SDC2130 PWM interference

6 years 11 months ago #29528341 by infoa
Hello,

We are using SDC2130 controller, and in the same system, we are using a analog video camera.
We can see a interference in the video signal, and this interference comes from pwm signal of SDC2130. When we change the pwm dutty (10Khz, 12Khz…20Khz) we change the interference in the video signal.

We share control power(12V) with other systems (controller+PCB+camera), and GND’s (12V and 24V) are join in the controller and are join in other point (near of main power supply).

We have well protected cables (twisted: power wire, shielded: for signal lines)...Can we install some type of filter to prevent such interference ? For example, One capacitor in 24V input ?


Thanks,

Please Log in or Create an account to join the conversation.

6 years 11 months ago #29528342 by roboteq
Replied by roboteq on topic SDC2130 PWM interference
You are doing the right things but eliminating the interferences will be very difficult. The controller switches with very high rise and fall times and this will generate high frequency harmonics and that cannot be changed. The main emission source will be the motor wires.

At this point, you just need to try all the tricks you can think of.

Try separating the power supply entirely. Try adding noise suppressing ferrites around the motor wires.

Please Log in or Create an account to join the conversation.

6 years 11 months ago #29528343 by LROBBINS
Replied by LROBBINS on topic SDC2130 PWM interference
What sort of bypass capacitors do you have on the power leads to the other devices - for example the camera? You might want to add a low-ESR electrolytic (e.g. >= 10 mF with ample WVDC headroom) and a ceramic (100 nF). You might also want to add a ferite around the power wires of the camera just where the leads enter it. Do you have a 100 nF bypass for Vdd and Vss of every chip in your board? "roboteq" is quite right that this can be a difficult noise to get rid of. While it's usually better to filter at the source, sometimes one has to filter at the other end as well.
Ciao,
Lenny
The following user(s) said Thank You: infoa

Please Log in or Create an account to join the conversation.

6 years 11 months ago #29528348 by infoa
Replied by infoa on topic SDC2130 PWM interference
Thanks you for your suggestions.

The PCB have some ICs, each IC have two capacitators: tantalo and ceramic.

The general 12Volts have one aluminium electrolytic capacitors (220uF/63V, low-ESR) and one ceramic multilayer capacitor (100nF). From this point my PCB, my controller and other circuit are powered. 24Volts line have our capacitator(2200uF/63V) near the controller.

I'll try separate both power supply, and I'll use ferrites around the motor wires.

A general question: does the 24 volts source and 12 volts source should have their GND joined to the chassis of a vehicle? What is your recommendation for this controller ?

Please Log in or Create an account to join the conversation.

6 years 11 months ago #29528349 by LROBBINS
Replied by LROBBINS on topic SDC2130 PWM interference
It does sound as though your bypassing is pretty thorough already.

I am familiar with only one kind of electric vehicle - wheelchairs - and for these both RESNA and ISO standards say that there must be NO low-impedance connection between any part of the electronics, including B-, and the frame. If there were a connection from B- to the frame, a short anywhere, especially anywhere upstream of fusing, could cause a fire, and a wheelchair driver can't hop off if a fire starts. Automobiles use the body as the ground return, but the high current wiring is short and (somewhat) protected by a fusible link or fuse and one assumes that if worst comes to worse the people can get out.

Of course, even though the frame is isolated from the electronics, that doesn't mean that the parts of the frame shouldn't be bonded to each other, and it doesn't hurt to have the electronics inside metal boxes, insulated from the electronics, that are connected to the frame. At least by the rules, one could also have a capacitor from ground to frame, but I've never seen one used on a wheelchair.

Do consider using a ferrite on the leads that enter your remote boxes. For example, Dynamic Controls uses one, with a double pass of wires (Batt+, Batt-, CANhi and CANlo) just inside the ACU joystick box.

Another possibility, if worst comes to worst, is to add common-mode filtering, but choosing the inductors and capacitors for that is not trivial (and beyond my ready recollection; I guess I've been lucky as I've not yet had to do that). Dynamic did have one control module on which they added common-mode filtering on the CAN lines. It reduced the noise, but it shifted the CAN timing enough that in some multi-node networks the bus became unreliable and they eventually stopped producing that module.

Perhaps someone else here will be able to offer you other suggestions.

Ciao,
Lenny

Please Log in or Create an account to join the conversation.

Moderators: tonysantoni
Time to create page: 0.092 seconds