I have a physical application that I believe would benefit from an LC filter. While I am familiar with RC filters for 5vdc analog signals I cannot say the same for power related LC filters and am hoping folks here can help educate me.

COMPONENTS
1) 36 vdc, 40 amp brushed motor (a minn-kota trolling motor).
2) Roboteq brushed motor controller, 50vdc max, 60 amp max, PWM operating at 18khz
3) 12 vdc ultrasonic transceiver operating at 32.767 khz

CONFIGURATION
A) The motors and controller are powered by a 36v (nominal) LiFePO4 battery.
B) The 12 vdc ultrasonic transceiver is powered by a 12v (nominal) LiFePO4 battery.
C) The two batteries (12v and 36v) are completely isolated from each other, ie no wiring intermingled, not even the grounds are connected together.
D) The ultrasonic transducer (antenna) is wired to the transceiver with a shielded RG58 cable.

PROBLEM
The ultrasonic transceiver, in receive mode, is picking up a "hum" from the motor controller which I assume is EMI, EMF, both? I am able to change the PWM frequency of the motor controller between 10khz and 24khz however doing so provided no relief other than a change in pitch of the "hum" which I pretty much expected.

QUESTIONS
1) Would a low pass LC power filter be appropriate in this instance?
2) What is an appropriate cutoff frequency given these components?
3) Where does such a filter get installed? On the power wires to the motors? On the power wires to the motor controller? On the ultrasonic transceiver?

 

QUESTIONS
1) Would a low pass LC power filter be appropriate in this instance?
2) What is an appropriate cutoff frequency given these components?
3) Where does such a filter get installed? On the power wires to the motors? On the power wires to the motor controller? On the ultrasonic transceiver?

1) Maybe. It depends on the frequency and amplitude of the interference signal. What does it look like on the oscilloscope? Is there interference on the power lines to the receiver? Is the receiver designed (balanced line) to minimise interference pickup?
Have you eliminated the possibility that the interference is being picked up acoustically?
2) Depends on 1
3) Depends on 2

 

"hum"

What frequency?

 

What frequency?

All frequencies that the motor controller PWM is capable of, from 10khz to 24khz. If I set the PWM to 10khz I hear an identical tone out of the ultrasonic receiver. If I set the PWM to 18khz I hear an 18khz tone out of the ultrasonic receiver.

 

1) Maybe. It depends on the frequency and amplitude of the interference signal. What does it look like on the oscilloscope? Is there interference on the power lines to the receiver? Is the receiver designed (balanced line) to minimise interference pickup?
Have you eliminated the possibility that the interference is being picked up acoustically?
2) Depends on 1
3) Depends on 2

I don't have access to an oscilloscope.
The receiver uses rg58 coax for the ultrasonic transducer so is unbalanced.
Regarding acoustic interference, the normal working condition is inside a large steel tank containing both the controllers and the ultrasonic receiver. The ultrasonic transducer is always outside the tank. If I move the receiver outside the tank there is no interference at all.

Unfortunately I don't have lots of testing equipment to properly answer your questions.

 

I suspect the trolling motor ESC is radiating EMI all over the boat if it is an aluminum hull. The common method is by parasitic coupling reactance.
It is harder to use a high inductance for such a low frequency considered as VLF< 30 kHz when the motor current is high. So rather than suppresses the source, I suggest improve isolation on the ultrasound transceiver. One method of EMI is from high common mode noise become differential by an unbalanced receiver from coupling capacitance. The common solution is to use ferrite torroids around the cable or use a well-balanced 1:1 CM choke with appropriate L and frequency range. It might not be easy to do with coaxial connectors and might not be the fix until the cause is found. Trying to make it worse or better offers clues, like changing proximity and orientation. Unfortunately, if it is an alum hull it can't be moved to the bow and epoxy to the hull.

 

All frequencies that the motor controller PWM is capable of, from 10khz to 24khz. If I set the PWM to 10khz I hear an identical tone out of the ultrasonic receiver. If I set the PWM to 18khz I hear an 18khz tone out of the ultrasonic receiver.

Don’t forget that the PWM is a squarewave, so it will be radiating all the odd-numbered harmonics as well.

 

I don't have access to an oscilloscope.
Unfortunately I don't have lots of testing equipment to properly answer your questions.

I hate to disappoint you, but you’re not going to fix without proper test equipment. You guess at probable causes and try solutions at random, but it might take a very long time.

 

I hate to disappoint you, but you’re not going to fix without proper test equipment. You guess at probable causes and try solutions at random, but it might take a very long time.

Acknowledged and agreed.

 

I suspect the trolling motor ESC is radiating EMI all over the boat if it is an aluminum hull. The common method is by parasitic coupling reactance.
It is harder to use a high inductance for such a low frequency considered as VLF< 30 kHz when the motor current is high. So rather than suppresses the source, I suggest improve isolation on the ultrasound transceiver. One method of EMI is from high common mode noise become differential by an unbalanced receiver from coupling capacitance. The common solution is to use ferrite torroids around the cable or use a well-balanced 1:1 CM choke with appropriate L and frequency range. It might not be easy to do with coaxial connectors and might not be the fix until the cause is found. Trying to make it worse or better offers clues, like changing proximity and orientation. Unfortunately, if it is an alum hull it can't be moved to the bow and epoxy to the hull.

Unfortunately repositioning the receiver offers no relief, except outside the tank which is fine for testing but not in application.

One other piece of anecdotal information I can provide, if it helps; these ultrasonic receivers are designed to operate with local "AA" batteries. The problem with these small batteries is short operational life. Others have reported to me the exact same problem when using an external battery (in this case a 304ah 12v LiFePO4) however when they power the unit directly with "AA" batteries the problem goes away completely. Is this meaningful other than building a budget for throw away batteries and replacing them continually?

 

Unfortunately repositioning the receiver offers no relief, except outside the tank which is fine for testing but not in application.

One other piece of anecdotal information I can provide, if it helps; these ultrasonic receivers are designed to operate with local "AA" batteries. The problem with these small batteries is short operational life. Others have reported to me the exact same problem when using an external battery (in this case a 304ah 12v LiFePO4) however when they power the unit directly with "AA" batteries the problem goes away completely. Is this meaningful other than building a budget for throw away batteries and replacing them continually?

This can help you understand how interference passes to the ultrasonic part of the system.
Put the battery in a separate metal box and connect the box body to the ground of the ultrasonic sensor . Isolate the box from the common ground .
Such a test will help to understand whether interference passes through the battery circuit

 

I imagine the DC wires from the battery for both the motor power and sensor power both have some loop area and proximity (parasitic) mutual inductance and coupling capacitance. A CM choke works with a large CM inductance and the null Diff. Mode, DM inductance. This adds isolation to the sensor like small floating "AA" batteries but not like a 304Ah which has a large gap and loop area for radiated "antenna" effects as well as conducted effects of the motor ESC pulsing the battery voltage. Without knowing the conducted Voltage ripple shared by the sensor, it could be conducted or radiated. If conducted, that is easily fixed. The problem is NOT knowing the root cause.

 

Not mentioned so far is the wiring from battery to motor controller and from motor controller to the motor. These need to as short as possible and positive and negative twisted together (or tie wrapped together at short intervals if very chunky) to minimise radiation from these. I would be reluctant to low pass filter the power to the motor without knowing how the controller works. If it is a type that detects back EMF on the pulses to control torque, an inappropriate filter could interfere with this. The battery feed to the controller is more amenable to filtering if there is not sufficient in the controller, but finding large enough inductors with a sufficient current rating will be hard.

 

Unfortunately I do not have much leeway with wire lengths in this application. From 35vdc battery to controller to motors totals about 15 feet. From 12vdc to communication transceiver totals about 7 feet. Because of this I feel filters and/or faraday cages may be my only options to limit the EMI.

I'm currently considering fabrication of an LC filter (1mH inductor and 10uF capacitor) producing a cutoff frequency of 1.5khz and connecting it directly to the 12vdc input power rails of the transceiver. Hoping that will have some impact...I don't have to get rid of all the noise, just enough to make comms easier.