eye on ball.
                              Hashim Khan

                              Everything not impossible is compulsory

                              "There's no pleasing these serpents"......Lewis Carroll

                              Comment


                              • #30
                                The main problem with using MOSFETs that are overspecified for the purpose is that gate capacitance can cause slow switching with a soft source, like the LM324 op-amp in this case. If the op-amp has an effective output resistance of 100 ohms, and the IRF540N MOSFET has 1220 pF gate capacitance, the time constant is about 120 nSec. During the switching time from on to off, there will be a condition where as much as half the supply voltage and half the current will exist simultaneously, so if the device is switching 10 amps and 12 volts, there may be about 30 watts dissipated, but for no longer than about 50 nSec. If the device is switching at 50 kHz, or 10 uSec between transitions, the duty cycle would be 50/10000 or 0.5%, and switching losses would be about 30 * 0.005 or 150 mW. If the ON resistance is 0.04 ohms, as it is for the IRF540N, the conduction losses will be as much as 100 * 0.04 = 4 watts at 100% duty cycle. The actual performance is a bit more complex, but this is the general idea.
                                http://pauleschoen.com/pix/PM08_P76_P54.png
                                Paul , P S Technology, Inc. and MrTibbs
                                USA Maryland 21030

                                Comment

                                Working...
                                X