Wilo Pump

[blauerscharik]

Hi guys,

I've got this Wilo circulating pump which stopped working. I cracked it open to see what failed.
The IC in the middle blew up. I de-soldered it.
It's was an FSB70250 https://eu.mouser.com/ProductDetail/onsemi-Fairchild/FSB70250?qs=USvxm8v2%2FcTBlvh8JEMFIg%3D%3D
It's a pretty complicated circuit board. I looked for a schematic but couldn't find anything. Can anyone help me understand how it it works and why that IC would explode and how I can test if everything else is good.

Thanks

https://ibb.co/ThzfCv2
https://ibb.co/wL1Zfc9
https://ibb.co/9GLjnkP

[shakalnokturn]

I'm not familiar with them myself but the basic idea is to rectify the mains and have a small VFD drive a PMSM (or BLDC?) motor.

So you'll have a diode bridge large capacitor, small SMPS for the microprocessor in charge of PWM control, soft-start, sensors... And a 3 phase H bridge to drive the motor, the component that blew-up.

The reasons for blowing up could be: Power surge, overcurrent (jammed pump), poor control inputs (weak onboard SMPS), dampness...

[blauerscharik]

I'm not familiar with them myself but the basic idea is to rectify the mains and have a small VFD drive a PMSM (or BLDC?) motor.

So you'll have a diode bridge large capacitor, small SMPS for the microprocessor in charge of PWM control, soft-start, sensors... And a 3 phase H bridge to drive the motor, the component that blew-up.

The reasons for blowing up could be: Power surge, overcurrent (jammed pump), poor control inputs (weak onboard SMPS), dampness...

Thanks.
That's right it got 3 phase going to the motor. I took that cover off as well and I think it's a BLDC, it got 6 coils. Weird is that there is an additional connection to the coils so it's actually 4 in total but one is covered and not connected anywhere...
 
Power surge is is the most likely. Poor control inputs, weak SMPS? Is that something I can measure?

When I apply power to it. I got this transistor marking says (ODN) that gets hot and makes a high pitch noise. Not sure if that is supposed to be like that or not.

Do you think there is a way to bypass this H-Bridge?

For testing I tried to apply mains voltage (trough a light bulb) to the coils and the pump does turn slowly. But then it gets stuck between two coils when you remove power and want to start again. A little push and it turns again.

[shakalnokturn]

Ideally you'd test the SMPS with similar loading to when the pump is running but measuring voltages and ripple as-is should be good enough.

What's the reference on the SMPS IC, the large one in DIP8 package with missing pin #6?

I don't think you should try to bypass the H bridge and run at 220V 50Hz, the motor just isn't designed for that use.
If replacing the blown H bridge is a problem because of SMD package, if free space allows, you may be able to find an electrically compatible substitute that you can wire in, it will need somewhere to dissipate heat, added wire inductance could lead to problems though.

I can't see mentioned transistor in photos.

[blauerscharik]

Ideally you'd test the SMPS with similar loading to when the pump is running but measuring voltages and ripple as-is should be good enough.

I get 310VDC at the cap and also on the U/V/W In on the blown IC.

What's the reference on the SMPS IC, the large one in DIP8 package with missing pin #6?

That's this fella right here: https://www.mouser.com/datasheet/2/149/FSL306LR-336188.pdf

I don't think you should try to bypass the H bridge and run at 220V 50Hz, the motor just isn't designed for that use.

Oh no I'm not going to. That was just for testing and experimenting/understanding how this works.

If replacing the blown H bridge is a problem because of SMD package, if free space allows, you may be able to find an electrically compatible substitute that you can wire in, it will need somewhere to dissipate heat, added wire inductance could lead to problems though.

Space shouldn't be an issue. I could have also something external of the pump if needed. I already looked at how to build an ESC with 6 Mosfets. I just don't know if putting Mosfets in the circuit alone will work though. There is all these Pin 1-13 or so. Not sure if they are actually needed.

I can't see mentioned transistor in photos.

I marked it on the picture.
The guy next to it is this one: https://pdf1.alldatasheet.com/datasheet-pdf/view/12073/ONSEMI/MC33063.html

[blauerscharik]

So I ordered a new FSB70250. I soldered it onto the board today. I then hooked it up to mains, through a light bulb and transformer and POP it smoked again.
Now it just gets hot, looking at it through the thermal camera. Don't know if it was my soldering or something else not right on the board. Well that was kinda waste of money but at least I learned something.   

Now my thought is to build my own ESC with 6 Mosfets and Arduino for PWM. What do you think? Or is there still a way to make it run again with the same board?

Thanks

[shakalnokturn]

Your SOT23-5 transistor that gets warm must be for a second step-down SMPS for the CPU I guess.

Did you check supply voltages before replacing the FSB70250? It isn't always obvious finding why things fail, it's a shame to not to investigate before replacing a part though, a good practice to save time and money.

Are you sure the rotor turns freely? I recently had a pump that wouldn't start on low speed. A small ball of linen had nested in the pump.
Are all windings similar in resistance and inductance?
Have you checked winding insulation resistance?
You seem to have an isolation transformer if you have an oscilloscope it could have been worth cheeking the control lines to the FSB70250 before replacing it. (The problem is that if there's a speed sensor the CPU may shut those off quite fast because the motor isn't really rotating.

You could give your own design a try, if you know where you're going. It could also be long and disappointing. The original H bridge also has current monitoring. (Not that it saved it from destruction anyway...)

[blauerscharik]

Did you check supply voltages before replacing the FSB70250? It isn't always obvious finding why things fail, it's a shame to not to investigate before replacing a part though, a good practice to save time and money.

Yes I checked it and everything seemed fine.

Are you sure the rotor turns freely? I recently had a pump that wouldn't start on low speed. A small ball of linen had nested in the pump.
Are all windings similar in resistance and inductance?
Have you checked winding insulation resistance?
You seem to have an isolation transformer if you have an oscilloscope it could have been worth cheeking the control lines to the FSB70250 before replacing it. (The problem is that if there's a speed sensor the CPU may shut those off quite fast because the motor isn't really rotating.

The rotor is fine I took everything apart, I even hooked it up to mains briefly and it was spinning.
Resistance is good and on all windings equal.
I can't check the insulation though, but I don't think that there is a problem.
I do have a scope, but that needs fixing as well   

You could give your own design a try, if you know where you're going. It could also be long and disappointing. The original H bridge also has current monitoring. (Not that it saved it from destruction anyway...)

I think it will be the long and disappointing route  but I might do it still, just to learn something.