Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

DC12V 12A car window motor - what spec for a 220V PSU?

  1. Jul 10, 2017 #1
    I have a combined motor and screw drive (to 90 deg).
    It's original purpose was to convert car windows to electric drive... my guess is that it is at least 25 years old, and made in Japan.
    For scale... the width of the motor casing is 50mm (as you look at it)


    I'm making a motorised turntable.

    The jig is coming along nicely, but I need to start thinking of a PSU and could do with some assistance.

    I don't have a PSU but I've tested the motor momentarily on an 8A battery charger.
    It rotates at a reasonable speed ie. around 120 rpm.
    (I attached an indicator wire to the output shaft and counted 60 turns to 30 seconds)
    This is in the ball park area (by amazing good fortune)

    It probably has more potential power than I need... primarily I'm looking to rotate up to a motorcycle barrel.
    Clearly it would be nice to control the speed.... perhaps a bit faster, or a bit slower (more likely).
    Maybe I need to have adjustable voltage.... perhaps that would be useful for bringing the turntable up to speed.

    I have a very nice Rollei PSU - AC 12V max - 100VA - click adjustable 110V, 120V, 130V, 220V, 230V, 240V
    I don't need to use it, but I note this for information purposes.

    Having looked on eBay... the transformers seem to be highlighted for LED strip, and CCTV usage.
    Typically 12V 15A 180W.

    None are controllable by 'twist knob' - but I've not gone into it in depth.
    I don't mind adding controllers.... I can solder.

    However, I thought it best to get an overview from the forum, on what direction I should be looking at.

  2. jcsd
  3. Jul 10, 2017 #2


    User Avatar

    Staff: Mentor

    What's a motorcycle barrel?
    You need to use PWM of the 12V source to control the DC motor speed. :smile:
  4. Jul 10, 2017 #3
    Thank you very much for that lead :)

    Re: "What's a motorcycle barrel"?
    It is the metal casting that contains the cylinder (for the piston).
    The outer part of the barrel is either heat radiating 'fins' (air cooled) or it holds a jacket of coolant (liquid cooled).

    For the particular barrel - it is in the 1.7Kg range.
    I'd be happy to get a result at that, and then learn from the experience.

    With all the friction involved with winding a window... I'm thinking that the motor can start a 2kg load.
    ... but maybe a soft start is always better.

    I researched PWM... this stands for 'Pulse Width Modulation'
    Here's a very good video introduction to PWM:
    (I presume it's good... he has 1.5m subscribers, and in electronics... it might be a worthy reference)

    But from his video, it seems to indicate that I should be choosing a 24V source.... with the median output being 12V.
    Have I understood that correctly (or is he talking bollocks)?

    From this, I'm thinking that I need a mains voltage transformer to 24V DC

    P = I x V
    144W = 12A x 12V

    Does this mean that the motor requires 144W regardless of the voltage?

    It is not yet clear to me, the spec of the transformer.
  5. Jul 10, 2017 #4

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    My advice is find a secondhand car battery charger at a yard sale or junkpile .
    I just returned from my scrap metal recycle yard where there's a pile of about twenty of them .
    Most often it's just a broken cord or meter or thermal overload that's gone wrong. The transformer and rectifier are usually salvageable.

    I suspect the D12A is a model number not its current rating. 144 watts is almost 1/5 horsepower which should lift 110 pounds 1 foot per second.

    It might draw that much if stalled though. You'll have to test it to see.

    DC motors draw current in direct proportion to their load. Unloaded it'll probably draw only an amp or two is my guess.

    Be aware that some window motors have a rubber safety clutch inside that round gear . Its purpose is to limit how much torque the motor can apply to the window lift mechanism so as to not cut off a limb when kids are playing with the electric windows. It shears and you have to replace it.
    Yours may not, some simply use a lower torque motor.
    With how long a lever arm, ie how much torque ?

    Windshield wiper motors are also worm gear drive and will handle continuous duty.

    Fun project. Look for hobbyist PWM kits they're not expensive.
  6. Jul 10, 2017 #5
    No. It means the amount of current can be as high as the rated 12 amps (provided it is 12 amps; see Jim's post #4). At 12 volts, and 12 amps, electrical input to the motor is 144 watts (12V x 12A =144 watts).

    How much current flows through the motor depends on mechanical load at the output shaft. As mechanical load increases, so does the amount of torque (current) required to turn it.

    For example, perhaps once you've loaded it, the motor demands 6 amps at 12 volts (72 watts). 72 watts of power requires 6 amps at full speed (12V). If the speed-to-power relationship is linear (possibly the case here, but not so with fans and pumps, and a variety of other loads) you can estimate how slow you can go. 6 volts (half speed) is required to create 72 watts at the 12 amp rating limit.
  7. Jul 11, 2017 #6
    Thanks for the contributions.

    It's hard to bear... the thought that it's not 12A... you know... similar size font and directly above the volts :)
    But yes, it could simply be product no. D-12A
    It shows that 'we see what we expect to see'.

    Re the windscreen wiper motor... perhaps they were the original source for this product.
    I haven't opened the gearbox as yet... I have resisted the temptation.
    I needed to first break the back of the jig.
    Once that's done, I'll check out the internals, and grease up.

    Re the motor amp demand - that's reassuring.
    I have a Makita fast charger DC1401
    It delivers 14.4V 4.5A = 64 watts

    The turntable isn't yet made - perhaps by the end of the day.
    Then we can test loads.
  8. Jul 11, 2017 #7
    I'd check what no-load current draw is, and what it increases to when loaded.
    Do you have a way to measure DC current?

    If not, get a 0.1Ω, 5 watt power resistor, place it in series with the motor, and measure voltage drop across it.

    I=V/R, so a reading of 100 mV across a 0.1 ohm resistor = 1 amp, 10 mV=0.1 amp, and so on.
    It isn't a bad idea to measure motor current this way even when a multimeter with a DC range is available; no worries about clearing the meter protection fuse due to start-up surges, etc.

    Another approach to speed controlling a low current, low voltage DC motor such as this is to use an LM317 adjustable voltage regulator. It is adjustable from 1.25 volts to 37 volts DC at 1.5 amp output. 1.25 volt minimum voltage output is one downside (it can't be adjusted to zero speed), and another is the 1.5 amp output limit. However, it isn't much more work to add a "pass" power transistor which allows higher current operation (depending on the transistor used, up to 6 to 8 amps).
    Last edited: Jul 11, 2017
  9. Jul 11, 2017 #8
    Most basic DMMS can measure up to 10A ( fused) - or use the 0.1 ohm shunt as pointed above.-- I would be concerned that this motor can draw more then 12A, soooo, you may want to try with a 6V battery and get and idea of how much this will draw.

    Controlling the DC supply. and therefore speed with a basic PWM is straight forward , once you know the parameters and limits of the motor.

    I am still puzzled by "rotate up to a motorcycle barrel", it seems the barrel is the engine casting. Can you post a photo or sketch?
  10. Jul 11, 2017 #9


    Staff: Mentor

    So your turntable is not to play music from vinyl disks. Are you making a lathe? Are you rotating objects for spray painting purposes? A lathe needs much more torque.
  11. Jul 11, 2017 #10
    Time for a pause, to see where we are
    This project concerns material surface modification through impact.
    This ranges from renovation, through dimensional modification, to densification.

    Most commercially produced vertical and rotational control systems are highly controlled.
    Both operate slowly.

    I always knew that (with help) I could easily control the rotational speed, however the effect of the moving metal (as it is impacted), is as yet unknown.
    We must presume that a circular delve becomes increasingly elliptical.

    Aerospace operations typically appear to be sub 10 rpm.
    ... but they have total vertical control... at this moment... we don't!
    We have an orange squeezer... but that is for primary examination of the rotary aspect of the jig.
    Vertical motorised control must be relegated to a 2nd project.

    Objects For Modification
    Soft aluminium, through cast iron, to hardened steel.
    This is a typical 2T m/bike barrel:
    MX50-barrel-top.jpg MX50-barrel-bottom.jpg

    The top photo shows the cylinder, surrounded by a cavity.
    Within the cavity, circulates coolant - driven through a radiator (by a pump).

    The bottom photo (top right) shows the large circular coolant duct.
    It is through this duct, that the coolant is pumped.
    Note: always carefully inspect this duct, to ensure it has been fully drilled, and that there are no obstructions - at pain of engine seizure.

    Staying with the bottom photo... there are two arced ducts top and bottom (transfer ports).
    The bottom duct (looking carefully) shows that it is split into two (at an angle).

    Move your eyes up to the two arced oblong ducts (ports) in the cylinder wall.
    These ports feed fuel into the cylinder (across the piston crown), and are fed via the transfer ports.

    The third transfer port (top left) feeds fuel through a port not shown.
    That port is angled upwards.
    The fuel gas is injected upwards, displacing the remaining combustion fumes, forcing them out through the exhaust.
    (note how the exhaust is opposite this transfer port)

    The elliptical delve above the ports is an 'oil retention reservoir'.
    Oil can become trapped in the rough surface, and drip into the cylinder, as the piston passes.

    The piston glides through this cylinder, riding on 'piston ring' contact.

    We need to carry oil molecules (oil bearings) between the piston and the cylinder.
    We also need to eliminate snagging of the piston ring edge between the ports and delves.
    We also wish to increase the hardness of both piston and cylinder, to enable them to act like true 'load bearing surfaces' and to minimise the transfer of aluminium (deposits) from the piston to the cylinder.

    Note: Evidently there are numerous complex objectives to hand - the piston traverses the cylinder typically over 300 times per second - with an explosion over 150 times per second.
    However, from this brief introduction, we can understand why we might wish to employ material modification techniques... and hence this project.​

    This is typical 'prototype engineering' on a shoestring.
    Older members will immediately recognise this as the true route to understanding the issues at stake.
    Younger members may wonder why the correct motor and gearbox (et al) was not specified first.

    The fact is... that the final definition anyway, rests in a fog.
    ... so we grab what we have available... and then with, community knowledge, make it happen.
    ... and then share the outcomes, back into the community.

    Ultimately, we will be able to determine an ideal spec... but (regardless) we will have got the prototype working correctly, and gained all the requisite knowledge.
    Hence the member requirement for fundamental motor data (for starters)

    Here's where the project is:
    • Window motor
    • Cut down 10mm bolt - ground to the motor 'spline drive'
    • Modified water pump 'threaded ground spigot' (over the bolt thread)
    • Water pump seal (to ground spigot)
    • The above fitted in an IP rated electrical box.
    • Outer guide bearing and cover plate (combined) - perspex/acrylic vehicle registration plate.
      Note: Drills cut small in this plastic - allowing final cutting with metal polish between the spigot and plastic
      An additional O ring will be added to further protect this bearing surface.

    turntable-drive-shaft.jpg turntable-internal-drive-shaft-seal.jpg

    Electrical Testing
    Here's a cautionary tale:

    I had left my multimeter ON (for whatever reason).
    When I came to it, I found the battery dead, the fuse broken, and the copper bar split.

    I replaced the fuse, and soldered the bar - for no other reason than, that was all I could do.
    My finger is touching where it was soldered:

    I do not know the implications of this failure... perhaps somebody can enlighten me.
    Here is the front face of the meter:

    Electrical parts

    I have numerous pieces of kit for spares.
    An old standard PC PSU might be useful:

    It seems to contain plenty of resistors.
    If anybody has any ideas on what bits might help us... that would be great.
    Tomorrow I will start checking resistor levels.
  12. Jul 11, 2017 #11
    Haha- thanks for the write up and pics. For me at this point the the speed control of the window motor is pretty well understood. How do you want to control this, with a knob, a digital setting? What is the budget? -- It can be done very cheap with discrete components, or for a little more digitally with something like an arduino - but still need to figure out how much current the motor will draw.

    So you want to mount the "barrel" on the turntable to do some process on it? The you mention "material surface modification through impact" - this part has nothing to do with the turntable, but you are looking to process the inside of the cylinder, this is with impact? -- kinda like honing - but with a "hammering" ? -- Sorry - it is just something I have not come across before - and am intrigued.
  13. Jul 11, 2017 #12


    Staff: Mentor

    Shot peening perhaps?
  14. Jul 12, 2017 #13
    Yes... but imagine whacking a static piece of metal with the ball end of a hammer: the delve would equate to the shape of the ball.
    But if the metal is sliding past - perhaps the delve is elliptical - perhaps densification is reduced due to the spreading of the impact force across a larger surface area.

    If the cylinder diameter is 47mm - the length of the steel = 147mm
    If rpm = 120
    In one second the metal has moved 294mm
    Therefore, in 1ms the metal has moved 0.3mm.

    This 'feels' way too much... and likely hence why the pro turntables run much slower.
    Some (for turbines) are in the 0.1 - 1 rpm range
    ... but let's not forget that the larger the diameter, the faster is the 'passing' metal.

    So really the rpm is derived from the permissible mm/sec movement of the surface.
    Ie. a smaller diameter shaft could be rotated at much greater rpm, to achieve the same effect.

    Vertical movement
    Ideally the impact point (or spread of impact points) would would drop after one rotation.
    This implies steps.
    However, due to the width of the impact band, a controlled descent will have the impacts slowly screwing down, with the impact band overlapping the previous, to achieve uniform impacts all the way down the cylinder.

    Oh yes... re budget

    For prototyping I try to use what I've got.
    The PWM controller must be bought... but they are inexpensive.
    I think that I have everything else.
  15. Jul 12, 2017 #14
    Re the administration of control.
    It would be evidently very nice to have a little computer.

    This would be great for calculating motor rpm from the diameter.
    Enter the diameter, and the turntable rotates at the correct speed.

    I think this would be a great 'phase 2'.

    In the meantime, a rotary knob would be ideal.
    Perhaps an rpm output is available - I need to check the pre assembled little units on ebay.

    This one has a soft start setup :http://www.ebay.co.uk/itm/DC-12V-80...oller-LED-Digital-Display-Panel-/332160641318
    ... but none so far indicate rpm.
  16. Jul 12, 2017 #15
    This controller doesn't have any terminals to connect a speed potentiometer ("speed knob"). It uses a pair of 'up/down' pushbuttons to set speed.
    I didn't see any way to zero and span the digital display - it is fixed to 0 to 100%. If meter zero and span were adjustable, what you'd do is set 'zero' to read '000.0' at zero speed. Next, turn on the drive, run it at maximum speed, independently measure turntable RPM (for instance, by using a stopwatch, and counting revolutions), then adjust meter 'span' to read out the measured RPM. Provided the relationship between actual turntable speed, and whatever it is the meter measures (for instance, armature voltage) are adequately linear, the meter is now scaled, and will display turntable RPM.
  17. Jul 12, 2017 #16
    Pretty sure - it is not really speed control, but just PWM control. The % is the % of the full voltage. Meaning the speed will vary as the load changes, may not be an issue in your case.

    Also - it still need the DC power supply.
  18. Jul 12, 2017 #17
    I'll have a look at the motor next.

    I wanted to get the turntable disc completed, because I knew it would be fiddly, and noisy to cut with a jig saw.
    Anyway, it's done, and here it is:

    This provides support options for large and small diameter workpieces.
  19. Jul 12, 2017 #18
    RE the comments concerning the PWM controller.

    Yes I need to be careful.
    I showed that link primarily because it had a 'soft start', which I thought was quite neat.

    Interesting that the speed would vary by load... say changing from a cast iron barrel to an aluminium piston.
    I'd never considered that.
    It looks like the ideal would be to digitally control the rpm, regardless of weight.
    Perhaps that's impossible with this motor.

    If so, a strong magnet attached to the disc might work.
    The sensor would need to be in the motor enclosure, and the field must pass via a steel grill 2.5mm thick:

    It would be possible to project the sensor external to the enclosure - perhaps through an enlarged hole.
    However, this is a messy solution, and best avoided if possible.
  20. Jul 12, 2017 #19

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    If you have an old fashioned "Variac" adjustable transformer, that and a car battery charger works very well. Would get your experiment going quickly.

    Old Ford windshield wiper motors had three brushes hence superior speed regulation. And lots of torque. If you can't close your speed control loop , consider that option for open loop.
  21. Jul 12, 2017 #20
    No I don't.
    I am currently testing with an 8A/4A battery charger.


    4A delivers 12V free of load
    8A delivers 13.55V free of load

    4A delivers 10.3V motor load - no weight
    8A delivers 12V motor load - no weight

    4A delivers 9V motor load - 1Kg
    8A delivers 10.9V motor load - 1kg

    4A delivers 8.7V motor load - 2Kg
    8A delivers 10.3V motor load - 2kg

    1. The Voltage delivered is varied - I've made a median guesstimate
    2. The RPM drops with weight, as predicted by project contributors
    Interesting that the voltage drop lessens, as the load increases.

    Next tests will be RPM, and Amps.
    Last edited: Jul 12, 2017
  22. Jul 12, 2017 #21

    4A delivers 12V free of load
    8A delivers 13.55V free of load

    4A delivers 10.3V motor load - no weight - 88 rpm -
    8A delivers 12V motor load - no weight - 104 rpm -

    4A delivers 9V motor load - 1 kg - 60 rpm -
    8A delivers 10.9V motor load - 1 kg - 80 rpm -

    4A delivers 8.7V motor load - 2 kg - 54 rpm -
    8A delivers 10.3V motor load - 2 kg - 68 rpm -

    To prevent the onset of madness... counting was restricted to 30 seconds - the figure then doubled.
    For the bulk of each period, the count establishes it's rhythm, allowing the eyes to stray to the stop watch.
    I would judge the accuracy to be +- 1 rpm
  23. Jul 12, 2017 #22
    Load - RPM - Amps

    4A delivers 12V free of load
    8A delivers 13.55V free of load

    4A delivers 10.3V motor load - no weight - 88 rpm - 1.5 A
    8A delivers 12V motor load - no weight - 104 rpm - 1.6 A

    4A delivers 9V motor load - 1 kg - 60 rpm - 2.1 A
    8A delivers 10.9V motor load - 1 kg - 80 rpm - 2.8 A

    4A delivers 8.7V motor load - 2 kg - 54 rpm - 2.3 A
    8A delivers 10.3V motor load - 2 kg - 68 rpm - 2.4 A

    Again, there was variability.
    Perhaps this is due to the past wear of the worm drive.
    Bear in mind that we are not dealing with magnetic or airless bearings.
    However, the results should provide a guide.

    Perhaps a spreadsheet plot might throw up any human misjudgement.
  24. Jul 12, 2017 #23
    What can we derive from the test results?

    I'm thinking that the voltage should be regulated to a fixed voltage - say a figure less than 10 volts.

    In this respect, the voltage would become fixed.
    I need to test with the Makita charger specced at 14.4V @ 4.5A

    But with a fixed voltage. this would stabilise the rpm, and only amperage would change.
    ... or is this unnecessary with PWM - in effect, itself stabilising the voltage?
  25. Jul 12, 2017 #24
    Rather than focusing on the drive controller for now, if turntable speed is currently in the ballpark, verify you can obtain the peining effect you are after. If you can (or it becomes apparent if might work at a slower turntable speed) then put effort into motor speed control.

    The PWM controller in your previous post is an open loop control, and outputs 0 to 100% of input voltage (for instance, the battery charger) in increments as small as 1%. Increasing motor load will still reduce supply voltage, and since speed is directly proportional to armature voltage, speed will still fall off as load increases.

    To obtain regulated speed control it is desirable to close the loop by measuring something representing speed, and using it for feedback. DC tach generators have a long tradition, but are ceding ground to rotary pulse generators and digital encoders. When measured speed drops below setpoint speed, the controller increases armature voltage to compensate, if speed is above setpoint it reduces armature voltage, and if measured speed equals setpoint it allows voltage output to remain unchanged.

    Each feature adds complexity, but a 'true' motor controller also has a way to set minimum and maximum speed, to control acceleration and deceleration (accel/decel, which performs essentially what 'soft start/stop' does in your example), and some form of current limiting to prevent mechanical overload from blowing up the drive and/or overheating the motor.

    The current measurement obtained on the 8A charger range with 1 kg load appears to be a mis-read (shown in red).

  26. Jul 12, 2017 #25
    Oh yes, I will certainly be trying it :)

    I have just tested the barrel (1.7Kg) @ 4A
    Bizarrely it rotates at 52 rpm.
    The expected rpm was 56.

    This may be because it is not balanced (due to the exhaust).
    The test weights were 'weight lifting' weights @ 1Kg each... so perfectly round.

    Using the scientific methodology of just looking at it spinning... I think we are entering the working range.
    Probably in the region of 30 rpm will be better.
    I'll experiment with resistance.

    I take your point regarding the complexities of a feedback system.
    ... but at this stage of the project it is always worthwhile examining options.

    Anyway, I wanted to supply the test results as per requests.
    This enables further input from the community.

    Apart from that... it may be the case that one quickly understands the settings for different loads.

    However... tomorrow I'm back on mechanical engineering (my primary domain).
    I need to create a new drive shaft that is longer - the one I was using was to hand, but is 10mm too short.
    I also must grommet the enclosure, and add flex.
    Plus the turntable studs to locate the barrel and eliminate 'a blind hole', to allow full cylinder penetration (and facilitate air flow).

    Vertical movement will be by hand :nb)
    I'll try and design a simple 'centralisation rest'.
    The problem is the robustness of the pressure lines - very numb and they exert their own pressure.
    ... having something to push against, central to the cylinder, is probably the minimum necessity.

    However, I have numerous barrels to work with, before attempting 'the task' on my new top end.
    Maybe tomorrow night, I'll have something to show for the work so far.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted