Tips to quite radiated emissions of small 24V DC brush motor

[Ge0rge Marutz]

Hello Gentleman (and ladies?),

My company manufactures an automotive assembly that uses a small 24VDC
motor to pump fluid. We recently went through a round of EMC testing
under SAE J1113-41 guidelines.

To make a long story short our product behaves very well until the
motor turns on. At this point emissions go through the roof. The
product is well above an acceptable limit.

A little background. The PCB and pump motor are housed in a custom
molded plastic enclosure. It is not possible to change the housing at
this time in case you were going to recommend metal. The motor sits on
top of the PCB and is housed in its own plastic pump sub-assembly
(again, cannot change). The circuit board is a 4 layer board with
dedicated power and ground planes. The pump is switched on and off
with a N-channel MOSFET low side drive transistor. Two wires
(approximately 30mm each) go from the circuit board to the pump. I have
1000pf capacitors placed on the PCB from the point of entry of each
wire to the PCB and ground (common mode filters I believe). I also
have a 100nF SMD capacitor placed between the pump motor wire leads at
the point where the wires leave the board. The PCB seems to be immune
to the high levels of RE. We have had no problems even considering the
fact that this noisy pump sits right on top of the PCB. My big concern
is what this motor might do to other electronics. We would like to
meet SAE J-1113 guidelines for at least a region IV device, preferably
region III or higher.

I tried to attenuate RE by connecting the motors metal body to the
ground plane of my PCB. It did not seem to help much.

Do you have any suggestions? Is there something I can do to my
assembly to quiet down RE ( like metalized paint on interior of
housing)? Is there something that can be changed with the pump design
to reduce RE?

Looking for any hints you have to offer.

Thank you,

George

 

[Ge0rge Marutz]

"We would like to meet SAE J-1113 guidelines for at least a region IV
device, preferably region III or higher."

Sorry, I was confused. I meant to say we would like our product to be
considered at least Class 2, prefereably Class 3 or higher.

Ge0rge

 

[Jim Thompson]

Hello Gentleman (and ladies?),

My company manufactures an automotive assembly that uses a small 24VDC
motor to pump fluid. We recently went through a round of EMC testing
under SAE J1113-41 guidelines.

To make a long story short our product behaves very well until the
motor turns on. At this point emissions go through the roof. The
product is well above an acceptable limit.

A little background. The PCB and pump motor are housed in a custom
molded plastic enclosure. It is not possible to change the housing at
this time in case you were going to recommend metal. The motor sits on
top of the PCB and is housed in its own plastic pump sub-assembly
(again, cannot change). The circuit board is a 4 layer board with
dedicated power and ground planes. The pump is switched on and off
with a N-channel MOSFET low side drive transistor. Two wires
(approximately 30mm each) go from the circuit board to the pump. I have
1000pf capacitors placed on the PCB from the point of entry of each
wire to the PCB and ground (common mode filters I believe). I also
have a 100nF SMD capacitor placed between the pump motor wire leads at
the point where the wires leave the board. The PCB seems to be immune
to the high levels of RE. We have had no problems even considering the
fact that this noisy pump sits right on top of the PCB. My big concern
is what this motor might do to other electronics. We would like to
meet SAE J-1113 guidelines for at least a region IV device, preferably
region III or higher.

I tried to attenuate RE by connecting the motors metal body to the
ground plane of my PCB. It did not seem to help much.

Do you have any suggestions? Is there something I can do to my
assembly to quiet down RE ( like metalized paint on interior of
housing)? Is there something that can be changed with the pump design
to reduce RE?

Looking for any hints you have to offer.

Thank you,

George

Could you change to BRUSHLESS EC (electronically commutated)?

...Jim Thompson

 

[Ge0rge Marutz]

Don't electrically cummutated motors require much more complex drive
circuitry? Could the circuit fit in the same package outline occupied
by a SOT-223 transistor and five or six 0603 caps and resistors? PCB
real estate is at an extreme premium for me.

Thank you for the feedback

Ge0rge

 

[Dan Mills]

Hello Gentleman (and ladies?),

My company manufactures an automotive assembly that uses a small 24VDC
motor to pump fluid. We recently went through a round of EMC testing
under SAE J1113-41 guidelines.

To make a long story short our product behaves very well until the
motor turns on. At this point emissions go through the roof. The
product is well above an acceptable limit.
[Chop]

I tried to attenuate RE by connecting the motors metal body to the
ground plane of my PCB. It did not seem to help much.

try a 1nF ceramic cap from each motor terminal to the motor body (short
leads!), also a ferrite ring with both motor leads passed thru it a few
turns can help.

The caps from the motor terminals is how the radio controlled toy cars
manage to work with really NASTY drive motors and cheap radios, so it may
solve your problem.

Just some random thoughts.

Regards, Dan.

 

[Larry Brasfield]

Hello Gentleman (and ladies?), Hi.
My company manufactures an automotive assembly that uses a small 24VDC
motor to pump fluid. We recently went through a round of EMC testing
under SAE J1113-41 guidelines.

To make a long story short our product behaves very well until the
motor turns on. At this point emissions go through the roof. The
product is well above an acceptable limit.

A little background. The PCB and pump motor are housed in a custom
molded plastic enclosure. It is not possible to change the housing at
this time in case you were going to recommend metal. The motor sits on
top of the PCB and is housed in its own plastic pump sub-assembly
(again, cannot change). The circuit board is a 4 layer board with
dedicated power and ground planes. The pump is switched on and off
with a N-channel MOSFET low side drive transistor. Two wires
(approximately 30mm each) go from the circuit board to the pump. I have
1000pf capacitors placed on the PCB from the point of entry of each
wire to the PCB and ground (common mode filters I believe). I also
have a 100nF SMD capacitor placed between the pump motor wire leads at
the point where the wires leave the board. The PCB seems to be immune
to the high levels of RE. We have had no problems even considering the
fact that this noisy pump sits right on top of the PCB. My big concern
is what this motor might do to other electronics. We would like to
meet SAE J-1113 guidelines for at least a region IV device, preferably
region III or higher.

I tried to attenuate RE by connecting the motors metal body to the
ground plane of my PCB. It did not seem to help much.

Do you have any suggestions? Is there something I can do to my
assembly to quiet down RE ( like metalized paint on interior of
housing)? Is there something that can be changed with the pump design
to reduce RE?

Looking for any hints you have to offer.

I once worked on a product which had a transducer
within a couple inches of a brushed DC motor where
the signal transduced was a few MHz wide and was
expected to be noise free down to the approximately
50 Ohm noise floor of the transducer. Needless to
say, brush noise was a real issue with that design.

One "fix" that did the most good was to place a small
ceramic cap right across the motor leads where they
exited the metal housing of the motor itself. From the
standpoint of localizing interfering currents, this is about
as good as you can do with a brushed motor. For this
to be maximally effective, the cap leads must be kept
as short as possible. Ideally, the RF current would not
have to leave the motor housing at all; the cap would be
placed inside across the brushes themselves.

If your motor is not in a metal housing, I doubt that you
will have much sucess bringing the brush noise down by
much. In that case, Jim's suggestion of a brushless motor
would be mine as well.

 

[Fred Bloggs]

I once worked on a product which had a transducer
within a couple inches of a brushed DC motor where
the signal transduced was a few MHz wide and was
expected to be noise free down to the approximately
50 Ohm noise floor of the transducer. Needless to
say, brush noise was a real issue with that design.

One "fix" that did the most good was to place a small
ceramic cap right across the motor leads where they
exited the metal housing of the motor itself. From the
standpoint of localizing interfering currents, this is about
as good as you can do with a brushed motor. For this
to be maximally effective, the cap leads must be kept
as short as possible. Ideally, the RF current would not
have to leave the motor housing at all; the cap would be
placed inside across the brushes themselves.

If your motor is not in a metal housing, I doubt that you
will have much sucess bringing the brush noise down by
much. In that case, Jim's suggestion of a brushless motor
would be mine as well.

As usual that is only half the answer- DC motor EMI suppression
strives to confine RF current circulation to paths of small dimension
relative to the frequencies involved- and this means they must be kept
off the power lead wires. It also means common mode low impedance shunt
at each line to GND as well as the differential shunt you mention. As
far as a good GND goes- he must make a high frequency low inductance
plane at the motor itself- and this can be a simple band around the
housing. This should be joined to the controller with a braided shield-
might as well be around the power feed.

 

[R Adsett]

Don't electrically cummutated motors require much more complex drive
circuitry? Could the circuit fit in the same package outline occupied
by a SOT-223 transistor and five or six 0603 caps and resistors? PCB
real estate is at an extreme premium for me.

More complex drive circuitry yes, but at least some small burushless
motors have the commutating circuitry built-in. Of course that may just
mean that the motor is too large for you.

Robert

 

[[email protected]]

yep, as said a cap as close to the motor terminals as poss, ie
definately mounted on the motor itself. Plus ferret beads - not sure
which part of the ferret theyre made from.

I dont rememebr trying this, but screened motor leads would likely help
with whatevers left after that. Ditto keeping the output traces running
side by side, to some extent.


NT

 

[JeffM]

Plus ferret beads - not sure which part of the ferret theyre made
from.

NT (bigcat @ meeow.co.uk)

As long as we're being silly, are those micro ferrets or nano ferrets?

 

[[email protected]]

As long as we're being silly, are those micro ferrets or nano

ferrets?

Micky ferret, Nanna ferret and Millie ferret.

NT

 

[JeffM]

micro ferrets or nano ferrets?

Micky ferret, Nanna ferret and Millie ferret.
NT (bigcat @ meeow.co.uk)

Even better.

 

[Terry Given]

Hello Gentleman (and ladies?),

My company manufactures an automotive assembly that uses a small 24VDC
motor to pump fluid. We recently went through a round of EMC testing
under SAE J1113-41 guidelines.

To make a long story short our product behaves very well until the
motor turns on. At this point emissions go through the roof. The
product is well above an acceptable limit.

A little background. The PCB and pump motor are housed in a custom
molded plastic enclosure. It is not possible to change the housing at
this time in case you were going to recommend metal. The motor sits on
top of the PCB and is housed in its own plastic pump sub-assembly
(again, cannot change). The circuit board is a 4 layer board with
dedicated power and ground planes. The pump is switched on and off
with a N-channel MOSFET low side drive transistor. Two wires
(approximately 30mm each) go from the circuit board to the pump. I have
1000pf capacitors placed on the PCB from the point of entry of each
wire to the PCB and ground (common mode filters I believe). I also
have a 100nF SMD capacitor placed between the pump motor wire leads at
the point where the wires leave the board. The PCB seems to be immune
to the high levels of RE. We have had no problems even considering the
fact that this noisy pump sits right on top of the PCB. My big concern
is what this motor might do to other electronics. We would like to
meet SAE J-1113 guidelines for at least a region IV device, preferably
region III or higher.

I tried to attenuate RE by connecting the motors metal body to the
ground plane of my PCB. It did not seem to help much.

Do you have any suggestions? Is there something I can do to my
assembly to quiet down RE ( like metalized paint on interior of
housing)? Is there something that can be changed with the pump design
to reduce RE?

Looking for any hints you have to offer.

Thank you,

George

twist the leads to the motor - about one twist per cm.

As well as FB & LBs suggestions.

Cheers
Terry