Building a motor controller for a Samsung brushless direct drive washer motor.

[Razerwire]

I have recently acquired a Samsung front loading washing machine. It was working fine except for the door latch mechanism was broken so the previous owner had used huge self stick velcro straps to hold the door shut... anyway I only wanted the guts, the rest I scrapped minus the drum for a fire pit. My question is, On the motor there is the usual 3 pin(3 phase) power connector and a rotor position sensor connector. The odd thing I can't find info about anywhere, is why the hall effect sensor connector is only 4 pin instead of 5 as it is on all sensored  brushless motors I've seen? Normally the RPS/Hall sensors have 5 wires. One power, three for Hall A B C, and one ground. Neither a controller I bought nor one I built will run this motor,  however I haven't tried using either. I have the control board and the separate inverter board which the motor and sensors connect to. I want to build a controller for this motor or even better just use the factory inverter board (fed with my own control signals) to run it.. any information or advice would be much appreciated. Once I understand the sensors, making a controller will be straightforward. I could of course tear into the sensor module and see the sensors physical form and layout but I want to avoid damaging anything. Besides, opening it up to basically find it has two hall sensors instead of 3, still tells me little about how to drive the motor. The pin identifiers for the sensor cable on the board and in the very simple block diagram in the technical info pamphlet simply say, Vcc, U, V and GND. I'll include pictures of the motor,  the 4 pin sensor plugs and the factory control board. The factory motor control inverter board has (from left to right in the picture) 2 pin molex power input, 3 pin molex power output to motor, 4 pin molex hall sensor feedback, and a 3 pin molex control signal input from the main control board.. No set plans for a use yet but I'm looking forward to the design challenge. Might eventually turn it into a generator if I can't find a good use for it. Thanks for sharing your time and knowledge. -Rex- 

[DavidAlfa]

Modern drivers don't require hall sensors, they're sensorless, search "sensorless bldc".
The 4th wire might be because it's connected in star, not triangle.
The best you could to is to investigate the original controller. It might not be so complicated after all.
You might get more help after providing high quality pictures where part numbers are readable.

[Razerwire]

Thanks, I will update and include all part numbers.

[Razerwire]

Quick question DavidAlfa, I understand that many newer BLDC motors are sensorless, however this motors wiring diagram clearly states for the 4 pin plug to the motor, "Hall sensor." There are three power phases going into the motor as well as a 4 wire "hall sensor," cable. You mentioned what "the 4th wire might be," which is a little vague since I don't know exactly which wire you're referring to since, total, the motor has seven wires ran to it. 3 power and 4 sensor. The 4 sensor wires are #1 Vcc(power) #2 U (?) #3 V (?) and #4 ground. When you say the 4th wire my mind was thinking the sensor ground wire, but that wouldn't make sense. Thanks for any clarification you can offer. I will investigate the stator winding wiring configuration, good tip.

[Razerwire]

The model number for the Samsung front loading washer is WF42H5000A. The part number for the main control board is DC92-01803D. The part number for the inverter board (the one that actually drives the motor) is DC92-01531B. I imagine that if I could see the signal that the main board feeds the inverter I would have a good idea what signal form I need to replicate at the 3 pin input to the inverter board. Finding out the voltage of the power going into the inverter board is easy, as is would be measuring the feed power to the hall sensor. My oscilloscope can't seem to get a definitive read on the three wire drive input signal though. I see nothing resembling a PWM or other pulsed signal. In fact it's reading as just plain old AC which can't be the case.

[fourtytwo42]

May be of no interest but I have read these make quite good wind turbine alternators, then you won't need the hall sensors!

[Benta]

Could be that there are only two Hall sensors in the motor.
You don't actually need three, except for direct commutation.
Otherwise two, spaced at any angle (except 180 degrees), will give the controller enough information to compute the 3-phase drive signal.

Try spinning the motor by hand and 'scoping the Hall signals.

[twospoons]

You will likely find hall sensors in washing machine motors because they reverse frequently and run slowly during the wash cycle, while under a heavy load. Back-EMF sensing becomes difficult under those conditions.

[Razerwire]

Spinning the motor and scoping the hall sensors is a great idea. Hadn't thought of that. I'll feed them with my bench power supply, hook up the scope and just see what comes out. I did end up trying to run the controller I bought and just leave one of the three hall wires off, alternating all the combinations just to see if it'd drive it with two hall wires....nope. The motor locks up and shakes back and  forth is all I get out of it. I am curious to see what the hall sensors scope out like. Thanks. Let me know if you think of anything else.

[Benta]

Well, use a battery screw driver to spin the motor, that should give a steady signal for the scope. Let us know what you find.

[Razerwire]

Will do

[amyk]

A 3-phase AC motor controlled by a VFD doesn't need a sensor either.

[Razerwire]

Yeah, I'm still just curious as to what the "hall sensors," on this motor actually are. Do you think its a brushless AC motor with VFD input? What leads you to that line of thought if so? Samsung keeps all of their non homeowner, real tech specs and diagrams under lock and key. The little homeowner tech pamphlet that came with the washer could call them flux capacitors on the block wiring diagram and most people would never notice anyway. I still haven't scoped the sensors. Got into my Cockroft-Walton voltage multiplier project.

[amyk]

I guess it might be speed feedback and/or stall detection.

[eugene]

I think I would study the PCB on the motor that contains the "Hall sensors." Might find some clues.

[Razerwire]

That could be the case and Samsung mislabeled them as Hall sensors in the diagram. The control boards have no legible information and are potted in at least a quarter inch of tough silicone. I haven't removed the sensor module from the motor. I will investigate and get back with you guys. It's incredible that I haven't been able to find and technical info on this motor or any detailed stories of anyone getting one to run. EXCEPT.... Two different instances where guys had removed the control boards from the washer, tricked the safety sensors and got the motor to run using the factory control boards with the motor running a cycle. This permitted maximum 5 min runs at full power during spin cycle. We may be breaking new ground a bit and if we can figure out how to have full manual control of this motor it could open up many project opportunities for folks. This particular motor is unlike many other direct drive "pancake," brushless washing machine motors such as the Fisher Paykel Smart drive motors which people have no issues running with a standard 3 hall sensor controller. Thanks for all the input. These motors have serious potential with their massive low speed torque and smooth running. All feedback is appreciated!

[DavidAlfa]

There are two connectors in the motor, both visible in your pictures.
One is 3-pin (power) and the other is 4-pin (Hall sensor).
Check the attached file, pages 21 and 22.

[Razerwire]

I have that same exact booklet with that same exact wiring diagram. And now see this is the kicker, what has really been grinding my gears... is if you'll look at the diagram at the wiring for the hall sensors... it says the wire colors are from left to right: pink, red, blue and orange. Yet the actual wires I'm looking at are L to R: blue orange red pink, at the motor. The diagram does not accurately represent the actual color coding layout. At least in my case, idk, but  there must be a way...

[DavidAlfa]

In this video the wire colors seem to match:




Unless it's some weird protocol, I think these 3 wires should do something like:
- Run/Stop
- Direction
- PWM signal for speed



I'd disassemble the power module and trace these wires.
Oh and don't forget to make good quality pictures ensuring the parts are readable.
Otherwise it'll be much harder for us to guess.


Remember that board is directlty connected to mains. It's just mains rectified.
So should be very careful when measuring using the oscilloscope!
If you only see AC, it might be some grounding issue?

[Razerwire]

So I removed the sensor module from the motor and included pics of it below. The two spurs sticking up in the pic with my fingers on them, I believe are the two hall sensors (every other hall as sensored BLDC I've ever seen had 3 hall sensors), or some other kind of double pronged sensor. Since 4 wires go into the sensor module and we know that one is power and one is ground then that leaves 2 wires for 2 hall sensors or some kind of sensor that uses two wires and is two pronged. The actual guts of the sensors are potted under that rectangular gray patch of tough silicone you see in the picture. I even found the model number for the motor finally, it was faint and corroded but legible on the rotor.  It's model # COM1500P DC31-00112A. It says: DC 310V.  0.6HP.  2.5A.  1400RPM.  3P.    NEWMOTECH CO. LTD (SUZHOUTAIXIN) 2016. 05. 12. 09:41:29.

[Razerwire]

I also discovered that the stator windings are lacquered aluminum and not copper which makes it very undesirable for any kind of repressing which exposes it to the elements f or extended periods such as a home hydro electric generator or wind mill generator, which are the two most popular purposes for this type and shape of motor. I just got unlucky and scored a unit with the aluminum stator windings. It'd still make one hell of a motor for a garage/indoor/nice weather project.

[DavidAlfa]

Check the driver in the board! You'll need to desolder it, the marking is often in the underside, the top is all heatsink.
Those kind of boards use igbt modules like SCM1242MF, FNA41560...

[Razerwire]

Ok I see what you're saying. Slight roadblock is that the board is potted under half inch of silicone. Any clever way you know of to get through that without too much butchery?

[Benta]

If the chematic is correct, it's like I said earlier: there are only two Hall sensors, U and V, the W phase being computed by the controller from those.

[ElectricLove1975]

Hello!

I am trying to solve the exact problem you guys talked about in this thread with the exact same motor, but I do not have the control board. I am trying to control one of these motors using a commerical controller with a standard 5 pin hall effect sensor.

I have started a new thread about this here:
https://www.eevblog.com/forum/projects/finding-a-controller-to-maximize-torque-in-salvaged-washing-machine-motor/

OP, did you end up using the controller from the original washing machine? I stupidly threw mine away when I took it apart so that isn't an option for me.

Is there a way to solve this 4 pin / 5 pin matchup issue outside of just replacing the sensor on the motor that I'm not seeing? Trying to replace the sensor seems like it could be really messy and very difficult to get exactly right as opposed to just figuring out how to convert the connection somehow.

Thanks!