Announcement

Collapse
No announcement yet.

A motor controller is a power supply, right?

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

    A motor controller is a power supply, right?

    Hoping someone here will be able to provide a few pointers with this. I've got a motor controller board for a device that moves a caravan. A DC motor on each wheel is supplied with 12v DC or -12V DC depending on whether the remote control is pushed in "forward" or "backward" mode. The control is arranged by having 2 relays per motor that, at rest, are all connected to the negative.. If the motor is to go forwards, then relay 1 activates meaning the motor sees "positive negative" on its terminals 1 and 2. If the motor is to go backwards then relay 2 activates meaning the motor sees "negative positive" on 1 and 2

    So, that part seemed ok.. The relays are fine, but there was no negative to the motors even tho the positive was OK


    Turned out there were a parallel pair of MOSFETs on the negative side of things for each motor and all the MOSFETs were blown/gave crazy readings. I swapped them out like for like and one channel now works fine, the other channel has an issue:

    It works fine if I use my multimeter to supply a bit of voltage to the MOSFET gates, but the board itself doesn't supply it. Having caught a comment on the web that "blown MOSFETs can cause the driver circuits to be damaged too" I'm checking the circuit that drives them. Because the board is more or less symmetrical in terms of components I traced back from the gates on each side to see where voltage differed, comparing working vs broken, and where it was the same. I'm no electronics genius, but it's proving to be an interesting learn

    So I've photographed the two relevant halves of the board. You can see one of the FETs in the top right (NMOS6) and that's the broken side - so the thin track that leads under each cap is the gate line:



    On the left side the gates are at 12v (red overlay) when the board is powered, on the right they're at 0v (blue overlay). Working vertically down the picture there are a pair of transistors - one's a J3Y and the other is really hard to ID - seems to have 3K written on it on the left, seems to have a hole in it on the right. I believe (if I'm understanding the datasheet schematic correctly) that this means the gates are connected to the emitters and on the working side the base pins are at 0v (blue), on the broken side the base pins are at 0.8v (orange). Everything else seems consistent each side of the board, though the layout differs the circuit seems to be the same. I stopped tracing back when I reached the same 5v (purple) or 12v (everythere else) so the majot difference seems to be this track that links all 3 transistors in some way - 0.8v on the broken side and 0v on the working

    I drew this:

    which I think is right, though the circuit simulator doesn't seem to indicate any 12v should reach the gates so it's probably incomplete, or I've got the BCEs the wrong way round, or something else to do with the transistors...

    Any assist on that I should be doing to repair this?

    Replacing the exploded transistor is first up, but can I raid any transistors out of a computer PSU i have kicking round? Or should I just be ordering exact replacements.. and what is the exact replacement for the one with the hole in?

    Will the J3Ys need replacing too or might the voltages I see be explained by damage on only the one with the hole in?

    Am I right in thinking these transistors are J3Y (easy to see) and a 3K (?) (It's a BC858B?) The transistor with the hole in has "3(hole)2" and the 2 is rotated 90 degrees. The "same" one on the working side of the board has "3KDO" written on it, except the DO is 90 degrees rotated and looks more like |)() because the letters have a break in the font at the top and bottom..

    It's a bit nasty too, but if I don't get any joy from fixing this side, sould severing the gates line and bridging it to the gate line on other side of the board also work?

    Thanks in advance for any pointers!
    Last edited by cjard; 10-14-2020, 01:41 PM.

    #2
    Re: A motor controller is a power supply, right?

    You need to show good clear straight shots of the whole top and bottom side of the board, and also how all the boards and other parts are connected to the boards.
    Never stop learning
    Basic LCD TV and Monitor troubleshooting guides.
    http://www.badcaps.net/forum/showthr...956#post305956

    Voltage Regulator (LDO) testing:
    http://www.badcaps.net/forum/showthr...999#post300999

    Inverter testing using old CFL:
    http://www.badcaps.net/forum/showthr...er+testing+cfl

    Tear down pictures : Hit the ">" Show Albums and stories" on the left side
    http://s807.photobucket.com/user/budm/library/

    TV Factory reset codes listing:
    http://www.badcaps.net/forum/showthread.php?t=24809

    Comment


      #3
      Re: A motor controller is a power supply, right?

      As requested:




      Note these images have the MOSFETs removed (except one known good as a test). On the base of the board most of what you see is either negative or positive. This board connects to nothing else except 12v DC In on the middle bolt pair (and 12v DC out on the outer bolt pairs depending on motor direction). The red fly leads go to some devices that are some kind of thermal cutout I believe- glued to the lids of the lower two relays (which are constant energized when power applied). The bit of paper wedged in the relay can be ignored too
      Last edited by cjard; 10-14-2020, 05:09 PM.

      Comment


        #4
        Re: A motor controller is a power supply, right?

        Is the voltages the way you measured on the board? Seems the circuit input is active low and when power needs to be supplied that input +5V needs to go to 0V.

        As colored up and voltaged, with that first npn transistor connected to +5V/high, the mosfet gates should be off and the mosfets themselves should be off. This would imply the upper left NPN transistor in the left side of the picture is shorted and that needs to be replaced.

        Comment


          #5
          Re: A motor controller is a power supply, right?

          hi eccerr0r

          Yep, the coloring in reflects the voltages I measured on-board when the board is powered up

          It's the right side of the colored-in pic that is faulty, the left side (with 0v between the transistors) works ok/ supplies 12v to the gates of the IRF3205s when the board is powered up..

          Comment


            #6
            Re: A motor controller is a power supply, right?

            So your "simulation" schematic is wrong because as drawn, the gate drive is supposed to be 0V and the mosfets off. You'll need to retrace again.

            In any case that transistor seems to have failed if indeed is an npn transistor, having the base at 0V, emitter and collector at +12V, something is very wrong with circuit biasing and that device is now shorted due to reverse breakdown.

            Comment


              #7
              Re: A motor controller is a power supply, right?

              Not quite sure I fully understood what was said there, but here's a bit more info (that schematic is what I see when I don't ask the board to do anything i.e. its what I see with it powered up but not activating any motors):

              These FETs (IRF3205 NPN are present on either side of the board and they seem to be used to turn the negative on and off to the motor - I say this because there is a large bolt to which the user should connect his battery, labelled N, and this bolt is clamped to a big rail that runs round the board. All the FET source pins connect to this rail, the drains connect to a pair of relays (that also receive a positive 12v from elsewhere) that are switched alternately to make the motor go backwards or forwards accordingly.

              On the side of the board that works, the FET gates receive 12v as soon as the board is powered on. On the side of the board that doesn't work, the gate line is at 0v.

              The relays all energize as they should to carry out the switching but the motor only turns on the side where the FET gates are at 12v. I've noticed, however, if I use my multimeter in diode mode to supply some volts to the gate line on the broken side of the board (with the neg probe on the source rail), then all the motor channels work as expected; seemingly the multimeter has activated these FETs in place of whatever is supposed to do it on the board proper. If I short the gate and drain then the motor becomes inoperative again

              I've checked and double checked different datasheets for these transistors, and they all seem to say that (looking top down with the one leg on the left and two legs on the right) the left leg is collector, top emitter and bottom base, and I've looked at that schematic about 20 times checking it - i'm sure it's right unless the components I have have a different BCE arrangement.. The voltages I've put in the schematic are those as measured with the board powered and no remote buttons pushed.

              When I push a remote button and activate the relays, then for either side (working or broken), the 5v falls gradually to 0v over about 5 seconds.
              On the working side the 0v between the transistors gradually rises to 3.6v.
              On the broken side, the 0.8v between transistors gradually falls to about 0.5v
              On the working side the gate line drops from 12v to 11.5v but I figured this was due to a weedy power supply running the relay, I'm not sure. I've only got a crappy grade multimeter..


              From previous queries/asking around it was suspected the MOSFETs might be used as a soft start system, but I don't know how to verify it
              Last edited by cjard; 10-15-2020, 11:19 AM.

              Comment


                #8
                Re: A motor controller is a power supply, right?

                In a properly working circuit the node that's common to all three bjt transistors (collector of first, bases of other two) should drop to near 0V when the "+5V" input to the first transistor and resistor is at +5V.

                The common to all three bjt transistors should reach +12V when that "+5V" input goes to 0V.

                The output of the two "final" bjt transistors should mirror what the "common to all three bjt transistors" node. So the side where it says "ningbo bainian" matches what I think is what should be proper operation of the transistors. The other side, if the circuit is measured properly, appears to have a faulty J3Y transistor. However this appears to be contrary to what you say is true so

                - your schematic is wrong (also: Where is that diode connected?)
                - your voltage measurements are wrong
                - the voltages measured in the first post are not steady state and in the 'on' condition and measured at the wrong time.
                Last edited by eccerr0r; 10-15-2020, 12:43 PM.

                Comment


                  #9
                  Re: A motor controller is a power supply, right?

                  I don't know where the diode connection goes - it's "the point beyond which the voltages are the same and I stopped measuring"

                  The schematic is "wrong" in the sense that the colors don't match what I measure - I didn't draw the colors; falstad's circuit simulator does that for me and I can't make the sim match the voltages I measure (dont know how). If I toggle the 5v between 0 and 5v I can see that the output of the NPN-PNP pair whose emitters are linked switches between 0v and 12v when the 5v goes to 0v (respectively)

                  Ignoring the schematic voltages for a moment, and just looking at the components, I'm sure that the schema I've drawn reflects the components placed on the board, but if you think i've made an error in terms of connecting the components, let me know..

                  The voltages measured and stated in the first post are "board powered on, no remote buttons pushed" - the voltages are steady, but not representative of what happens when the remote button is pushed.

                  I can't explain what I measure on the "non NINGBO" side of the board; I know you're saying it's wrong, but that's the side that "works" in that it turns the motor.. I'll make a video of the volt checks soon. I'm not sure if you're saying I'm wrong/have measured incorrectly/am blind, or that the circuit is not functioning correctly per the schematic design? Perhaps if these fets are supposed to provide a soft start I should see some pulsing (not sure my meter is fast enough), or at least that the gate line should have some sort of switching behavior and shouldnt immediately go to 12v when the board is powered on?

                  The NINGBO side of the board had a "3K" (I think it means BC858) PNP transistor with a hole blown in the top - i've taken it off the board and get a reading of 240 between base and collector on my meter (leads any way round), nothing for anything else.. I understand that for a PNP I should have had a reading of about 700 emitter(pos lead) to base(neg lead) and same ish for moving pos lead to collector, on the multimeter I have.. So I think it's a fair bet that transistor is a goner

                  Also for the ningbo side, that gate line is at 0v and stays that way.. So perhaps both sides of the board are faulty but for different reasons; the non-NINGBO side activates the MOSFETs permanently when it should switch them and the NINGBO side never activates them

                  I've found an A733 PNP transistor in an old power supply; is it a compatbile replacement for the BC858 with the hole in the top? Specs seem broadly similar but I've no idea what I'm reading..
                  Last edited by cjard; 10-15-2020, 03:06 PM.

                  Comment


                    #10
                    Re: A motor controller is a power supply, right?

                    A733 is TO92 but for your specific case it should work fine.

                    Sorry I can't help anymore, the data you give doesn't correlate to each other, too many either board track extract errors or measurement errors in the data you collected.

                    Comment


                      #11
                      Re: A motor controller is a power supply, right?

                      Yes, it's a different package but I can rearrange the legs so it perches in the right place

                      So you'd say there's a possibility that some of the J3Y are damaged too? This could explain some things; I've been following "how to test your transistor" guides and I just couldn't get them to match up to how the components behaved. I thought it was just because I was testing them in-circuit (which I understand can't always work out because other components present on the circuit will confuse the readings) but I couldn't get the readings to match from one side of the board to the other either

                      I think it's probably better to just pull all the transistors out of both sides and bench test them all..

                      Another couple of questions for you if you will: I've also managed to raid a PC power supply for a few SC8050 transistors. Will these work adequately equivalently to any J3Y that I find defective?

                      Is there any way to tell from the readings I have given which J3Y is damaged/which I should pull and test first ( on the non-NINGBO side; the size that puts its mosfet gate line straight to 12v?)
                      Last edited by cjard; 10-15-2020, 11:36 PM.

                      Comment


                        #12
                        Re: A motor controller is a power supply, right?

                        I was able to resolve this, hopefully; turned out every transistor in the schematic above, was toasted bar one J3Y.. Schematic was right, but the volts I had measured were completely out of whack because of the number of damaged components

                        Not having enough time to wait for the exact replacements to arrive I raided 3 PC power supplies and was able to cobble together a pair of SC8050 and an A733 for one side, and a pair of 2222As and an A1015 for the other; reshaping the legs suitably so they perch where the SMDs used to be

                        Now have MOSFET gate lines that steadily rise in voltage the longer the button is held down and same for the actual motor output lines. The tech department for the device were able to confirm that the MOSFETS implement a soft start

                        Thanks to eccerr0r for the pointers on what to check and replace!
                        Attached Files
                        Last edited by cjard; 10-16-2020, 07:07 PM.

                        Comment

                        Working...
                        X